main - opencv on mbed

Users » joeverbout » Code » main

opencv on mbed

opencv2/calib3d.hpp@0:ea44dc9ed014, 2016-03-31 (annotated)

Committer:: joeverbout
Date:: Thu Mar 31 21:16:38 2016 +0000
Revision:: 0:ea44dc9ed014

OpenCV on mbed attempt

Who changed what in which revision?

User	Revision	Line number	New contents of line
joeverbout	0:ea44dc9ed014	1	/*M///////////////////////////////////////////////////////////////////////////////////////
joeverbout	0:ea44dc9ed014	2	//
joeverbout	0:ea44dc9ed014	3	// IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING.
joeverbout	0:ea44dc9ed014	4	//
joeverbout	0:ea44dc9ed014	5	// By downloading, copying, installing or using the software you agree to this license.
joeverbout	0:ea44dc9ed014	6	// If you do not agree to this license, do not download, install,
joeverbout	0:ea44dc9ed014	7	// copy or use the software.
joeverbout	0:ea44dc9ed014	8	//
joeverbout	0:ea44dc9ed014	9	//
joeverbout	0:ea44dc9ed014	10	// License Agreement
joeverbout	0:ea44dc9ed014	11	// For Open Source Computer Vision Library
joeverbout	0:ea44dc9ed014	12	//
joeverbout	0:ea44dc9ed014	13	// Copyright (C) 2000-2008, Intel Corporation, all rights reserved.
joeverbout	0:ea44dc9ed014	14	// Copyright (C) 2009, Willow Garage Inc., all rights reserved.
joeverbout	0:ea44dc9ed014	15	// Copyright (C) 2013, OpenCV Foundation, all rights reserved.
joeverbout	0:ea44dc9ed014	16	// Third party copyrights are property of their respective owners.
joeverbout	0:ea44dc9ed014	17	//
joeverbout	0:ea44dc9ed014	18	// Redistribution and use in source and binary forms, with or without modification,
joeverbout	0:ea44dc9ed014	19	// are permitted provided that the following conditions are met:
joeverbout	0:ea44dc9ed014	20	//
joeverbout	0:ea44dc9ed014	21	// * Redistribution's of source code must retain the above copyright notice,
joeverbout	0:ea44dc9ed014	22	// this list of conditions and the following disclaimer.
joeverbout	0:ea44dc9ed014	23	//
joeverbout	0:ea44dc9ed014	24	// * Redistribution's in binary form must reproduce the above copyright notice,
joeverbout	0:ea44dc9ed014	25	// this list of conditions and the following disclaimer in the documentation
joeverbout	0:ea44dc9ed014	26	// and/or other materials provided with the distribution.
joeverbout	0:ea44dc9ed014	27	//
joeverbout	0:ea44dc9ed014	28	// * The name of the copyright holders may not be used to endorse or promote products
joeverbout	0:ea44dc9ed014	29	// derived from this software without specific prior written permission.
joeverbout	0:ea44dc9ed014	30	//
joeverbout	0:ea44dc9ed014	31	// This software is provided by the copyright holders and contributors "as is" and
joeverbout	0:ea44dc9ed014	32	// any express or implied warranties, including, but not limited to, the implied
joeverbout	0:ea44dc9ed014	33	// warranties of merchantability and fitness for a particular purpose are disclaimed.
joeverbout	0:ea44dc9ed014	34	// In no event shall the Intel Corporation or contributors be liable for any direct,
joeverbout	0:ea44dc9ed014	35	// indirect, incidental, special, exemplary, or consequential damages
joeverbout	0:ea44dc9ed014	36	// (including, but not limited to, procurement of substitute goods or services;
joeverbout	0:ea44dc9ed014	37	// loss of use, data, or profits; or business interruption) however caused
joeverbout	0:ea44dc9ed014	38	// and on any theory of liability, whether in contract, strict liability,
joeverbout	0:ea44dc9ed014	39	// or tort (including negligence or otherwise) arising in any way out of
joeverbout	0:ea44dc9ed014	40	// the use of this software, even if advised of the possibility of such damage.
joeverbout	0:ea44dc9ed014	41	//
joeverbout	0:ea44dc9ed014	42	//M*/
joeverbout	0:ea44dc9ed014	43
joeverbout	0:ea44dc9ed014	44	#ifndef __OPENCV_CALIB3D_HPP__
joeverbout	0:ea44dc9ed014	45	#define __OPENCV_CALIB3D_HPP__
joeverbout	0:ea44dc9ed014	46
joeverbout	0:ea44dc9ed014	47	#include "opencv2/core.hpp"
joeverbout	0:ea44dc9ed014	48	#include "opencv2/features2d.hpp"
joeverbout	0:ea44dc9ed014	49	#include "opencv2/core/affine.hpp"
joeverbout	0:ea44dc9ed014	50
joeverbout	0:ea44dc9ed014	51	/**
joeverbout	0:ea44dc9ed014	52	@defgroup calib3d Camera Calibration and 3D Reconstruction
joeverbout	0:ea44dc9ed014	53
joeverbout	0:ea44dc9ed014	54	The functions in this section use a so-called pinhole camera model. In this model, a scene view is
joeverbout	0:ea44dc9ed014	55	formed by projecting 3D points into the image plane using a perspective transformation.
joeverbout	0:ea44dc9ed014	56
joeverbout	0:ea44dc9ed014	57	\f[s \; m' = A [R\|t] M'\f]
joeverbout	0:ea44dc9ed014	58
joeverbout	0:ea44dc9ed014	59	or
joeverbout	0:ea44dc9ed014	60
joeverbout	0:ea44dc9ed014	61	\f[s \vecthree{u}{v}{1} = \vecthreethree{f_x}{0}{c_x}{0}{f_y}{c_y}{0}{0}{1}
joeverbout	0:ea44dc9ed014	62	\begin{bmatrix}
joeverbout	0:ea44dc9ed014	63	r_{11} & r_{12} & r_{13} & t_1 \\
joeverbout	0:ea44dc9ed014	64	r_{21} & r_{22} & r_{23} & t_2 \\
joeverbout	0:ea44dc9ed014	65	r_{31} & r_{32} & r_{33} & t_3
joeverbout	0:ea44dc9ed014	66	\end{bmatrix}
joeverbout	0:ea44dc9ed014	67	\begin{bmatrix}
joeverbout	0:ea44dc9ed014	68	X \\
joeverbout	0:ea44dc9ed014	69	Y \\
joeverbout	0:ea44dc9ed014	70	Z \\
joeverbout	0:ea44dc9ed014	71	1
joeverbout	0:ea44dc9ed014	72	\end{bmatrix}\f]
joeverbout	0:ea44dc9ed014	73
joeverbout	0:ea44dc9ed014	74	where:
joeverbout	0:ea44dc9ed014	75
joeverbout	0:ea44dc9ed014	76	- \f$(X, Y, Z)\f$ are the coordinates of a 3D point in the world coordinate space
joeverbout	0:ea44dc9ed014	77	- \f$(u, v)\f$ are the coordinates of the projection point in pixels
joeverbout	0:ea44dc9ed014	78	- \f$A\f$ is a camera matrix, or a matrix of intrinsic parameters
joeverbout	0:ea44dc9ed014	79	- \f$(cx, cy)\f$ is a principal point that is usually at the image center
joeverbout	0:ea44dc9ed014	80	- \f$fx, fy\f$ are the focal lengths expressed in pixel units.
joeverbout	0:ea44dc9ed014	81
joeverbout	0:ea44dc9ed014	82	Thus, if an image from the camera is scaled by a factor, all of these parameters should be scaled
joeverbout	0:ea44dc9ed014	83	(multiplied/divided, respectively) by the same factor. The matrix of intrinsic parameters does not
joeverbout	0:ea44dc9ed014	84	depend on the scene viewed. So, once estimated, it can be re-used as long as the focal length is
joeverbout	0:ea44dc9ed014	85	fixed (in case of zoom lens). The joint rotation-translation matrix \f$[R\|t]\f$ is called a matrix of
joeverbout	0:ea44dc9ed014	86	extrinsic parameters. It is used to describe the camera motion around a static scene, or vice versa,
joeverbout	0:ea44dc9ed014	87	rigid motion of an object in front of a still camera. That is, \f$[R\|t]\f$ translates coordinates of a
joeverbout	0:ea44dc9ed014	88	point \f$(X, Y, Z)\f$ to a coordinate system, fixed with respect to the camera. The transformation above
joeverbout	0:ea44dc9ed014	89	is equivalent to the following (when \f$z \ne 0\f$ ):
joeverbout	0:ea44dc9ed014	90
joeverbout	0:ea44dc9ed014	91	\f[\begin{array}{l}
joeverbout	0:ea44dc9ed014	92	\vecthree{x}{y}{z} = R \vecthree{X}{Y}{Z} + t \\
joeverbout	0:ea44dc9ed014	93	x' = x/z \\
joeverbout	0:ea44dc9ed014	94	y' = y/z \\
joeverbout	0:ea44dc9ed014	95	u = f_x*x' + c_x \\
joeverbout	0:ea44dc9ed014	96	v = f_y*y' + c_y
joeverbout	0:ea44dc9ed014	97	\end{array}\f]
joeverbout	0:ea44dc9ed014	98
joeverbout	0:ea44dc9ed014	99	Real lenses usually have some distortion, mostly radial distortion and slight tangential distortion.
joeverbout	0:ea44dc9ed014	100	So, the above model is extended as:
joeverbout	0:ea44dc9ed014	101
joeverbout	0:ea44dc9ed014	102	\f[\begin{array}{l}
joeverbout	0:ea44dc9ed014	103	\vecthree{x}{y}{z} = R \vecthree{X}{Y}{Z} + t \\
joeverbout	0:ea44dc9ed014	104	x' = x/z \\
joeverbout	0:ea44dc9ed014	105	y' = y/z \\
joeverbout	0:ea44dc9ed014	106	x'' = x' \frac{1 + k_1 r^2 + k_2 r^4 + k_3 r^6}{1 + k_4 r^2 + k_5 r^4 + k_6 r^6} + 2 p_1 x' y' + p_2(r^2 + 2 x'^2) + s_1 r^2 + s_2 r^4 \\
joeverbout	0:ea44dc9ed014	107	y'' = y' \frac{1 + k_1 r^2 + k_2 r^4 + k_3 r^6}{1 + k_4 r^2 + k_5 r^4 + k_6 r^6} + p_1 (r^2 + 2 y'^2) + 2 p_2 x' y' + s_3 r^2 + s_4 r^4 \\
joeverbout	0:ea44dc9ed014	108	\text{where} \quad r^2 = x'^2 + y'^2 \\
joeverbout	0:ea44dc9ed014	109	u = f_x*x'' + c_x \\
joeverbout	0:ea44dc9ed014	110	v = f_y*y'' + c_y
joeverbout	0:ea44dc9ed014	111	\end{array}\f]
joeverbout	0:ea44dc9ed014	112
joeverbout	0:ea44dc9ed014	113	\f$k_1\f$, \f$k_2\f$, \f$k_3\f$, \f$k_4\f$, \f$k_5\f$, and \f$k_6\f$ are radial distortion coefficients. \f$p_1\f$ and \f$p_2\f$ are
joeverbout	0:ea44dc9ed014	114	tangential distortion coefficients. \f$s_1\f$, \f$s_2\f$, \f$s_3\f$, and \f$s_4\f$, are the thin prism distortion
joeverbout	0:ea44dc9ed014	115	coefficients. Higher-order coefficients are not considered in OpenCV.
joeverbout	0:ea44dc9ed014	116
joeverbout	0:ea44dc9ed014	117	In some cases the image sensor may be tilted in order to focus an oblique plane in front of the
joeverbout	0:ea44dc9ed014	118	camera (Scheimpfug condition). This can be useful for particle image velocimetry (PIV) or
joeverbout	0:ea44dc9ed014	119	triangulation with a laser fan. The tilt causes a perspective distortion of \f$x''\f$ and
joeverbout	0:ea44dc9ed014	120	\f$y''\f$. This distortion can be modelled in the following way, see e.g. @cite Louhichi07.
joeverbout	0:ea44dc9ed014	121
joeverbout	0:ea44dc9ed014	122	\f[\begin{array}{l}
joeverbout	0:ea44dc9ed014	123	s\vecthree{x'''}{y'''}{1} =
joeverbout	0:ea44dc9ed014	124	\vecthreethree{R_{33}(\tau_x, \tau_y)}{0}{-R_{13}(\tau_x, \tau_y)}
joeverbout	0:ea44dc9ed014	125	{0}{R_{33}(\tau_x, \tau_y)}{-R_{23}(\tau_x, \tau_y)}
joeverbout	0:ea44dc9ed014	126	{0}{0}{1} R(\tau_x, \tau_y) \vecthree{x''}{y''}{1}\\
joeverbout	0:ea44dc9ed014	127	u = f_x*x''' + c_x \\
joeverbout	0:ea44dc9ed014	128	v = f_y*y''' + c_y
joeverbout	0:ea44dc9ed014	129	\end{array}\f]
joeverbout	0:ea44dc9ed014	130
joeverbout	0:ea44dc9ed014	131	where the matrix \f$R(\tau_x, \tau_y)\f$ is defined by two rotations with angular parameter \f$\tau_x\f$
joeverbout	0:ea44dc9ed014	132	and \f$\tau_y\f$, respectively,
joeverbout	0:ea44dc9ed014	133
joeverbout	0:ea44dc9ed014	134	\f[
joeverbout	0:ea44dc9ed014	135	R(\tau_x, \tau_y) =
joeverbout	0:ea44dc9ed014	136	\vecthreethree{\cos(\tau_y)}{0}{-\sin(\tau_y)}{0}{1}{0}{\sin(\tau_y)}{0}{\cos(\tau_y)}
joeverbout	0:ea44dc9ed014	137	\vecthreethree{1}{0}{0}{0}{\cos(\tau_x)}{\sin(\tau_x)}{0}{-\sin(\tau_x)}{\cos(\tau_x)} =
joeverbout	0:ea44dc9ed014	138	\vecthreethree{\cos(\tau_y)}{\sin(\tau_y)\sin(\tau_x)}{-\sin(\tau_y)\cos(\tau_x)}
joeverbout	0:ea44dc9ed014	139	{0}{\cos(\tau_x)}{\sin(\tau_x)}
joeverbout	0:ea44dc9ed014	140	{\sin(\tau_y)}{-\cos(\tau_y)\sin(\tau_x)}{\cos(\tau_y)\cos(\tau_x)}.
joeverbout	0:ea44dc9ed014	141	\f]
joeverbout	0:ea44dc9ed014	142
joeverbout	0:ea44dc9ed014	143	In the functions below the coefficients are passed or returned as
joeverbout	0:ea44dc9ed014	144
joeverbout	0:ea44dc9ed014	145	\f[(k_1, k_2, p_1, p_2[, k_3[, k_4, k_5, k_6 [, s_1, s_2, s_3, s_4[, \tau_x, \tau_y]]]])\f]
joeverbout	0:ea44dc9ed014	146
joeverbout	0:ea44dc9ed014	147	vector. That is, if the vector contains four elements, it means that \f$k_3=0\f$ . The distortion
joeverbout	0:ea44dc9ed014	148	coefficients do not depend on the scene viewed. Thus, they also belong to the intrinsic camera
joeverbout	0:ea44dc9ed014	149	parameters. And they remain the same regardless of the captured image resolution. If, for example, a
joeverbout	0:ea44dc9ed014	150	camera has been calibrated on images of 320 x 240 resolution, absolutely the same distortion
joeverbout	0:ea44dc9ed014	151	coefficients can be used for 640 x 480 images from the same camera while \f$f_x\f$, \f$f_y\f$, \f$c_x\f$, and
joeverbout	0:ea44dc9ed014	152	\f$c_y\f$ need to be scaled appropriately.
joeverbout	0:ea44dc9ed014	153
joeverbout	0:ea44dc9ed014	154	The functions below use the above model to do the following:
joeverbout	0:ea44dc9ed014	155
joeverbout	0:ea44dc9ed014	156	- Project 3D points to the image plane given intrinsic and extrinsic parameters.
joeverbout	0:ea44dc9ed014	157	- Compute extrinsic parameters given intrinsic parameters, a few 3D points, and their
joeverbout	0:ea44dc9ed014	158	projections.
joeverbout	0:ea44dc9ed014	159	- Estimate intrinsic and extrinsic camera parameters from several views of a known calibration
joeverbout	0:ea44dc9ed014	160	pattern (every view is described by several 3D-2D point correspondences).
joeverbout	0:ea44dc9ed014	161	- Estimate the relative position and orientation of the stereo camera "heads" and compute the
joeverbout	0:ea44dc9ed014	162	rectification transformation that makes the camera optical axes parallel.
joeverbout	0:ea44dc9ed014	163
joeverbout	0:ea44dc9ed014	164	@note
joeverbout	0:ea44dc9ed014	165	- A calibration sample for 3 cameras in horizontal position can be found at
joeverbout	0:ea44dc9ed014	166	opencv_source_code/samples/cpp/3calibration.cpp
joeverbout	0:ea44dc9ed014	167	- A calibration sample based on a sequence of images can be found at
joeverbout	0:ea44dc9ed014	168	opencv_source_code/samples/cpp/calibration.cpp
joeverbout	0:ea44dc9ed014	169	- A calibration sample in order to do 3D reconstruction can be found at
joeverbout	0:ea44dc9ed014	170	opencv_source_code/samples/cpp/build3dmodel.cpp
joeverbout	0:ea44dc9ed014	171	- A calibration sample of an artificially generated camera and chessboard patterns can be
joeverbout	0:ea44dc9ed014	172	found at opencv_source_code/samples/cpp/calibration_artificial.cpp
joeverbout	0:ea44dc9ed014	173	- A calibration example on stereo calibration can be found at
joeverbout	0:ea44dc9ed014	174	opencv_source_code/samples/cpp/stereo_calib.cpp
joeverbout	0:ea44dc9ed014	175	- A calibration example on stereo matching can be found at
joeverbout	0:ea44dc9ed014	176	opencv_source_code/samples/cpp/stereo_match.cpp
joeverbout	0:ea44dc9ed014	177	- (Python) A camera calibration sample can be found at
joeverbout	0:ea44dc9ed014	178	opencv_source_code/samples/python/calibrate.py
joeverbout	0:ea44dc9ed014	179
joeverbout	0:ea44dc9ed014	180	@{
joeverbout	0:ea44dc9ed014	181	@defgroup calib3d_fisheye Fisheye camera model
joeverbout	0:ea44dc9ed014	182
joeverbout	0:ea44dc9ed014	183	Definitions: Let P be a point in 3D of coordinates X in the world reference frame (stored in the
joeverbout	0:ea44dc9ed014	184	matrix X) The coordinate vector of P in the camera reference frame is:
joeverbout	0:ea44dc9ed014	185
joeverbout	0:ea44dc9ed014	186	\f[Xc = R X + T\f]
joeverbout	0:ea44dc9ed014	187
joeverbout	0:ea44dc9ed014	188	where R is the rotation matrix corresponding to the rotation vector om: R = rodrigues(om); call x, y
joeverbout	0:ea44dc9ed014	189	and z the 3 coordinates of Xc:
joeverbout	0:ea44dc9ed014	190
joeverbout	0:ea44dc9ed014	191	\f[x = Xc_1 \\ y = Xc_2 \\ z = Xc_3\f]
joeverbout	0:ea44dc9ed014	192
joeverbout	0:ea44dc9ed014	193	The pinehole projection coordinates of P is [a; b] where
joeverbout	0:ea44dc9ed014	194
joeverbout	0:ea44dc9ed014	195	\f[a = x / z \ and \ b = y / z \\ r^2 = a^2 + b^2 \\ \theta = atan(r)\f]
joeverbout	0:ea44dc9ed014	196
joeverbout	0:ea44dc9ed014	197	Fisheye distortion:
joeverbout	0:ea44dc9ed014	198
joeverbout	0:ea44dc9ed014	199	\f[\theta_d = \theta (1 + k_1 \theta^2 + k_2 \theta^4 + k_3 \theta^6 + k_4 \theta^8)\f]
joeverbout	0:ea44dc9ed014	200
joeverbout	0:ea44dc9ed014	201	The distorted point coordinates are [x'; y'] where
joeverbout	0:ea44dc9ed014	202
joeverbout	0:ea44dc9ed014	203	\f[x' = (\theta_d / r) x \\ y' = (\theta_d / r) y \f]
joeverbout	0:ea44dc9ed014	204
joeverbout	0:ea44dc9ed014	205	Finally, conversion into pixel coordinates: The final pixel coordinates vector [u; v] where:
joeverbout	0:ea44dc9ed014	206
joeverbout	0:ea44dc9ed014	207	\f[u = f_x (x' + \alpha y') + c_x \\
joeverbout	0:ea44dc9ed014	208	v = f_y yy + c_y\f]
joeverbout	0:ea44dc9ed014	209
joeverbout	0:ea44dc9ed014	210	@defgroup calib3d_c C API
joeverbout	0:ea44dc9ed014	211
joeverbout	0:ea44dc9ed014	212	@}
joeverbout	0:ea44dc9ed014	213	*/
joeverbout	0:ea44dc9ed014	214
joeverbout	0:ea44dc9ed014	215	namespace cv
joeverbout	0:ea44dc9ed014	216	{
joeverbout	0:ea44dc9ed014	217
joeverbout	0:ea44dc9ed014	218	//! @addtogroup calib3d
joeverbout	0:ea44dc9ed014	219	//! @{
joeverbout	0:ea44dc9ed014	220
joeverbout	0:ea44dc9ed014	221	//! type of the robust estimation algorithm
joeverbout	0:ea44dc9ed014	222	enum { LMEDS = 4, //!< least-median algorithm
joeverbout	0:ea44dc9ed014	223	RANSAC = 8, //!< RANSAC algorithm
joeverbout	0:ea44dc9ed014	224	RHO = 16 //!< RHO algorithm
joeverbout	0:ea44dc9ed014	225	};
joeverbout	0:ea44dc9ed014	226
joeverbout	0:ea44dc9ed014	227	enum { SOLVEPNP_ITERATIVE = 0,
joeverbout	0:ea44dc9ed014	228	SOLVEPNP_EPNP = 1, //!< EPnP: Efficient Perspective-n-Point Camera Pose Estimation @cite lepetit2009epnp
joeverbout	0:ea44dc9ed014	229	SOLVEPNP_P3P = 2, //!< Complete Solution Classification for the Perspective-Three-Point Problem @cite gao2003complete
joeverbout	0:ea44dc9ed014	230	SOLVEPNP_DLS = 3, //!< A Direct Least-Squares (DLS) Method for PnP @cite hesch2011direct
joeverbout	0:ea44dc9ed014	231	SOLVEPNP_UPNP = 4 //!< Exhaustive Linearization for Robust Camera Pose and Focal Length Estimation @cite penate2013exhaustive
joeverbout	0:ea44dc9ed014	232
joeverbout	0:ea44dc9ed014	233	};
joeverbout	0:ea44dc9ed014	234
joeverbout	0:ea44dc9ed014	235	enum { CALIB_CB_ADAPTIVE_THRESH = 1,
joeverbout	0:ea44dc9ed014	236	CALIB_CB_NORMALIZE_IMAGE = 2,
joeverbout	0:ea44dc9ed014	237	CALIB_CB_FILTER_QUADS = 4,
joeverbout	0:ea44dc9ed014	238	CALIB_CB_FAST_CHECK = 8
joeverbout	0:ea44dc9ed014	239	};
joeverbout	0:ea44dc9ed014	240
joeverbout	0:ea44dc9ed014	241	enum { CALIB_CB_SYMMETRIC_GRID = 1,
joeverbout	0:ea44dc9ed014	242	CALIB_CB_ASYMMETRIC_GRID = 2,
joeverbout	0:ea44dc9ed014	243	CALIB_CB_CLUSTERING = 4
joeverbout	0:ea44dc9ed014	244	};
joeverbout	0:ea44dc9ed014	245
joeverbout	0:ea44dc9ed014	246	enum { CALIB_USE_INTRINSIC_GUESS = 0x00001,
joeverbout	0:ea44dc9ed014	247	CALIB_FIX_ASPECT_RATIO = 0x00002,
joeverbout	0:ea44dc9ed014	248	CALIB_FIX_PRINCIPAL_POINT = 0x00004,
joeverbout	0:ea44dc9ed014	249	CALIB_ZERO_TANGENT_DIST = 0x00008,
joeverbout	0:ea44dc9ed014	250	CALIB_FIX_FOCAL_LENGTH = 0x00010,
joeverbout	0:ea44dc9ed014	251	CALIB_FIX_K1 = 0x00020,
joeverbout	0:ea44dc9ed014	252	CALIB_FIX_K2 = 0x00040,
joeverbout	0:ea44dc9ed014	253	CALIB_FIX_K3 = 0x00080,
joeverbout	0:ea44dc9ed014	254	CALIB_FIX_K4 = 0x00800,
joeverbout	0:ea44dc9ed014	255	CALIB_FIX_K5 = 0x01000,
joeverbout	0:ea44dc9ed014	256	CALIB_FIX_K6 = 0x02000,
joeverbout	0:ea44dc9ed014	257	CALIB_RATIONAL_MODEL = 0x04000,
joeverbout	0:ea44dc9ed014	258	CALIB_THIN_PRISM_MODEL = 0x08000,
joeverbout	0:ea44dc9ed014	259	CALIB_FIX_S1_S2_S3_S4 = 0x10000,
joeverbout	0:ea44dc9ed014	260	CALIB_TILTED_MODEL = 0x40000,
joeverbout	0:ea44dc9ed014	261	CALIB_FIX_TAUX_TAUY = 0x80000,
joeverbout	0:ea44dc9ed014	262	// only for stereo
joeverbout	0:ea44dc9ed014	263	CALIB_FIX_INTRINSIC = 0x00100,
joeverbout	0:ea44dc9ed014	264	CALIB_SAME_FOCAL_LENGTH = 0x00200,
joeverbout	0:ea44dc9ed014	265	// for stereo rectification
joeverbout	0:ea44dc9ed014	266	CALIB_ZERO_DISPARITY = 0x00400,
joeverbout	0:ea44dc9ed014	267	CALIB_USE_LU = (1 << 17), //!< use LU instead of SVD decomposition for solving. much faster but potentially less precise
joeverbout	0:ea44dc9ed014	268	};
joeverbout	0:ea44dc9ed014	269
joeverbout	0:ea44dc9ed014	270	//! the algorithm for finding fundamental matrix
joeverbout	0:ea44dc9ed014	271	enum { FM_7POINT = 1, //!< 7-point algorithm
joeverbout	0:ea44dc9ed014	272	FM_8POINT = 2, //!< 8-point algorithm
joeverbout	0:ea44dc9ed014	273	FM_LMEDS = 4, //!< least-median algorithm
joeverbout	0:ea44dc9ed014	274	FM_RANSAC = 8 //!< RANSAC algorithm
joeverbout	0:ea44dc9ed014	275	};
joeverbout	0:ea44dc9ed014	276
joeverbout	0:ea44dc9ed014	277
joeverbout	0:ea44dc9ed014	278
joeverbout	0:ea44dc9ed014	279	/** @brief Converts a rotation matrix to a rotation vector or vice versa.
joeverbout	0:ea44dc9ed014	280
joeverbout	0:ea44dc9ed014	281	@param src Input rotation vector (3x1 or 1x3) or rotation matrix (3x3).
joeverbout	0:ea44dc9ed014	282	@param dst Output rotation matrix (3x3) or rotation vector (3x1 or 1x3), respectively.
joeverbout	0:ea44dc9ed014	283	@param jacobian Optional output Jacobian matrix, 3x9 or 9x3, which is a matrix of partial
joeverbout	0:ea44dc9ed014	284	derivatives of the output array components with respect to the input array components.
joeverbout	0:ea44dc9ed014	285
joeverbout	0:ea44dc9ed014	286	\f[\begin{array}{l} \theta \leftarrow norm(r) \\ r \leftarrow r/ \theta \\ R = \cos{\theta} I + (1- \cos{\theta} ) r r^T + \sin{\theta} \vecthreethree{0}{-r_z}{r_y}{r_z}{0}{-r_x}{-r_y}{r_x}{0} \end{array}\f]
joeverbout	0:ea44dc9ed014	287
joeverbout	0:ea44dc9ed014	288	Inverse transformation can be also done easily, since
joeverbout	0:ea44dc9ed014	289
joeverbout	0:ea44dc9ed014	290	\f[\sin ( \theta ) \vecthreethree{0}{-r_z}{r_y}{r_z}{0}{-r_x}{-r_y}{r_x}{0} = \frac{R - R^T}{2}\f]
joeverbout	0:ea44dc9ed014	291
joeverbout	0:ea44dc9ed014	292	A rotation vector is a convenient and most compact representation of a rotation matrix (since any
joeverbout	0:ea44dc9ed014	293	rotation matrix has just 3 degrees of freedom). The representation is used in the global 3D geometry
joeverbout	0:ea44dc9ed014	294	optimization procedures like calibrateCamera, stereoCalibrate, or solvePnP .
joeverbout	0:ea44dc9ed014	295	*/
joeverbout	0:ea44dc9ed014	296	CV_EXPORTS_W void Rodrigues( InputArray src, OutputArray dst, OutputArray jacobian = noArray() );
joeverbout	0:ea44dc9ed014	297
joeverbout	0:ea44dc9ed014	298	/** @brief Finds a perspective transformation between two planes.
joeverbout	0:ea44dc9ed014	299
joeverbout	0:ea44dc9ed014	300	@param srcPoints Coordinates of the points in the original plane, a matrix of the type CV_32FC2
joeverbout	0:ea44dc9ed014	301	or vector\<Point2f\> .
joeverbout	0:ea44dc9ed014	302	@param dstPoints Coordinates of the points in the target plane, a matrix of the type CV_32FC2 or
joeverbout	0:ea44dc9ed014	303	a vector\<Point2f\> .
joeverbout	0:ea44dc9ed014	304	@param method Method used to computed a homography matrix. The following methods are possible:
joeverbout	0:ea44dc9ed014	305	- 0 - a regular method using all the points
joeverbout	0:ea44dc9ed014	306	- RANSAC - RANSAC-based robust method
joeverbout	0:ea44dc9ed014	307	- LMEDS - Least-Median robust method
joeverbout	0:ea44dc9ed014	308	- RHO - PROSAC-based robust method
joeverbout	0:ea44dc9ed014	309	@param ransacReprojThreshold Maximum allowed reprojection error to treat a point pair as an inlier
joeverbout	0:ea44dc9ed014	310	(used in the RANSAC and RHO methods only). That is, if
joeverbout	0:ea44dc9ed014	311	\f[\\| \texttt{dstPoints} _i - \texttt{convertPointsHomogeneous} ( \texttt{H} * \texttt{srcPoints} _i) \\| > \texttt{ransacReprojThreshold}\f]
joeverbout	0:ea44dc9ed014	312	then the point \f$i\f$ is considered an outlier. If srcPoints and dstPoints are measured in pixels,
joeverbout	0:ea44dc9ed014	313	it usually makes sense to set this parameter somewhere in the range of 1 to 10.
joeverbout	0:ea44dc9ed014	314	@param mask Optional output mask set by a robust method ( RANSAC or LMEDS ). Note that the input
joeverbout	0:ea44dc9ed014	315	mask values are ignored.
joeverbout	0:ea44dc9ed014	316	@param maxIters The maximum number of RANSAC iterations, 2000 is the maximum it can be.
joeverbout	0:ea44dc9ed014	317	@param confidence Confidence level, between 0 and 1.
joeverbout	0:ea44dc9ed014	318
joeverbout	0:ea44dc9ed014	319	The functions find and return the perspective transformation \f$H\f$ between the source and the
joeverbout	0:ea44dc9ed014	320	destination planes:
joeverbout	0:ea44dc9ed014	321
joeverbout	0:ea44dc9ed014	322	\f[s_i \vecthree{x'_i}{y'_i}{1} \sim H \vecthree{x_i}{y_i}{1}\f]
joeverbout	0:ea44dc9ed014	323
joeverbout	0:ea44dc9ed014	324	so that the back-projection error
joeverbout	0:ea44dc9ed014	325
joeverbout	0:ea44dc9ed014	326	\f[\sum _i \left ( x'_i- \frac{h_{11} x_i + h_{12} y_i + h_{13}}{h_{31} x_i + h_{32} y_i + h_{33}} \right )^2+ \left ( y'_i- \frac{h_{21} x_i + h_{22} y_i + h_{23}}{h_{31} x_i + h_{32} y_i + h_{33}} \right )^2\f]
joeverbout	0:ea44dc9ed014	327
joeverbout	0:ea44dc9ed014	328	is minimized. If the parameter method is set to the default value 0, the function uses all the point
joeverbout	0:ea44dc9ed014	329	pairs to compute an initial homography estimate with a simple least-squares scheme.
joeverbout	0:ea44dc9ed014	330
joeverbout	0:ea44dc9ed014	331	However, if not all of the point pairs ( \f$srcPoints_i\f$, \f$dstPoints_i\f$ ) fit the rigid perspective
joeverbout	0:ea44dc9ed014	332	transformation (that is, there are some outliers), this initial estimate will be poor. In this case,
joeverbout	0:ea44dc9ed014	333	you can use one of the three robust methods. The methods RANSAC, LMeDS and RHO try many different
joeverbout	0:ea44dc9ed014	334	random subsets of the corresponding point pairs (of four pairs each), estimate the homography matrix
joeverbout	0:ea44dc9ed014	335	using this subset and a simple least-square algorithm, and then compute the quality/goodness of the
joeverbout	0:ea44dc9ed014	336	computed homography (which is the number of inliers for RANSAC or the median re-projection error for
joeverbout	0:ea44dc9ed014	337	LMeDs). The best subset is then used to produce the initial estimate of the homography matrix and
joeverbout	0:ea44dc9ed014	338	the mask of inliers/outliers.
joeverbout	0:ea44dc9ed014	339
joeverbout	0:ea44dc9ed014	340	Regardless of the method, robust or not, the computed homography matrix is refined further (using
joeverbout	0:ea44dc9ed014	341	inliers only in case of a robust method) with the Levenberg-Marquardt method to reduce the
joeverbout	0:ea44dc9ed014	342	re-projection error even more.
joeverbout	0:ea44dc9ed014	343
joeverbout	0:ea44dc9ed014	344	The methods RANSAC and RHO can handle practically any ratio of outliers but need a threshold to
joeverbout	0:ea44dc9ed014	345	distinguish inliers from outliers. The method LMeDS does not need any threshold but it works
joeverbout	0:ea44dc9ed014	346	correctly only when there are more than 50% of inliers. Finally, if there are no outliers and the
joeverbout	0:ea44dc9ed014	347	noise is rather small, use the default method (method=0).
joeverbout	0:ea44dc9ed014	348
joeverbout	0:ea44dc9ed014	349	The function is used to find initial intrinsic and extrinsic matrices. Homography matrix is
joeverbout	0:ea44dc9ed014	350	determined up to a scale. Thus, it is normalized so that \f$h_{33}=1\f$. Note that whenever an H matrix
joeverbout	0:ea44dc9ed014	351	cannot be estimated, an empty one will be returned.
joeverbout	0:ea44dc9ed014	352
joeverbout	0:ea44dc9ed014	353	@sa
joeverbout	0:ea44dc9ed014	354	getAffineTransform, getPerspectiveTransform, estimateRigidTransform, warpPerspective,
joeverbout	0:ea44dc9ed014	355	perspectiveTransform
joeverbout	0:ea44dc9ed014	356
joeverbout	0:ea44dc9ed014	357	@note
joeverbout	0:ea44dc9ed014	358	- A example on calculating a homography for image matching can be found at
joeverbout	0:ea44dc9ed014	359	opencv_source_code/samples/cpp/video_homography.cpp
joeverbout	0:ea44dc9ed014	360
joeverbout	0:ea44dc9ed014	361	*/
joeverbout	0:ea44dc9ed014	362	CV_EXPORTS_W Mat findHomography( InputArray srcPoints, InputArray dstPoints,
joeverbout	0:ea44dc9ed014	363	int method = 0, double ransacReprojThreshold = 3,
joeverbout	0:ea44dc9ed014	364	OutputArray mask=noArray(), const int maxIters = 2000,
joeverbout	0:ea44dc9ed014	365	const double confidence = 0.995);
joeverbout	0:ea44dc9ed014	366
joeverbout	0:ea44dc9ed014	367	/** @overload */
joeverbout	0:ea44dc9ed014	368	CV_EXPORTS Mat findHomography( InputArray srcPoints, InputArray dstPoints,
joeverbout	0:ea44dc9ed014	369	OutputArray mask, int method = 0, double ransacReprojThreshold = 3 );
joeverbout	0:ea44dc9ed014	370
joeverbout	0:ea44dc9ed014	371	/** @brief Computes an RQ decomposition of 3x3 matrices.
joeverbout	0:ea44dc9ed014	372
joeverbout	0:ea44dc9ed014	373	@param src 3x3 input matrix.
joeverbout	0:ea44dc9ed014	374	@param mtxR Output 3x3 upper-triangular matrix.
joeverbout	0:ea44dc9ed014	375	@param mtxQ Output 3x3 orthogonal matrix.
joeverbout	0:ea44dc9ed014	376	@param Qx Optional output 3x3 rotation matrix around x-axis.
joeverbout	0:ea44dc9ed014	377	@param Qy Optional output 3x3 rotation matrix around y-axis.
joeverbout	0:ea44dc9ed014	378	@param Qz Optional output 3x3 rotation matrix around z-axis.
joeverbout	0:ea44dc9ed014	379
joeverbout	0:ea44dc9ed014	380	The function computes a RQ decomposition using the given rotations. This function is used in
joeverbout	0:ea44dc9ed014	381	decomposeProjectionMatrix to decompose the left 3x3 submatrix of a projection matrix into a camera
joeverbout	0:ea44dc9ed014	382	and a rotation matrix.
joeverbout	0:ea44dc9ed014	383
joeverbout	0:ea44dc9ed014	384	It optionally returns three rotation matrices, one for each axis, and the three Euler angles in
joeverbout	0:ea44dc9ed014	385	degrees (as the return value) that could be used in OpenGL. Note, there is always more than one
joeverbout	0:ea44dc9ed014	386	sequence of rotations about the three principle axes that results in the same orientation of an
joeverbout	0:ea44dc9ed014	387	object, eg. see @cite Slabaugh . Returned tree rotation matrices and corresponding three Euler angules
joeverbout	0:ea44dc9ed014	388	are only one of the possible solutions.
joeverbout	0:ea44dc9ed014	389	*/
joeverbout	0:ea44dc9ed014	390	CV_EXPORTS_W Vec3d RQDecomp3x3( InputArray src, OutputArray mtxR, OutputArray mtxQ,
joeverbout	0:ea44dc9ed014	391	OutputArray Qx = noArray(),
joeverbout	0:ea44dc9ed014	392	OutputArray Qy = noArray(),
joeverbout	0:ea44dc9ed014	393	OutputArray Qz = noArray());
joeverbout	0:ea44dc9ed014	394
joeverbout	0:ea44dc9ed014	395	/** @brief Decomposes a projection matrix into a rotation matrix and a camera matrix.
joeverbout	0:ea44dc9ed014	396
joeverbout	0:ea44dc9ed014	397	@param projMatrix 3x4 input projection matrix P.
joeverbout	0:ea44dc9ed014	398	@param cameraMatrix Output 3x3 camera matrix K.
joeverbout	0:ea44dc9ed014	399	@param rotMatrix Output 3x3 external rotation matrix R.
joeverbout	0:ea44dc9ed014	400	@param transVect Output 4x1 translation vector T.
joeverbout	0:ea44dc9ed014	401	@param rotMatrixX Optional 3x3 rotation matrix around x-axis.
joeverbout	0:ea44dc9ed014	402	@param rotMatrixY Optional 3x3 rotation matrix around y-axis.
joeverbout	0:ea44dc9ed014	403	@param rotMatrixZ Optional 3x3 rotation matrix around z-axis.
joeverbout	0:ea44dc9ed014	404	@param eulerAngles Optional three-element vector containing three Euler angles of rotation in
joeverbout	0:ea44dc9ed014	405	degrees.
joeverbout	0:ea44dc9ed014	406
joeverbout	0:ea44dc9ed014	407	The function computes a decomposition of a projection matrix into a calibration and a rotation
joeverbout	0:ea44dc9ed014	408	matrix and the position of a camera.
joeverbout	0:ea44dc9ed014	409
joeverbout	0:ea44dc9ed014	410	It optionally returns three rotation matrices, one for each axis, and three Euler angles that could
joeverbout	0:ea44dc9ed014	411	be used in OpenGL. Note, there is always more than one sequence of rotations about the three
joeverbout	0:ea44dc9ed014	412	principle axes that results in the same orientation of an object, eg. see @cite Slabaugh . Returned
joeverbout	0:ea44dc9ed014	413	tree rotation matrices and corresponding three Euler angules are only one of the possible solutions.
joeverbout	0:ea44dc9ed014	414
joeverbout	0:ea44dc9ed014	415	The function is based on RQDecomp3x3 .
joeverbout	0:ea44dc9ed014	416	*/
joeverbout	0:ea44dc9ed014	417	CV_EXPORTS_W void decomposeProjectionMatrix( InputArray projMatrix, OutputArray cameraMatrix,
joeverbout	0:ea44dc9ed014	418	OutputArray rotMatrix, OutputArray transVect,
joeverbout	0:ea44dc9ed014	419	OutputArray rotMatrixX = noArray(),
joeverbout	0:ea44dc9ed014	420	OutputArray rotMatrixY = noArray(),
joeverbout	0:ea44dc9ed014	421	OutputArray rotMatrixZ = noArray(),
joeverbout	0:ea44dc9ed014	422	OutputArray eulerAngles =noArray() );
joeverbout	0:ea44dc9ed014	423
joeverbout	0:ea44dc9ed014	424	/** @brief Computes partial derivatives of the matrix product for each multiplied matrix.
joeverbout	0:ea44dc9ed014	425
joeverbout	0:ea44dc9ed014	426	@param A First multiplied matrix.
joeverbout	0:ea44dc9ed014	427	@param B Second multiplied matrix.
joeverbout	0:ea44dc9ed014	428	@param dABdA First output derivative matrix d(A\*B)/dA of size
joeverbout	0:ea44dc9ed014	429	\f$\texttt{A.rowsB.cols} \times {A.rowsA.cols}\f$ .
joeverbout	0:ea44dc9ed014	430	@param dABdB Second output derivative matrix d(A\*B)/dB of size
joeverbout	0:ea44dc9ed014	431	\f$\texttt{A.rowsB.cols} \times {B.rowsB.cols}\f$ .
joeverbout	0:ea44dc9ed014	432
joeverbout	0:ea44dc9ed014	433	The function computes partial derivatives of the elements of the matrix product \f$A*B\f$ with regard to
joeverbout	0:ea44dc9ed014	434	the elements of each of the two input matrices. The function is used to compute the Jacobian
joeverbout	0:ea44dc9ed014	435	matrices in stereoCalibrate but can also be used in any other similar optimization function.
joeverbout	0:ea44dc9ed014	436	*/
joeverbout	0:ea44dc9ed014	437	CV_EXPORTS_W void matMulDeriv( InputArray A, InputArray B, OutputArray dABdA, OutputArray dABdB );
joeverbout	0:ea44dc9ed014	438
joeverbout	0:ea44dc9ed014	439	/** @brief Combines two rotation-and-shift transformations.
joeverbout	0:ea44dc9ed014	440
joeverbout	0:ea44dc9ed014	441	@param rvec1 First rotation vector.
joeverbout	0:ea44dc9ed014	442	@param tvec1 First translation vector.
joeverbout	0:ea44dc9ed014	443	@param rvec2 Second rotation vector.
joeverbout	0:ea44dc9ed014	444	@param tvec2 Second translation vector.
joeverbout	0:ea44dc9ed014	445	@param rvec3 Output rotation vector of the superposition.
joeverbout	0:ea44dc9ed014	446	@param tvec3 Output translation vector of the superposition.
joeverbout	0:ea44dc9ed014	447	@param dr3dr1
joeverbout	0:ea44dc9ed014	448	@param dr3dt1
joeverbout	0:ea44dc9ed014	449	@param dr3dr2
joeverbout	0:ea44dc9ed014	450	@param dr3dt2
joeverbout	0:ea44dc9ed014	451	@param dt3dr1
joeverbout	0:ea44dc9ed014	452	@param dt3dt1
joeverbout	0:ea44dc9ed014	453	@param dt3dr2
joeverbout	0:ea44dc9ed014	454	@param dt3dt2 Optional output derivatives of rvec3 or tvec3 with regard to rvec1, rvec2, tvec1 and
joeverbout	0:ea44dc9ed014	455	tvec2, respectively.
joeverbout	0:ea44dc9ed014	456
joeverbout	0:ea44dc9ed014	457	The functions compute:
joeverbout	0:ea44dc9ed014	458
joeverbout	0:ea44dc9ed014	459	\f[\begin{array}{l} \texttt{rvec3} = \mathrm{rodrigues} ^{-1} \left ( \mathrm{rodrigues} ( \texttt{rvec2} ) \cdot \mathrm{rodrigues} ( \texttt{rvec1} ) \right ) \\ \texttt{tvec3} = \mathrm{rodrigues} ( \texttt{rvec2} ) \cdot \texttt{tvec1} + \texttt{tvec2} \end{array} ,\f]
joeverbout	0:ea44dc9ed014	460
joeverbout	0:ea44dc9ed014	461	where \f$\mathrm{rodrigues}\f$ denotes a rotation vector to a rotation matrix transformation, and
joeverbout	0:ea44dc9ed014	462	\f$\mathrm{rodrigues}^{-1}\f$ denotes the inverse transformation. See Rodrigues for details.
joeverbout	0:ea44dc9ed014	463
joeverbout	0:ea44dc9ed014	464	Also, the functions can compute the derivatives of the output vectors with regards to the input
joeverbout	0:ea44dc9ed014	465	vectors (see matMulDeriv ). The functions are used inside stereoCalibrate but can also be used in
joeverbout	0:ea44dc9ed014	466	your own code where Levenberg-Marquardt or another gradient-based solver is used to optimize a
joeverbout	0:ea44dc9ed014	467	function that contains a matrix multiplication.
joeverbout	0:ea44dc9ed014	468	*/
joeverbout	0:ea44dc9ed014	469	CV_EXPORTS_W void composeRT( InputArray rvec1, InputArray tvec1,
joeverbout	0:ea44dc9ed014	470	InputArray rvec2, InputArray tvec2,
joeverbout	0:ea44dc9ed014	471	OutputArray rvec3, OutputArray tvec3,
joeverbout	0:ea44dc9ed014	472	OutputArray dr3dr1 = noArray(), OutputArray dr3dt1 = noArray(),
joeverbout	0:ea44dc9ed014	473	OutputArray dr3dr2 = noArray(), OutputArray dr3dt2 = noArray(),
joeverbout	0:ea44dc9ed014	474	OutputArray dt3dr1 = noArray(), OutputArray dt3dt1 = noArray(),
joeverbout	0:ea44dc9ed014	475	OutputArray dt3dr2 = noArray(), OutputArray dt3dt2 = noArray() );
joeverbout	0:ea44dc9ed014	476
joeverbout	0:ea44dc9ed014	477	/** @brief Projects 3D points to an image plane.
joeverbout	0:ea44dc9ed014	478
joeverbout	0:ea44dc9ed014	479	@param objectPoints Array of object points, 3xN/Nx3 1-channel or 1xN/Nx1 3-channel (or
joeverbout	0:ea44dc9ed014	480	vector\<Point3f\> ), where N is the number of points in the view.
joeverbout	0:ea44dc9ed014	481	@param rvec Rotation vector. See Rodrigues for details.
joeverbout	0:ea44dc9ed014	482	@param tvec Translation vector.
joeverbout	0:ea44dc9ed014	483	@param cameraMatrix Camera matrix \f$A = \vecthreethree{f_x}{0}{c_x}{0}{f_y}{c_y}{0}{0}{_1}\f$ .
joeverbout	0:ea44dc9ed014	484	@param distCoeffs Input vector of distortion coefficients
joeverbout	0:ea44dc9ed014	485	\f$(k_1, k_2, p_1, p_2[, k_3[, k_4, k_5, k_6 [, s_1, s_2, s_3, s_4[, \tau_x, \tau_y]]]])\f$ of
joeverbout	0:ea44dc9ed014	486	4, 5, 8, 12 or 14 elements. If the vector is empty, the zero distortion coefficients are assumed.
joeverbout	0:ea44dc9ed014	487	@param imagePoints Output array of image points, 2xN/Nx2 1-channel or 1xN/Nx1 2-channel, or
joeverbout	0:ea44dc9ed014	488	vector\<Point2f\> .
joeverbout	0:ea44dc9ed014	489	@param jacobian Optional output 2Nx(10+\<numDistCoeffs\>) jacobian matrix of derivatives of image
joeverbout	0:ea44dc9ed014	490	points with respect to components of the rotation vector, translation vector, focal lengths,
joeverbout	0:ea44dc9ed014	491	coordinates of the principal point and the distortion coefficients. In the old interface different
joeverbout	0:ea44dc9ed014	492	components of the jacobian are returned via different output parameters.
joeverbout	0:ea44dc9ed014	493	@param aspectRatio Optional "fixed aspect ratio" parameter. If the parameter is not 0, the
joeverbout	0:ea44dc9ed014	494	function assumes that the aspect ratio (fx/fy) is fixed and correspondingly adjusts the jacobian
joeverbout	0:ea44dc9ed014	495	matrix.
joeverbout	0:ea44dc9ed014	496
joeverbout	0:ea44dc9ed014	497	The function computes projections of 3D points to the image plane given intrinsic and extrinsic
joeverbout	0:ea44dc9ed014	498	camera parameters. Optionally, the function computes Jacobians - matrices of partial derivatives of
joeverbout	0:ea44dc9ed014	499	image points coordinates (as functions of all the input parameters) with respect to the particular
joeverbout	0:ea44dc9ed014	500	parameters, intrinsic and/or extrinsic. The Jacobians are used during the global optimization in
joeverbout	0:ea44dc9ed014	501	calibrateCamera, solvePnP, and stereoCalibrate . The function itself can also be used to compute a
joeverbout	0:ea44dc9ed014	502	re-projection error given the current intrinsic and extrinsic parameters.
joeverbout	0:ea44dc9ed014	503
joeverbout	0:ea44dc9ed014	504	@note By setting rvec=tvec=(0,0,0) or by setting cameraMatrix to a 3x3 identity matrix, or by
joeverbout	0:ea44dc9ed014	505	passing zero distortion coefficients, you can get various useful partial cases of the function. This
joeverbout	0:ea44dc9ed014	506	means that you can compute the distorted coordinates for a sparse set of points or apply a
joeverbout	0:ea44dc9ed014	507	perspective transformation (and also compute the derivatives) in the ideal zero-distortion setup.
joeverbout	0:ea44dc9ed014	508	*/
joeverbout	0:ea44dc9ed014	509	CV_EXPORTS_W void projectPoints( InputArray objectPoints,
joeverbout	0:ea44dc9ed014	510	InputArray rvec, InputArray tvec,
joeverbout	0:ea44dc9ed014	511	InputArray cameraMatrix, InputArray distCoeffs,
joeverbout	0:ea44dc9ed014	512	OutputArray imagePoints,
joeverbout	0:ea44dc9ed014	513	OutputArray jacobian = noArray(),
joeverbout	0:ea44dc9ed014	514	double aspectRatio = 0 );
joeverbout	0:ea44dc9ed014	515
joeverbout	0:ea44dc9ed014	516	/** @brief Finds an object pose from 3D-2D point correspondences.
joeverbout	0:ea44dc9ed014	517
joeverbout	0:ea44dc9ed014	518	@param objectPoints Array of object points in the object coordinate space, 3xN/Nx3 1-channel or
joeverbout	0:ea44dc9ed014	519	1xN/Nx1 3-channel, where N is the number of points. vector\<Point3f\> can be also passed here.
joeverbout	0:ea44dc9ed014	520	@param imagePoints Array of corresponding image points, 2xN/Nx2 1-channel or 1xN/Nx1 2-channel,
joeverbout	0:ea44dc9ed014	521	where N is the number of points. vector\<Point2f\> can be also passed here.
joeverbout	0:ea44dc9ed014	522	@param cameraMatrix Input camera matrix \f$A = \vecthreethree{fx}{0}{cx}{0}{fy}{cy}{0}{0}{1}\f$ .
joeverbout	0:ea44dc9ed014	523	@param distCoeffs Input vector of distortion coefficients
joeverbout	0:ea44dc9ed014	524	\f$(k_1, k_2, p_1, p_2[, k_3[, k_4, k_5, k_6 [, s_1, s_2, s_3, s_4[, \tau_x, \tau_y]]]])\f$ of
joeverbout	0:ea44dc9ed014	525	4, 5, 8, 12 or 14 elements. If the vector is NULL/empty, the zero distortion coefficients are
joeverbout	0:ea44dc9ed014	526	assumed.
joeverbout	0:ea44dc9ed014	527	@param rvec Output rotation vector (see Rodrigues ) that, together with tvec , brings points from
joeverbout	0:ea44dc9ed014	528	the model coordinate system to the camera coordinate system.
joeverbout	0:ea44dc9ed014	529	@param tvec Output translation vector.
joeverbout	0:ea44dc9ed014	530	@param useExtrinsicGuess Parameter used for SOLVEPNP_ITERATIVE. If true (1), the function uses
joeverbout	0:ea44dc9ed014	531	the provided rvec and tvec values as initial approximations of the rotation and translation
joeverbout	0:ea44dc9ed014	532	vectors, respectively, and further optimizes them.
joeverbout	0:ea44dc9ed014	533	@param flags Method for solving a PnP problem:
joeverbout	0:ea44dc9ed014	534	- SOLVEPNP_ITERATIVE Iterative method is based on Levenberg-Marquardt optimization. In
joeverbout	0:ea44dc9ed014	535	this case the function finds such a pose that minimizes reprojection error, that is the sum
joeverbout	0:ea44dc9ed014	536	of squared distances between the observed projections imagePoints and the projected (using
joeverbout	0:ea44dc9ed014	537	projectPoints ) objectPoints .
joeverbout	0:ea44dc9ed014	538	- SOLVEPNP_P3P Method is based on the paper of X.S. Gao, X.-R. Hou, J. Tang, H.-F. Chang
joeverbout	0:ea44dc9ed014	539	"Complete Solution Classification for the Perspective-Three-Point Problem". In this case the
joeverbout	0:ea44dc9ed014	540	function requires exactly four object and image points.
joeverbout	0:ea44dc9ed014	541	- SOLVEPNP_EPNP Method has been introduced by F.Moreno-Noguer, V.Lepetit and P.Fua in the
joeverbout	0:ea44dc9ed014	542	paper "EPnP: Efficient Perspective-n-Point Camera Pose Estimation".
joeverbout	0:ea44dc9ed014	543	- SOLVEPNP_DLS Method is based on the paper of Joel A. Hesch and Stergios I. Roumeliotis.
joeverbout	0:ea44dc9ed014	544	"A Direct Least-Squares (DLS) Method for PnP".
joeverbout	0:ea44dc9ed014	545	- SOLVEPNP_UPNP Method is based on the paper of A.Penate-Sanchez, J.Andrade-Cetto,
joeverbout	0:ea44dc9ed014	546	F.Moreno-Noguer. "Exhaustive Linearization for Robust Camera Pose and Focal Length
joeverbout	0:ea44dc9ed014	547	Estimation". In this case the function also estimates the parameters \f$f_x\f$ and \f$f_y\f$
joeverbout	0:ea44dc9ed014	548	assuming that both have the same value. Then the cameraMatrix is updated with the estimated
joeverbout	0:ea44dc9ed014	549	focal length.
joeverbout	0:ea44dc9ed014	550
joeverbout	0:ea44dc9ed014	551	The function estimates the object pose given a set of object points, their corresponding image
joeverbout	0:ea44dc9ed014	552	projections, as well as the camera matrix and the distortion coefficients.
joeverbout	0:ea44dc9ed014	553
joeverbout	0:ea44dc9ed014	554	@note
joeverbout	0:ea44dc9ed014	555	- An example of how to use solvePnP for planar augmented reality can be found at
joeverbout	0:ea44dc9ed014	556	opencv_source_code/samples/python/plane_ar.py
joeverbout	0:ea44dc9ed014	557	- If you are using Python:
joeverbout	0:ea44dc9ed014	558	- Numpy array slices won't work as input because solvePnP requires contiguous
joeverbout	0:ea44dc9ed014	559	arrays (enforced by the assertion using cv::Mat::checkVector() around line 55 of
joeverbout	0:ea44dc9ed014	560	modules/calib3d/src/solvepnp.cpp version 2.4.9)
joeverbout	0:ea44dc9ed014	561	- The P3P algorithm requires image points to be in an array of shape (N,1,2) due
joeverbout	0:ea44dc9ed014	562	to its calling of cv::undistortPoints (around line 75 of modules/calib3d/src/solvepnp.cpp version 2.4.9)
joeverbout	0:ea44dc9ed014	563	which requires 2-channel information.
joeverbout	0:ea44dc9ed014	564	- Thus, given some data D = np.array(...) where D.shape = (N,M), in order to use a subset of
joeverbout	0:ea44dc9ed014	565	it as, e.g., imagePoints, one must effectively copy it into a new array: imagePoints =
joeverbout	0:ea44dc9ed014	566	np.ascontiguousarray(D[:,:2]).reshape((N,1,2))
joeverbout	0:ea44dc9ed014	567	*/
joeverbout	0:ea44dc9ed014	568	CV_EXPORTS_W bool solvePnP( InputArray objectPoints, InputArray imagePoints,
joeverbout	0:ea44dc9ed014	569	InputArray cameraMatrix, InputArray distCoeffs,
joeverbout	0:ea44dc9ed014	570	OutputArray rvec, OutputArray tvec,
joeverbout	0:ea44dc9ed014	571	bool useExtrinsicGuess = false, int flags = SOLVEPNP_ITERATIVE );
joeverbout	0:ea44dc9ed014	572
joeverbout	0:ea44dc9ed014	573	/** @brief Finds an object pose from 3D-2D point correspondences using the RANSAC scheme.
joeverbout	0:ea44dc9ed014	574
joeverbout	0:ea44dc9ed014	575	@param objectPoints Array of object points in the object coordinate space, 3xN/Nx3 1-channel or
joeverbout	0:ea44dc9ed014	576	1xN/Nx1 3-channel, where N is the number of points. vector\<Point3f\> can be also passed here.
joeverbout	0:ea44dc9ed014	577	@param imagePoints Array of corresponding image points, 2xN/Nx2 1-channel or 1xN/Nx1 2-channel,
joeverbout	0:ea44dc9ed014	578	where N is the number of points. vector\<Point2f\> can be also passed here.
joeverbout	0:ea44dc9ed014	579	@param cameraMatrix Input camera matrix \f$A = \vecthreethree{fx}{0}{cx}{0}{fy}{cy}{0}{0}{1}\f$ .
joeverbout	0:ea44dc9ed014	580	@param distCoeffs Input vector of distortion coefficients
joeverbout	0:ea44dc9ed014	581	\f$(k_1, k_2, p_1, p_2[, k_3[, k_4, k_5, k_6 [, s_1, s_2, s_3, s_4[, \tau_x, \tau_y]]]])\f$ of
joeverbout	0:ea44dc9ed014	582	4, 5, 8, 12 or 14 elements. If the vector is NULL/empty, the zero distortion coefficients are
joeverbout	0:ea44dc9ed014	583	assumed.
joeverbout	0:ea44dc9ed014	584	@param rvec Output rotation vector (see Rodrigues ) that, together with tvec , brings points from
joeverbout	0:ea44dc9ed014	585	the model coordinate system to the camera coordinate system.
joeverbout	0:ea44dc9ed014	586	@param tvec Output translation vector.
joeverbout	0:ea44dc9ed014	587	@param useExtrinsicGuess Parameter used for SOLVEPNP_ITERATIVE. If true (1), the function uses
joeverbout	0:ea44dc9ed014	588	the provided rvec and tvec values as initial approximations of the rotation and translation
joeverbout	0:ea44dc9ed014	589	vectors, respectively, and further optimizes them.
joeverbout	0:ea44dc9ed014	590	@param iterationsCount Number of iterations.
joeverbout	0:ea44dc9ed014	591	@param reprojectionError Inlier threshold value used by the RANSAC procedure. The parameter value
joeverbout	0:ea44dc9ed014	592	is the maximum allowed distance between the observed and computed point projections to consider it
joeverbout	0:ea44dc9ed014	593	an inlier.
joeverbout	0:ea44dc9ed014	594	@param confidence The probability that the algorithm produces a useful result.
joeverbout	0:ea44dc9ed014	595	@param inliers Output vector that contains indices of inliers in objectPoints and imagePoints .
joeverbout	0:ea44dc9ed014	596	@param flags Method for solving a PnP problem (see solvePnP ).
joeverbout	0:ea44dc9ed014	597
joeverbout	0:ea44dc9ed014	598	The function estimates an object pose given a set of object points, their corresponding image
joeverbout	0:ea44dc9ed014	599	projections, as well as the camera matrix and the distortion coefficients. This function finds such
joeverbout	0:ea44dc9ed014	600	a pose that minimizes reprojection error, that is, the sum of squared distances between the observed
joeverbout	0:ea44dc9ed014	601	projections imagePoints and the projected (using projectPoints ) objectPoints. The use of RANSAC
joeverbout	0:ea44dc9ed014	602	makes the function resistant to outliers.
joeverbout	0:ea44dc9ed014	603
joeverbout	0:ea44dc9ed014	604	@note
joeverbout	0:ea44dc9ed014	605	- An example of how to use solvePNPRansac for object detection can be found at
joeverbout	0:ea44dc9ed014	606	opencv_source_code/samples/cpp/tutorial_code/calib3d/real_time_pose_estimation/
joeverbout	0:ea44dc9ed014	607	*/
joeverbout	0:ea44dc9ed014	608	CV_EXPORTS_W bool solvePnPRansac( InputArray objectPoints, InputArray imagePoints,
joeverbout	0:ea44dc9ed014	609	InputArray cameraMatrix, InputArray distCoeffs,
joeverbout	0:ea44dc9ed014	610	OutputArray rvec, OutputArray tvec,
joeverbout	0:ea44dc9ed014	611	bool useExtrinsicGuess = false, int iterationsCount = 100,
joeverbout	0:ea44dc9ed014	612	float reprojectionError = 8.0, double confidence = 0.99,
joeverbout	0:ea44dc9ed014	613	OutputArray inliers = noArray(), int flags = SOLVEPNP_ITERATIVE );
joeverbout	0:ea44dc9ed014	614
joeverbout	0:ea44dc9ed014	615	/** @brief Finds an initial camera matrix from 3D-2D point correspondences.
joeverbout	0:ea44dc9ed014	616
joeverbout	0:ea44dc9ed014	617	@param objectPoints Vector of vectors of the calibration pattern points in the calibration pattern
joeverbout	0:ea44dc9ed014	618	coordinate space. In the old interface all the per-view vectors are concatenated. See
joeverbout	0:ea44dc9ed014	619	calibrateCamera for details.
joeverbout	0:ea44dc9ed014	620	@param imagePoints Vector of vectors of the projections of the calibration pattern points. In the
joeverbout	0:ea44dc9ed014	621	old interface all the per-view vectors are concatenated.
joeverbout	0:ea44dc9ed014	622	@param imageSize Image size in pixels used to initialize the principal point.
joeverbout	0:ea44dc9ed014	623	@param aspectRatio If it is zero or negative, both \f$f_x\f$ and \f$f_y\f$ are estimated independently.
joeverbout	0:ea44dc9ed014	624	Otherwise, \f$f_x = f_y * \texttt{aspectRatio}\f$ .
joeverbout	0:ea44dc9ed014	625
joeverbout	0:ea44dc9ed014	626	The function estimates and returns an initial camera matrix for the camera calibration process.
joeverbout	0:ea44dc9ed014	627	Currently, the function only supports planar calibration patterns, which are patterns where each
joeverbout	0:ea44dc9ed014	628	object point has z-coordinate =0.
joeverbout	0:ea44dc9ed014	629	*/
joeverbout	0:ea44dc9ed014	630	CV_EXPORTS_W Mat initCameraMatrix2D( InputArrayOfArrays objectPoints,
joeverbout	0:ea44dc9ed014	631	InputArrayOfArrays imagePoints,
joeverbout	0:ea44dc9ed014	632	Size imageSize, double aspectRatio = 1.0 );
joeverbout	0:ea44dc9ed014	633
joeverbout	0:ea44dc9ed014	634	/** @brief Finds the positions of internal corners of the chessboard.
joeverbout	0:ea44dc9ed014	635
joeverbout	0:ea44dc9ed014	636	@param image Source chessboard view. It must be an 8-bit grayscale or color image.
joeverbout	0:ea44dc9ed014	637	@param patternSize Number of inner corners per a chessboard row and column
joeverbout	0:ea44dc9ed014	638	( patternSize = cvSize(points_per_row,points_per_colum) = cvSize(columns,rows) ).
joeverbout	0:ea44dc9ed014	639	@param corners Output array of detected corners.
joeverbout	0:ea44dc9ed014	640	@param flags Various operation flags that can be zero or a combination of the following values:
joeverbout	0:ea44dc9ed014	641	- CV_CALIB_CB_ADAPTIVE_THRESH Use adaptive thresholding to convert the image to black
joeverbout	0:ea44dc9ed014	642	and white, rather than a fixed threshold level (computed from the average image brightness).
joeverbout	0:ea44dc9ed014	643	- CV_CALIB_CB_NORMALIZE_IMAGE Normalize the image gamma with equalizeHist before
joeverbout	0:ea44dc9ed014	644	applying fixed or adaptive thresholding.
joeverbout	0:ea44dc9ed014	645	- CV_CALIB_CB_FILTER_QUADS Use additional criteria (like contour area, perimeter,
joeverbout	0:ea44dc9ed014	646	square-like shape) to filter out false quads extracted at the contour retrieval stage.
joeverbout	0:ea44dc9ed014	647	- CALIB_CB_FAST_CHECK Run a fast check on the image that looks for chessboard corners,
joeverbout	0:ea44dc9ed014	648	and shortcut the call if none is found. This can drastically speed up the call in the
joeverbout	0:ea44dc9ed014	649	degenerate condition when no chessboard is observed.
joeverbout	0:ea44dc9ed014	650
joeverbout	0:ea44dc9ed014	651	The function attempts to determine whether the input image is a view of the chessboard pattern and
joeverbout	0:ea44dc9ed014	652	locate the internal chessboard corners. The function returns a non-zero value if all of the corners
joeverbout	0:ea44dc9ed014	653	are found and they are placed in a certain order (row by row, left to right in every row).
joeverbout	0:ea44dc9ed014	654	Otherwise, if the function fails to find all the corners or reorder them, it returns 0. For example,
joeverbout	0:ea44dc9ed014	655	a regular chessboard has 8 x 8 squares and 7 x 7 internal corners, that is, points where the black
joeverbout	0:ea44dc9ed014	656	squares touch each other. The detected coordinates are approximate, and to determine their positions
joeverbout	0:ea44dc9ed014	657	more accurately, the function calls cornerSubPix. You also may use the function cornerSubPix with
joeverbout	0:ea44dc9ed014	658	different parameters if returned coordinates are not accurate enough.
joeverbout	0:ea44dc9ed014	659
joeverbout	0:ea44dc9ed014	660	Sample usage of detecting and drawing chessboard corners: :
joeverbout	0:ea44dc9ed014	661	@code
joeverbout	0:ea44dc9ed014	662	Size patternsize(8,6); //interior number of corners
joeverbout	0:ea44dc9ed014	663	Mat gray = ....; //source image
joeverbout	0:ea44dc9ed014	664	vector<Point2f> corners; //this will be filled by the detected corners
joeverbout	0:ea44dc9ed014	665
joeverbout	0:ea44dc9ed014	666	//CALIB_CB_FAST_CHECK saves a lot of time on images
joeverbout	0:ea44dc9ed014	667	//that do not contain any chessboard corners
joeverbout	0:ea44dc9ed014	668	bool patternfound = findChessboardCorners(gray, patternsize, corners,
joeverbout	0:ea44dc9ed014	669	CALIB_CB_ADAPTIVE_THRESH + CALIB_CB_NORMALIZE_IMAGE
joeverbout	0:ea44dc9ed014	670	+ CALIB_CB_FAST_CHECK);
joeverbout	0:ea44dc9ed014	671
joeverbout	0:ea44dc9ed014	672	if(patternfound)
joeverbout	0:ea44dc9ed014	673	cornerSubPix(gray, corners, Size(11, 11), Size(-1, -1),
joeverbout	0:ea44dc9ed014	674	TermCriteria(CV_TERMCRIT_EPS + CV_TERMCRIT_ITER, 30, 0.1));
joeverbout	0:ea44dc9ed014	675
joeverbout	0:ea44dc9ed014	676	drawChessboardCorners(img, patternsize, Mat(corners), patternfound);
joeverbout	0:ea44dc9ed014	677	@endcode
joeverbout	0:ea44dc9ed014	678	@note The function requires white space (like a square-thick border, the wider the better) around
joeverbout	0:ea44dc9ed014	679	the board to make the detection more robust in various environments. Otherwise, if there is no
joeverbout	0:ea44dc9ed014	680	border and the background is dark, the outer black squares cannot be segmented properly and so the
joeverbout	0:ea44dc9ed014	681	square grouping and ordering algorithm fails.
joeverbout	0:ea44dc9ed014	682	*/
joeverbout	0:ea44dc9ed014	683	CV_EXPORTS_W bool findChessboardCorners( InputArray image, Size patternSize, OutputArray corners,
joeverbout	0:ea44dc9ed014	684	int flags = CALIB_CB_ADAPTIVE_THRESH + CALIB_CB_NORMALIZE_IMAGE );
joeverbout	0:ea44dc9ed014	685
joeverbout	0:ea44dc9ed014	686	//! finds subpixel-accurate positions of the chessboard corners
joeverbout	0:ea44dc9ed014	687	CV_EXPORTS bool find4QuadCornerSubpix( InputArray img, InputOutputArray corners, Size region_size );
joeverbout	0:ea44dc9ed014	688
joeverbout	0:ea44dc9ed014	689	/** @brief Renders the detected chessboard corners.
joeverbout	0:ea44dc9ed014	690
joeverbout	0:ea44dc9ed014	691	@param image Destination image. It must be an 8-bit color image.
joeverbout	0:ea44dc9ed014	692	@param patternSize Number of inner corners per a chessboard row and column
joeverbout	0:ea44dc9ed014	693	(patternSize = cv::Size(points_per_row,points_per_column)).
joeverbout	0:ea44dc9ed014	694	@param corners Array of detected corners, the output of findChessboardCorners.
joeverbout	0:ea44dc9ed014	695	@param patternWasFound Parameter indicating whether the complete board was found or not. The
joeverbout	0:ea44dc9ed014	696	return value of findChessboardCorners should be passed here.
joeverbout	0:ea44dc9ed014	697
joeverbout	0:ea44dc9ed014	698	The function draws individual chessboard corners detected either as red circles if the board was not
joeverbout	0:ea44dc9ed014	699	found, or as colored corners connected with lines if the board was found.
joeverbout	0:ea44dc9ed014	700	*/
joeverbout	0:ea44dc9ed014	701	CV_EXPORTS_W void drawChessboardCorners( InputOutputArray image, Size patternSize,
joeverbout	0:ea44dc9ed014	702	InputArray corners, bool patternWasFound );
joeverbout	0:ea44dc9ed014	703
joeverbout	0:ea44dc9ed014	704	/** @brief Finds centers in the grid of circles.
joeverbout	0:ea44dc9ed014	705
joeverbout	0:ea44dc9ed014	706	@param image grid view of input circles; it must be an 8-bit grayscale or color image.
joeverbout	0:ea44dc9ed014	707	@param patternSize number of circles per row and column
joeverbout	0:ea44dc9ed014	708	( patternSize = Size(points_per_row, points_per_colum) ).
joeverbout	0:ea44dc9ed014	709	@param centers output array of detected centers.
joeverbout	0:ea44dc9ed014	710	@param flags various operation flags that can be one of the following values:
joeverbout	0:ea44dc9ed014	711	- CALIB_CB_SYMMETRIC_GRID uses symmetric pattern of circles.
joeverbout	0:ea44dc9ed014	712	- CALIB_CB_ASYMMETRIC_GRID uses asymmetric pattern of circles.
joeverbout	0:ea44dc9ed014	713	- CALIB_CB_CLUSTERING uses a special algorithm for grid detection. It is more robust to
joeverbout	0:ea44dc9ed014	714	perspective distortions but much more sensitive to background clutter.
joeverbout	0:ea44dc9ed014	715	@param blobDetector feature detector that finds blobs like dark circles on light background.
joeverbout	0:ea44dc9ed014	716
joeverbout	0:ea44dc9ed014	717	The function attempts to determine whether the input image contains a grid of circles. If it is, the
joeverbout	0:ea44dc9ed014	718	function locates centers of the circles. The function returns a non-zero value if all of the centers
joeverbout	0:ea44dc9ed014	719	have been found and they have been placed in a certain order (row by row, left to right in every
joeverbout	0:ea44dc9ed014	720	row). Otherwise, if the function fails to find all the corners or reorder them, it returns 0.
joeverbout	0:ea44dc9ed014	721
joeverbout	0:ea44dc9ed014	722	Sample usage of detecting and drawing the centers of circles: :
joeverbout	0:ea44dc9ed014	723	@code
joeverbout	0:ea44dc9ed014	724	Size patternsize(7,7); //number of centers
joeverbout	0:ea44dc9ed014	725	Mat gray = ....; //source image
joeverbout	0:ea44dc9ed014	726	vector<Point2f> centers; //this will be filled by the detected centers
joeverbout	0:ea44dc9ed014	727
joeverbout	0:ea44dc9ed014	728	bool patternfound = findCirclesGrid(gray, patternsize, centers);
joeverbout	0:ea44dc9ed014	729
joeverbout	0:ea44dc9ed014	730	drawChessboardCorners(img, patternsize, Mat(centers), patternfound);
joeverbout	0:ea44dc9ed014	731	@endcode
joeverbout	0:ea44dc9ed014	732	@note The function requires white space (like a square-thick border, the wider the better) around
joeverbout	0:ea44dc9ed014	733	the board to make the detection more robust in various environments.
joeverbout	0:ea44dc9ed014	734	*/
joeverbout	0:ea44dc9ed014	735	CV_EXPORTS_W bool findCirclesGrid( InputArray image, Size patternSize,
joeverbout	0:ea44dc9ed014	736	OutputArray centers, int flags = CALIB_CB_SYMMETRIC_GRID,
joeverbout	0:ea44dc9ed014	737	const Ptr<FeatureDetector> &blobDetector = SimpleBlobDetector::create());
joeverbout	0:ea44dc9ed014	738
joeverbout	0:ea44dc9ed014	739	/** @brief Finds the camera intrinsic and extrinsic parameters from several views of a calibration pattern.
joeverbout	0:ea44dc9ed014	740
joeverbout	0:ea44dc9ed014	741	@param objectPoints In the new interface it is a vector of vectors of calibration pattern points in
joeverbout	0:ea44dc9ed014	742	the calibration pattern coordinate space (e.g. std::vector<std::vector<cv::Vec3f>>). The outer
joeverbout	0:ea44dc9ed014	743	vector contains as many elements as the number of the pattern views. If the same calibration pattern
joeverbout	0:ea44dc9ed014	744	is shown in each view and it is fully visible, all the vectors will be the same. Although, it is
joeverbout	0:ea44dc9ed014	745	possible to use partially occluded patterns, or even different patterns in different views. Then,
joeverbout	0:ea44dc9ed014	746	the vectors will be different. The points are 3D, but since they are in a pattern coordinate system,
joeverbout	0:ea44dc9ed014	747	then, if the rig is planar, it may make sense to put the model to a XY coordinate plane so that
joeverbout	0:ea44dc9ed014	748	Z-coordinate of each input object point is 0.
joeverbout	0:ea44dc9ed014	749	In the old interface all the vectors of object points from different views are concatenated
joeverbout	0:ea44dc9ed014	750	together.
joeverbout	0:ea44dc9ed014	751	@param imagePoints In the new interface it is a vector of vectors of the projections of calibration
joeverbout	0:ea44dc9ed014	752	pattern points (e.g. std::vector<std::vector<cv::Vec2f>>). imagePoints.size() and
joeverbout	0:ea44dc9ed014	753	objectPoints.size() and imagePoints[i].size() must be equal to objectPoints[i].size() for each i.
joeverbout	0:ea44dc9ed014	754	In the old interface all the vectors of object points from different views are concatenated
joeverbout	0:ea44dc9ed014	755	together.
joeverbout	0:ea44dc9ed014	756	@param imageSize Size of the image used only to initialize the intrinsic camera matrix.
joeverbout	0:ea44dc9ed014	757	@param cameraMatrix Output 3x3 floating-point camera matrix
joeverbout	0:ea44dc9ed014	758	\f$A = \vecthreethree{f_x}{0}{c_x}{0}{f_y}{c_y}{0}{0}{1}\f$ . If CV\_CALIB\_USE\_INTRINSIC\_GUESS
joeverbout	0:ea44dc9ed014	759	and/or CV_CALIB_FIX_ASPECT_RATIO are specified, some or all of fx, fy, cx, cy must be
joeverbout	0:ea44dc9ed014	760	initialized before calling the function.
joeverbout	0:ea44dc9ed014	761	@param distCoeffs Output vector of distortion coefficients
joeverbout	0:ea44dc9ed014	762	\f$(k_1, k_2, p_1, p_2[, k_3[, k_4, k_5, k_6 [, s_1, s_2, s_3, s_4[, \tau_x, \tau_y]]]])\f$ of
joeverbout	0:ea44dc9ed014	763	4, 5, 8, 12 or 14 elements.
joeverbout	0:ea44dc9ed014	764	@param rvecs Output vector of rotation vectors (see Rodrigues ) estimated for each pattern view
joeverbout	0:ea44dc9ed014	765	(e.g. std::vector<cv::Mat>>). That is, each k-th rotation vector together with the corresponding
joeverbout	0:ea44dc9ed014	766	k-th translation vector (see the next output parameter description) brings the calibration pattern
joeverbout	0:ea44dc9ed014	767	from the model coordinate space (in which object points are specified) to the world coordinate
joeverbout	0:ea44dc9ed014	768	space, that is, a real position of the calibration pattern in the k-th pattern view (k=0.. M -1).
joeverbout	0:ea44dc9ed014	769	@param tvecs Output vector of translation vectors estimated for each pattern view.
joeverbout	0:ea44dc9ed014	770	@param flags Different flags that may be zero or a combination of the following values:
joeverbout	0:ea44dc9ed014	771	- CV_CALIB_USE_INTRINSIC_GUESS cameraMatrix contains valid initial values of
joeverbout	0:ea44dc9ed014	772	fx, fy, cx, cy that are optimized further. Otherwise, (cx, cy) is initially set to the image
joeverbout	0:ea44dc9ed014	773	center ( imageSize is used), and focal distances are computed in a least-squares fashion.
joeverbout	0:ea44dc9ed014	774	Note, that if intrinsic parameters are known, there is no need to use this function just to
joeverbout	0:ea44dc9ed014	775	estimate extrinsic parameters. Use solvePnP instead.
joeverbout	0:ea44dc9ed014	776	- CV_CALIB_FIX_PRINCIPAL_POINT The principal point is not changed during the global
joeverbout	0:ea44dc9ed014	777	optimization. It stays at the center or at a different location specified when
joeverbout	0:ea44dc9ed014	778	CV_CALIB_USE_INTRINSIC_GUESS is set too.
joeverbout	0:ea44dc9ed014	779	- CV_CALIB_FIX_ASPECT_RATIO The functions considers only fy as a free parameter. The
joeverbout	0:ea44dc9ed014	780	ratio fx/fy stays the same as in the input cameraMatrix . When
joeverbout	0:ea44dc9ed014	781	CV_CALIB_USE_INTRINSIC_GUESS is not set, the actual input values of fx and fy are
joeverbout	0:ea44dc9ed014	782	ignored, only their ratio is computed and used further.
joeverbout	0:ea44dc9ed014	783	- CV_CALIB_ZERO_TANGENT_DIST Tangential distortion coefficients \f$(p_1, p_2)\f$ are set
joeverbout	0:ea44dc9ed014	784	to zeros and stay zero.
joeverbout	0:ea44dc9ed014	785	- CV_CALIB_FIX_K1,...,CV_CALIB_FIX_K6 The corresponding radial distortion
joeverbout	0:ea44dc9ed014	786	coefficient is not changed during the optimization. If CV_CALIB_USE_INTRINSIC_GUESS is
joeverbout	0:ea44dc9ed014	787	set, the coefficient from the supplied distCoeffs matrix is used. Otherwise, it is set to 0.
joeverbout	0:ea44dc9ed014	788	- CV_CALIB_RATIONAL_MODEL Coefficients k4, k5, and k6 are enabled. To provide the
joeverbout	0:ea44dc9ed014	789	backward compatibility, this extra flag should be explicitly specified to make the
joeverbout	0:ea44dc9ed014	790	calibration function use the rational model and return 8 coefficients. If the flag is not
joeverbout	0:ea44dc9ed014	791	set, the function computes and returns only 5 distortion coefficients.
joeverbout	0:ea44dc9ed014	792	- CALIB_THIN_PRISM_MODEL Coefficients s1, s2, s3 and s4 are enabled. To provide the
joeverbout	0:ea44dc9ed014	793	backward compatibility, this extra flag should be explicitly specified to make the
joeverbout	0:ea44dc9ed014	794	calibration function use the thin prism model and return 12 coefficients. If the flag is not
joeverbout	0:ea44dc9ed014	795	set, the function computes and returns only 5 distortion coefficients.
joeverbout	0:ea44dc9ed014	796	- CALIB_FIX_S1_S2_S3_S4 The thin prism distortion coefficients are not changed during
joeverbout	0:ea44dc9ed014	797	the optimization. If CV_CALIB_USE_INTRINSIC_GUESS is set, the coefficient from the
joeverbout	0:ea44dc9ed014	798	supplied distCoeffs matrix is used. Otherwise, it is set to 0.
joeverbout	0:ea44dc9ed014	799	- CALIB_TILTED_MODEL Coefficients tauX and tauY are enabled. To provide the
joeverbout	0:ea44dc9ed014	800	backward compatibility, this extra flag should be explicitly specified to make the
joeverbout	0:ea44dc9ed014	801	calibration function use the tilted sensor model and return 14 coefficients. If the flag is not
joeverbout	0:ea44dc9ed014	802	set, the function computes and returns only 5 distortion coefficients.
joeverbout	0:ea44dc9ed014	803	- CALIB_FIX_TAUX_TAUY The coefficients of the tilted sensor model are not changed during
joeverbout	0:ea44dc9ed014	804	the optimization. If CV_CALIB_USE_INTRINSIC_GUESS is set, the coefficient from the
joeverbout	0:ea44dc9ed014	805	supplied distCoeffs matrix is used. Otherwise, it is set to 0.
joeverbout	0:ea44dc9ed014	806	@param criteria Termination criteria for the iterative optimization algorithm.
joeverbout	0:ea44dc9ed014	807
joeverbout	0:ea44dc9ed014	808	The function estimates the intrinsic camera parameters and extrinsic parameters for each of the
joeverbout	0:ea44dc9ed014	809	views. The algorithm is based on @cite Zhang2000 and @cite BouguetMCT . The coordinates of 3D object
joeverbout	0:ea44dc9ed014	810	points and their corresponding 2D projections in each view must be specified. That may be achieved
joeverbout	0:ea44dc9ed014	811	by using an object with a known geometry and easily detectable feature points. Such an object is
joeverbout	0:ea44dc9ed014	812	called a calibration rig or calibration pattern, and OpenCV has built-in support for a chessboard as
joeverbout	0:ea44dc9ed014	813	a calibration rig (see findChessboardCorners ). Currently, initialization of intrinsic parameters
joeverbout	0:ea44dc9ed014	814	(when CV_CALIB_USE_INTRINSIC_GUESS is not set) is only implemented for planar calibration
joeverbout	0:ea44dc9ed014	815	patterns (where Z-coordinates of the object points must be all zeros). 3D calibration rigs can also
joeverbout	0:ea44dc9ed014	816	be used as long as initial cameraMatrix is provided.
joeverbout	0:ea44dc9ed014	817
joeverbout	0:ea44dc9ed014	818	The algorithm performs the following steps:
joeverbout	0:ea44dc9ed014	819
joeverbout	0:ea44dc9ed014	820	- Compute the initial intrinsic parameters (the option only available for planar calibration
joeverbout	0:ea44dc9ed014	821	patterns) or read them from the input parameters. The distortion coefficients are all set to
joeverbout	0:ea44dc9ed014	822	zeros initially unless some of CV_CALIB_FIX_K? are specified.
joeverbout	0:ea44dc9ed014	823
joeverbout	0:ea44dc9ed014	824	- Estimate the initial camera pose as if the intrinsic parameters have been already known. This is
joeverbout	0:ea44dc9ed014	825	done using solvePnP .
joeverbout	0:ea44dc9ed014	826
joeverbout	0:ea44dc9ed014	827	- Run the global Levenberg-Marquardt optimization algorithm to minimize the reprojection error,
joeverbout	0:ea44dc9ed014	828	that is, the total sum of squared distances between the observed feature points imagePoints and
joeverbout	0:ea44dc9ed014	829	the projected (using the current estimates for camera parameters and the poses) object points
joeverbout	0:ea44dc9ed014	830	objectPoints. See projectPoints for details.
joeverbout	0:ea44dc9ed014	831
joeverbout	0:ea44dc9ed014	832	The function returns the final re-projection error.
joeverbout	0:ea44dc9ed014	833
joeverbout	0:ea44dc9ed014	834	@note
joeverbout	0:ea44dc9ed014	835	If you use a non-square (=non-NxN) grid and findChessboardCorners for calibration, and
joeverbout	0:ea44dc9ed014	836	calibrateCamera returns bad values (zero distortion coefficients, an image center very far from
joeverbout	0:ea44dc9ed014	837	(w/2-0.5,h/2-0.5), and/or large differences between \f$f_x\f$ and \f$f_y\f$ (ratios of 10:1 or more)),
joeverbout	0:ea44dc9ed014	838	then you have probably used patternSize=cvSize(rows,cols) instead of using
joeverbout	0:ea44dc9ed014	839	patternSize=cvSize(cols,rows) in findChessboardCorners .
joeverbout	0:ea44dc9ed014	840
joeverbout	0:ea44dc9ed014	841	@sa
joeverbout	0:ea44dc9ed014	842	findChessboardCorners, solvePnP, initCameraMatrix2D, stereoCalibrate, undistort
joeverbout	0:ea44dc9ed014	843	*/
joeverbout	0:ea44dc9ed014	844	CV_EXPORTS_W double calibrateCamera( InputArrayOfArrays objectPoints,
joeverbout	0:ea44dc9ed014	845	InputArrayOfArrays imagePoints, Size imageSize,
joeverbout	0:ea44dc9ed014	846	InputOutputArray cameraMatrix, InputOutputArray distCoeffs,
joeverbout	0:ea44dc9ed014	847	OutputArrayOfArrays rvecs, OutputArrayOfArrays tvecs,
joeverbout	0:ea44dc9ed014	848	int flags = 0, TermCriteria criteria = TermCriteria(
joeverbout	0:ea44dc9ed014	849	TermCriteria::COUNT + TermCriteria::EPS, 30, DBL_EPSILON) );
joeverbout	0:ea44dc9ed014	850
joeverbout	0:ea44dc9ed014	851	/** @brief Computes useful camera characteristics from the camera matrix.
joeverbout	0:ea44dc9ed014	852
joeverbout	0:ea44dc9ed014	853	@param cameraMatrix Input camera matrix that can be estimated by calibrateCamera or
joeverbout	0:ea44dc9ed014	854	stereoCalibrate .
joeverbout	0:ea44dc9ed014	855	@param imageSize Input image size in pixels.
joeverbout	0:ea44dc9ed014	856	@param apertureWidth Physical width in mm of the sensor.
joeverbout	0:ea44dc9ed014	857	@param apertureHeight Physical height in mm of the sensor.
joeverbout	0:ea44dc9ed014	858	@param fovx Output field of view in degrees along the horizontal sensor axis.
joeverbout	0:ea44dc9ed014	859	@param fovy Output field of view in degrees along the vertical sensor axis.
joeverbout	0:ea44dc9ed014	860	@param focalLength Focal length of the lens in mm.
joeverbout	0:ea44dc9ed014	861	@param principalPoint Principal point in mm.
joeverbout	0:ea44dc9ed014	862	@param aspectRatio \f$f_y/f_x\f$
joeverbout	0:ea44dc9ed014	863
joeverbout	0:ea44dc9ed014	864	The function computes various useful camera characteristics from the previously estimated camera
joeverbout	0:ea44dc9ed014	865	matrix.
joeverbout	0:ea44dc9ed014	866
joeverbout	0:ea44dc9ed014	867	@note
joeverbout	0:ea44dc9ed014	868	Do keep in mind that the unity measure 'mm' stands for whatever unit of measure one chooses for
joeverbout	0:ea44dc9ed014	869	the chessboard pitch (it can thus be any value).
joeverbout	0:ea44dc9ed014	870	*/
joeverbout	0:ea44dc9ed014	871	CV_EXPORTS_W void calibrationMatrixValues( InputArray cameraMatrix, Size imageSize,
joeverbout	0:ea44dc9ed014	872	double apertureWidth, double apertureHeight,
joeverbout	0:ea44dc9ed014	873	CV_OUT double& fovx, CV_OUT double& fovy,
joeverbout	0:ea44dc9ed014	874	CV_OUT double& focalLength, CV_OUT Point2d& principalPoint,
joeverbout	0:ea44dc9ed014	875	CV_OUT double& aspectRatio );
joeverbout	0:ea44dc9ed014	876
joeverbout	0:ea44dc9ed014	877	/** @brief Calibrates the stereo camera.
joeverbout	0:ea44dc9ed014	878
joeverbout	0:ea44dc9ed014	879	@param objectPoints Vector of vectors of the calibration pattern points.
joeverbout	0:ea44dc9ed014	880	@param imagePoints1 Vector of vectors of the projections of the calibration pattern points,
joeverbout	0:ea44dc9ed014	881	observed by the first camera.
joeverbout	0:ea44dc9ed014	882	@param imagePoints2 Vector of vectors of the projections of the calibration pattern points,
joeverbout	0:ea44dc9ed014	883	observed by the second camera.
joeverbout	0:ea44dc9ed014	884	@param cameraMatrix1 Input/output first camera matrix:
joeverbout	0:ea44dc9ed014	885	\f$\vecthreethree{f_x^{(j)}}{0}{c_x^{(j)}}{0}{f_y^{(j)}}{c_y^{(j)}}{0}{0}{1}\f$ , \f$j = 0,\, 1\f$ . If
joeverbout	0:ea44dc9ed014	886	any of CV_CALIB_USE_INTRINSIC_GUESS , CV_CALIB_FIX_ASPECT_RATIO ,
joeverbout	0:ea44dc9ed014	887	CV_CALIB_FIX_INTRINSIC , or CV_CALIB_FIX_FOCAL_LENGTH are specified, some or all of the
joeverbout	0:ea44dc9ed014	888	matrix components must be initialized. See the flags description for details.
joeverbout	0:ea44dc9ed014	889	@param distCoeffs1 Input/output vector of distortion coefficients
joeverbout	0:ea44dc9ed014	890	\f$(k_1, k_2, p_1, p_2[, k_3[, k_4, k_5, k_6 [, s_1, s_2, s_3, s_4[, \tau_x, \tau_y]]]])\f$ of
joeverbout	0:ea44dc9ed014	891	4, 5, 8, 12 or 14 elements. The output vector length depends on the flags.
joeverbout	0:ea44dc9ed014	892	@param cameraMatrix2 Input/output second camera matrix. The parameter is similar to cameraMatrix1
joeverbout	0:ea44dc9ed014	893	@param distCoeffs2 Input/output lens distortion coefficients for the second camera. The parameter
joeverbout	0:ea44dc9ed014	894	is similar to distCoeffs1 .
joeverbout	0:ea44dc9ed014	895	@param imageSize Size of the image used only to initialize intrinsic camera matrix.
joeverbout	0:ea44dc9ed014	896	@param R Output rotation matrix between the 1st and the 2nd camera coordinate systems.
joeverbout	0:ea44dc9ed014	897	@param T Output translation vector between the coordinate systems of the cameras.
joeverbout	0:ea44dc9ed014	898	@param E Output essential matrix.
joeverbout	0:ea44dc9ed014	899	@param F Output fundamental matrix.
joeverbout	0:ea44dc9ed014	900	@param flags Different flags that may be zero or a combination of the following values:
joeverbout	0:ea44dc9ed014	901	- CV_CALIB_FIX_INTRINSIC Fix cameraMatrix? and distCoeffs? so that only R, T, E , and F
joeverbout	0:ea44dc9ed014	902	matrices are estimated.
joeverbout	0:ea44dc9ed014	903	- CV_CALIB_USE_INTRINSIC_GUESS Optimize some or all of the intrinsic parameters
joeverbout	0:ea44dc9ed014	904	according to the specified flags. Initial values are provided by the user.
joeverbout	0:ea44dc9ed014	905	- CV_CALIB_FIX_PRINCIPAL_POINT Fix the principal points during the optimization.
joeverbout	0:ea44dc9ed014	906	- CV_CALIB_FIX_FOCAL_LENGTH Fix \f$f^{(j)}_x\f$ and \f$f^{(j)}_y\f$ .
joeverbout	0:ea44dc9ed014	907	- CV_CALIB_FIX_ASPECT_RATIO Optimize \f$f^{(j)}_y\f$ . Fix the ratio \f$f^{(j)}_x/f^{(j)}_y\f$
joeverbout	0:ea44dc9ed014	908	.
joeverbout	0:ea44dc9ed014	909	- CV_CALIB_SAME_FOCAL_LENGTH Enforce \f$f^{(0)}_x=f^{(1)}_x\f$ and \f$f^{(0)}_y=f^{(1)}_y\f$ .
joeverbout	0:ea44dc9ed014	910	- CV_CALIB_ZERO_TANGENT_DIST Set tangential distortion coefficients for each camera to
joeverbout	0:ea44dc9ed014	911	zeros and fix there.
joeverbout	0:ea44dc9ed014	912	- CV_CALIB_FIX_K1,...,CV_CALIB_FIX_K6 Do not change the corresponding radial
joeverbout	0:ea44dc9ed014	913	distortion coefficient during the optimization. If CV_CALIB_USE_INTRINSIC_GUESS is set,
joeverbout	0:ea44dc9ed014	914	the coefficient from the supplied distCoeffs matrix is used. Otherwise, it is set to 0.
joeverbout	0:ea44dc9ed014	915	- CV_CALIB_RATIONAL_MODEL Enable coefficients k4, k5, and k6. To provide the backward
joeverbout	0:ea44dc9ed014	916	compatibility, this extra flag should be explicitly specified to make the calibration
joeverbout	0:ea44dc9ed014	917	function use the rational model and return 8 coefficients. If the flag is not set, the
joeverbout	0:ea44dc9ed014	918	function computes and returns only 5 distortion coefficients.
joeverbout	0:ea44dc9ed014	919	- CALIB_THIN_PRISM_MODEL Coefficients s1, s2, s3 and s4 are enabled. To provide the
joeverbout	0:ea44dc9ed014	920	backward compatibility, this extra flag should be explicitly specified to make the
joeverbout	0:ea44dc9ed014	921	calibration function use the thin prism model and return 12 coefficients. If the flag is not
joeverbout	0:ea44dc9ed014	922	set, the function computes and returns only 5 distortion coefficients.
joeverbout	0:ea44dc9ed014	923	- CALIB_FIX_S1_S2_S3_S4 The thin prism distortion coefficients are not changed during
joeverbout	0:ea44dc9ed014	924	the optimization. If CV_CALIB_USE_INTRINSIC_GUESS is set, the coefficient from the
joeverbout	0:ea44dc9ed014	925	supplied distCoeffs matrix is used. Otherwise, it is set to 0.
joeverbout	0:ea44dc9ed014	926	- CALIB_TILTED_MODEL Coefficients tauX and tauY are enabled. To provide the
joeverbout	0:ea44dc9ed014	927	backward compatibility, this extra flag should be explicitly specified to make the
joeverbout	0:ea44dc9ed014	928	calibration function use the tilted sensor model and return 14 coefficients. If the flag is not
joeverbout	0:ea44dc9ed014	929	set, the function computes and returns only 5 distortion coefficients.
joeverbout	0:ea44dc9ed014	930	- CALIB_FIX_TAUX_TAUY The coefficients of the tilted sensor model are not changed during
joeverbout	0:ea44dc9ed014	931	the optimization. If CV_CALIB_USE_INTRINSIC_GUESS is set, the coefficient from the
joeverbout	0:ea44dc9ed014	932	supplied distCoeffs matrix is used. Otherwise, it is set to 0.
joeverbout	0:ea44dc9ed014	933	@param criteria Termination criteria for the iterative optimization algorithm.
joeverbout	0:ea44dc9ed014	934
joeverbout	0:ea44dc9ed014	935	The function estimates transformation between two cameras making a stereo pair. If you have a stereo
joeverbout	0:ea44dc9ed014	936	camera where the relative position and orientation of two cameras is fixed, and if you computed
joeverbout	0:ea44dc9ed014	937	poses of an object relative to the first camera and to the second camera, (R1, T1) and (R2, T2),
joeverbout	0:ea44dc9ed014	938	respectively (this can be done with solvePnP ), then those poses definitely relate to each other.
joeverbout	0:ea44dc9ed014	939	This means that, given ( \f$R_1\f$,\f$T_1\f$ ), it should be possible to compute ( \f$R_2\f$,\f$T_2\f$ ). You only
joeverbout	0:ea44dc9ed014	940	need to know the position and orientation of the second camera relative to the first camera. This is
joeverbout	0:ea44dc9ed014	941	what the described function does. It computes ( \f$R\f$,\f$T\f$ ) so that:
joeverbout	0:ea44dc9ed014	942
joeverbout	0:ea44dc9ed014	943	\f[R_2=R*R_1
joeverbout	0:ea44dc9ed014	944	T_2=R*T_1 + T,\f]
joeverbout	0:ea44dc9ed014	945
joeverbout	0:ea44dc9ed014	946	Optionally, it computes the essential matrix E:
joeverbout	0:ea44dc9ed014	947
joeverbout	0:ea44dc9ed014	948	\f[E= \vecthreethree{0}{-T_2}{T_1}{T_2}{0}{-T_0}{-T_1}{T_0}{0} *R\f]
joeverbout	0:ea44dc9ed014	949
joeverbout	0:ea44dc9ed014	950	where \f$T_i\f$ are components of the translation vector \f$T\f$ : \f$T=[T_0, T_1, T_2]^T\f$ . And the function
joeverbout	0:ea44dc9ed014	951	can also compute the fundamental matrix F:
joeverbout	0:ea44dc9ed014	952
joeverbout	0:ea44dc9ed014	953	\f[F = cameraMatrix2^{-T} E cameraMatrix1^{-1}\f]
joeverbout	0:ea44dc9ed014	954
joeverbout	0:ea44dc9ed014	955	Besides the stereo-related information, the function can also perform a full calibration of each of
joeverbout	0:ea44dc9ed014	956	two cameras. However, due to the high dimensionality of the parameter space and noise in the input
joeverbout	0:ea44dc9ed014	957	data, the function can diverge from the correct solution. If the intrinsic parameters can be
joeverbout	0:ea44dc9ed014	958	estimated with high accuracy for each of the cameras individually (for example, using
joeverbout	0:ea44dc9ed014	959	calibrateCamera ), you are recommended to do so and then pass CV_CALIB_FIX_INTRINSIC flag to the
joeverbout	0:ea44dc9ed014	960	function along with the computed intrinsic parameters. Otherwise, if all the parameters are
joeverbout	0:ea44dc9ed014	961	estimated at once, it makes sense to restrict some parameters, for example, pass
joeverbout	0:ea44dc9ed014	962	CV_CALIB_SAME_FOCAL_LENGTH and CV_CALIB_ZERO_TANGENT_DIST flags, which is usually a
joeverbout	0:ea44dc9ed014	963	reasonable assumption.
joeverbout	0:ea44dc9ed014	964
joeverbout	0:ea44dc9ed014	965	Similarly to calibrateCamera , the function minimizes the total re-projection error for all the
joeverbout	0:ea44dc9ed014	966	points in all the available views from both cameras. The function returns the final value of the
joeverbout	0:ea44dc9ed014	967	re-projection error.
joeverbout	0:ea44dc9ed014	968	*/
joeverbout	0:ea44dc9ed014	969	CV_EXPORTS_W double stereoCalibrate( InputArrayOfArrays objectPoints,
joeverbout	0:ea44dc9ed014	970	InputArrayOfArrays imagePoints1, InputArrayOfArrays imagePoints2,
joeverbout	0:ea44dc9ed014	971	InputOutputArray cameraMatrix1, InputOutputArray distCoeffs1,
joeverbout	0:ea44dc9ed014	972	InputOutputArray cameraMatrix2, InputOutputArray distCoeffs2,
joeverbout	0:ea44dc9ed014	973	Size imageSize, OutputArray R,OutputArray T, OutputArray E, OutputArray F,
joeverbout	0:ea44dc9ed014	974	int flags = CALIB_FIX_INTRINSIC,
joeverbout	0:ea44dc9ed014	975	TermCriteria criteria = TermCriteria(TermCriteria::COUNT+TermCriteria::EPS, 30, 1e-6) );
joeverbout	0:ea44dc9ed014	976
joeverbout	0:ea44dc9ed014	977
joeverbout	0:ea44dc9ed014	978	/** @brief Computes rectification transforms for each head of a calibrated stereo camera.
joeverbout	0:ea44dc9ed014	979
joeverbout	0:ea44dc9ed014	980	@param cameraMatrix1 First camera matrix.
joeverbout	0:ea44dc9ed014	981	@param distCoeffs1 First camera distortion parameters.
joeverbout	0:ea44dc9ed014	982	@param cameraMatrix2 Second camera matrix.
joeverbout	0:ea44dc9ed014	983	@param distCoeffs2 Second camera distortion parameters.
joeverbout	0:ea44dc9ed014	984	@param imageSize Size of the image used for stereo calibration.
joeverbout	0:ea44dc9ed014	985	@param R Rotation matrix between the coordinate systems of the first and the second cameras.
joeverbout	0:ea44dc9ed014	986	@param T Translation vector between coordinate systems of the cameras.
joeverbout	0:ea44dc9ed014	987	@param R1 Output 3x3 rectification transform (rotation matrix) for the first camera.
joeverbout	0:ea44dc9ed014	988	@param R2 Output 3x3 rectification transform (rotation matrix) for the second camera.
joeverbout	0:ea44dc9ed014	989	@param P1 Output 3x4 projection matrix in the new (rectified) coordinate systems for the first
joeverbout	0:ea44dc9ed014	990	camera.
joeverbout	0:ea44dc9ed014	991	@param P2 Output 3x4 projection matrix in the new (rectified) coordinate systems for the second
joeverbout	0:ea44dc9ed014	992	camera.
joeverbout	0:ea44dc9ed014	993	@param Q Output \f$4 \times 4\f$ disparity-to-depth mapping matrix (see reprojectImageTo3D ).
joeverbout	0:ea44dc9ed014	994	@param flags Operation flags that may be zero or CV_CALIB_ZERO_DISPARITY . If the flag is set,
joeverbout	0:ea44dc9ed014	995	the function makes the principal points of each camera have the same pixel coordinates in the
joeverbout	0:ea44dc9ed014	996	rectified views. And if the flag is not set, the function may still shift the images in the
joeverbout	0:ea44dc9ed014	997	horizontal or vertical direction (depending on the orientation of epipolar lines) to maximize the
joeverbout	0:ea44dc9ed014	998	useful image area.
joeverbout	0:ea44dc9ed014	999	@param alpha Free scaling parameter. If it is -1 or absent, the function performs the default
joeverbout	0:ea44dc9ed014	1000	scaling. Otherwise, the parameter should be between 0 and 1. alpha=0 means that the rectified
joeverbout	0:ea44dc9ed014	1001	images are zoomed and shifted so that only valid pixels are visible (no black areas after
joeverbout	0:ea44dc9ed014	1002	rectification). alpha=1 means that the rectified image is decimated and shifted so that all the
joeverbout	0:ea44dc9ed014	1003	pixels from the original images from the cameras are retained in the rectified images (no source
joeverbout	0:ea44dc9ed014	1004	image pixels are lost). Obviously, any intermediate value yields an intermediate result between
joeverbout	0:ea44dc9ed014	1005	those two extreme cases.
joeverbout	0:ea44dc9ed014	1006	@param newImageSize New image resolution after rectification. The same size should be passed to
joeverbout	0:ea44dc9ed014	1007	initUndistortRectifyMap (see the stereo_calib.cpp sample in OpenCV samples directory). When (0,0)
joeverbout	0:ea44dc9ed014	1008	is passed (default), it is set to the original imageSize . Setting it to larger value can help you
joeverbout	0:ea44dc9ed014	1009	preserve details in the original image, especially when there is a big radial distortion.
joeverbout	0:ea44dc9ed014	1010	@param validPixROI1 Optional output rectangles inside the rectified images where all the pixels
joeverbout	0:ea44dc9ed014	1011	are valid. If alpha=0 , the ROIs cover the whole images. Otherwise, they are likely to be smaller
joeverbout	0:ea44dc9ed014	1012	(see the picture below).
joeverbout	0:ea44dc9ed014	1013	@param validPixROI2 Optional output rectangles inside the rectified images where all the pixels
joeverbout	0:ea44dc9ed014	1014	are valid. If alpha=0 , the ROIs cover the whole images. Otherwise, they are likely to be smaller
joeverbout	0:ea44dc9ed014	1015	(see the picture below).
joeverbout	0:ea44dc9ed014	1016
joeverbout	0:ea44dc9ed014	1017	The function computes the rotation matrices for each camera that (virtually) make both camera image
joeverbout	0:ea44dc9ed014	1018	planes the same plane. Consequently, this makes all the epipolar lines parallel and thus simplifies
joeverbout	0:ea44dc9ed014	1019	the dense stereo correspondence problem. The function takes the matrices computed by stereoCalibrate
joeverbout	0:ea44dc9ed014	1020	as input. As output, it provides two rotation matrices and also two projection matrices in the new
joeverbout	0:ea44dc9ed014	1021	coordinates. The function distinguishes the following two cases:
joeverbout	0:ea44dc9ed014	1022
joeverbout	0:ea44dc9ed014	1023	- Horizontal stereo: the first and the second camera views are shifted relative to each other
joeverbout	0:ea44dc9ed014	1024	mainly along the x axis (with possible small vertical shift). In the rectified images, the
joeverbout	0:ea44dc9ed014	1025	corresponding epipolar lines in the left and right cameras are horizontal and have the same
joeverbout	0:ea44dc9ed014	1026	y-coordinate. P1 and P2 look like:
joeverbout	0:ea44dc9ed014	1027
joeverbout	0:ea44dc9ed014	1028	\f[\texttt{P1} = \begin{bmatrix} f & 0 & cx_1 & 0 \\ 0 & f & cy & 0 \\ 0 & 0 & 1 & 0 \end{bmatrix}\f]
joeverbout	0:ea44dc9ed014	1029
joeverbout	0:ea44dc9ed014	1030	\f[\texttt{P2} = \begin{bmatrix} f & 0 & cx_2 & T_x*f \\ 0 & f & cy & 0 \\ 0 & 0 & 1 & 0 \end{bmatrix} ,\f]
joeverbout	0:ea44dc9ed014	1031
joeverbout	0:ea44dc9ed014	1032	where \f$T_x\f$ is a horizontal shift between the cameras and \f$cx_1=cx_2\f$ if
joeverbout	0:ea44dc9ed014	1033	CV_CALIB_ZERO_DISPARITY is set.
joeverbout	0:ea44dc9ed014	1034
joeverbout	0:ea44dc9ed014	1035	- Vertical stereo: the first and the second camera views are shifted relative to each other
joeverbout	0:ea44dc9ed014	1036	mainly in vertical direction (and probably a bit in the horizontal direction too). The epipolar
joeverbout	0:ea44dc9ed014	1037	lines in the rectified images are vertical and have the same x-coordinate. P1 and P2 look like:
joeverbout	0:ea44dc9ed014	1038
joeverbout	0:ea44dc9ed014	1039	\f[\texttt{P1} = \begin{bmatrix} f & 0 & cx & 0 \\ 0 & f & cy_1 & 0 \\ 0 & 0 & 1 & 0 \end{bmatrix}\f]
joeverbout	0:ea44dc9ed014	1040
joeverbout	0:ea44dc9ed014	1041	\f[\texttt{P2} = \begin{bmatrix} f & 0 & cx & 0 \\ 0 & f & cy_2 & T_y*f \\ 0 & 0 & 1 & 0 \end{bmatrix} ,\f]
joeverbout	0:ea44dc9ed014	1042
joeverbout	0:ea44dc9ed014	1043	where \f$T_y\f$ is a vertical shift between the cameras and \f$cy_1=cy_2\f$ if CALIB_ZERO_DISPARITY is
joeverbout	0:ea44dc9ed014	1044	set.
joeverbout	0:ea44dc9ed014	1045
joeverbout	0:ea44dc9ed014	1046	As you can see, the first three columns of P1 and P2 will effectively be the new "rectified" camera
joeverbout	0:ea44dc9ed014	1047	matrices. The matrices, together with R1 and R2 , can then be passed to initUndistortRectifyMap to
joeverbout	0:ea44dc9ed014	1048	initialize the rectification map for each camera.
joeverbout	0:ea44dc9ed014	1049
joeverbout	0:ea44dc9ed014	1050	See below the screenshot from the stereo_calib.cpp sample. Some red horizontal lines pass through
joeverbout	0:ea44dc9ed014	1051	the corresponding image regions. This means that the images are well rectified, which is what most
joeverbout	0:ea44dc9ed014	1052	stereo correspondence algorithms rely on. The green rectangles are roi1 and roi2 . You see that
joeverbout	0:ea44dc9ed014	1053	their interiors are all valid pixels.
joeverbout	0:ea44dc9ed014	1054
joeverbout	0:ea44dc9ed014	1055	![image](pics/stereo_undistort.jpg)
joeverbout	0:ea44dc9ed014	1056	*/
joeverbout	0:ea44dc9ed014	1057	CV_EXPORTS_W void stereoRectify( InputArray cameraMatrix1, InputArray distCoeffs1,
joeverbout	0:ea44dc9ed014	1058	InputArray cameraMatrix2, InputArray distCoeffs2,
joeverbout	0:ea44dc9ed014	1059	Size imageSize, InputArray R, InputArray T,
joeverbout	0:ea44dc9ed014	1060	OutputArray R1, OutputArray R2,
joeverbout	0:ea44dc9ed014	1061	OutputArray P1, OutputArray P2,
joeverbout	0:ea44dc9ed014	1062	OutputArray Q, int flags = CALIB_ZERO_DISPARITY,
joeverbout	0:ea44dc9ed014	1063	double alpha = -1, Size newImageSize = Size(),
joeverbout	0:ea44dc9ed014	1064	CV_OUT Rect* validPixROI1 = 0, CV_OUT Rect* validPixROI2 = 0 );
joeverbout	0:ea44dc9ed014	1065
joeverbout	0:ea44dc9ed014	1066	/** @brief Computes a rectification transform for an uncalibrated stereo camera.
joeverbout	0:ea44dc9ed014	1067
joeverbout	0:ea44dc9ed014	1068	@param points1 Array of feature points in the first image.
joeverbout	0:ea44dc9ed014	1069	@param points2 The corresponding points in the second image. The same formats as in
joeverbout	0:ea44dc9ed014	1070	findFundamentalMat are supported.
joeverbout	0:ea44dc9ed014	1071	@param F Input fundamental matrix. It can be computed from the same set of point pairs using
joeverbout	0:ea44dc9ed014	1072	findFundamentalMat .
joeverbout	0:ea44dc9ed014	1073	@param imgSize Size of the image.
joeverbout	0:ea44dc9ed014	1074	@param H1 Output rectification homography matrix for the first image.
joeverbout	0:ea44dc9ed014	1075	@param H2 Output rectification homography matrix for the second image.
joeverbout	0:ea44dc9ed014	1076	@param threshold Optional threshold used to filter out the outliers. If the parameter is greater
joeverbout	0:ea44dc9ed014	1077	than zero, all the point pairs that do not comply with the epipolar geometry (that is, the points
joeverbout	0:ea44dc9ed014	1078	for which \f$\|\texttt{points2[i]}^T\texttt{F}\texttt{points1[i]}\|>\texttt{threshold}\f$ ) are
joeverbout	0:ea44dc9ed014	1079	rejected prior to computing the homographies. Otherwise,all the points are considered inliers.
joeverbout	0:ea44dc9ed014	1080
joeverbout	0:ea44dc9ed014	1081	The function computes the rectification transformations without knowing intrinsic parameters of the
joeverbout	0:ea44dc9ed014	1082	cameras and their relative position in the space, which explains the suffix "uncalibrated". Another
joeverbout	0:ea44dc9ed014	1083	related difference from stereoRectify is that the function outputs not the rectification
joeverbout	0:ea44dc9ed014	1084	transformations in the object (3D) space, but the planar perspective transformations encoded by the
joeverbout	0:ea44dc9ed014	1085	homography matrices H1 and H2 . The function implements the algorithm @cite Hartley99 .
joeverbout	0:ea44dc9ed014	1086
joeverbout	0:ea44dc9ed014	1087	@note
joeverbout	0:ea44dc9ed014	1088	While the algorithm does not need to know the intrinsic parameters of the cameras, it heavily
joeverbout	0:ea44dc9ed014	1089	depends on the epipolar geometry. Therefore, if the camera lenses have a significant distortion,
joeverbout	0:ea44dc9ed014	1090	it would be better to correct it before computing the fundamental matrix and calling this
joeverbout	0:ea44dc9ed014	1091	function. For example, distortion coefficients can be estimated for each head of stereo camera
joeverbout	0:ea44dc9ed014	1092	separately by using calibrateCamera . Then, the images can be corrected using undistort , or
joeverbout	0:ea44dc9ed014	1093	just the point coordinates can be corrected with undistortPoints .
joeverbout	0:ea44dc9ed014	1094	*/
joeverbout	0:ea44dc9ed014	1095	CV_EXPORTS_W bool stereoRectifyUncalibrated( InputArray points1, InputArray points2,
joeverbout	0:ea44dc9ed014	1096	InputArray F, Size imgSize,
joeverbout	0:ea44dc9ed014	1097	OutputArray H1, OutputArray H2,
joeverbout	0:ea44dc9ed014	1098	double threshold = 5 );
joeverbout	0:ea44dc9ed014	1099
joeverbout	0:ea44dc9ed014	1100	//! computes the rectification transformations for 3-head camera, where all the heads are on the same line.
joeverbout	0:ea44dc9ed014	1101	CV_EXPORTS_W float rectify3Collinear( InputArray cameraMatrix1, InputArray distCoeffs1,
joeverbout	0:ea44dc9ed014	1102	InputArray cameraMatrix2, InputArray distCoeffs2,
joeverbout	0:ea44dc9ed014	1103	InputArray cameraMatrix3, InputArray distCoeffs3,
joeverbout	0:ea44dc9ed014	1104	InputArrayOfArrays imgpt1, InputArrayOfArrays imgpt3,
joeverbout	0:ea44dc9ed014	1105	Size imageSize, InputArray R12, InputArray T12,
joeverbout	0:ea44dc9ed014	1106	InputArray R13, InputArray T13,
joeverbout	0:ea44dc9ed014	1107	OutputArray R1, OutputArray R2, OutputArray R3,
joeverbout	0:ea44dc9ed014	1108	OutputArray P1, OutputArray P2, OutputArray P3,
joeverbout	0:ea44dc9ed014	1109	OutputArray Q, double alpha, Size newImgSize,
joeverbout	0:ea44dc9ed014	1110	CV_OUT Rect* roi1, CV_OUT Rect* roi2, int flags );
joeverbout	0:ea44dc9ed014	1111
joeverbout	0:ea44dc9ed014	1112	/** @brief Returns the new camera matrix based on the free scaling parameter.
joeverbout	0:ea44dc9ed014	1113
joeverbout	0:ea44dc9ed014	1114	@param cameraMatrix Input camera matrix.
joeverbout	0:ea44dc9ed014	1115	@param distCoeffs Input vector of distortion coefficients
joeverbout	0:ea44dc9ed014	1116	\f$(k_1, k_2, p_1, p_2[, k_3[, k_4, k_5, k_6 [, s_1, s_2, s_3, s_4[, \tau_x, \tau_y]]]])\f$ of
joeverbout	0:ea44dc9ed014	1117	4, 5, 8, 12 or 14 elements. If the vector is NULL/empty, the zero distortion coefficients are
joeverbout	0:ea44dc9ed014	1118	assumed.
joeverbout	0:ea44dc9ed014	1119	@param imageSize Original image size.
joeverbout	0:ea44dc9ed014	1120	@param alpha Free scaling parameter between 0 (when all the pixels in the undistorted image are
joeverbout	0:ea44dc9ed014	1121	valid) and 1 (when all the source image pixels are retained in the undistorted image). See
joeverbout	0:ea44dc9ed014	1122	stereoRectify for details.
joeverbout	0:ea44dc9ed014	1123	@param newImgSize Image size after rectification. By default,it is set to imageSize .
joeverbout	0:ea44dc9ed014	1124	@param validPixROI Optional output rectangle that outlines all-good-pixels region in the
joeverbout	0:ea44dc9ed014	1125	undistorted image. See roi1, roi2 description in stereoRectify .
joeverbout	0:ea44dc9ed014	1126	@param centerPrincipalPoint Optional flag that indicates whether in the new camera matrix the
joeverbout	0:ea44dc9ed014	1127	principal point should be at the image center or not. By default, the principal point is chosen to
joeverbout	0:ea44dc9ed014	1128	best fit a subset of the source image (determined by alpha) to the corrected image.
joeverbout	0:ea44dc9ed014	1129	@return new_camera_matrix Output new camera matrix.
joeverbout	0:ea44dc9ed014	1130
joeverbout	0:ea44dc9ed014	1131	The function computes and returns the optimal new camera matrix based on the free scaling parameter.
joeverbout	0:ea44dc9ed014	1132	By varying this parameter, you may retrieve only sensible pixels alpha=0 , keep all the original
joeverbout	0:ea44dc9ed014	1133	image pixels if there is valuable information in the corners alpha=1 , or get something in between.
joeverbout	0:ea44dc9ed014	1134	When alpha\>0 , the undistortion result is likely to have some black pixels corresponding to
joeverbout	0:ea44dc9ed014	1135	"virtual" pixels outside of the captured distorted image. The original camera matrix, distortion
joeverbout	0:ea44dc9ed014	1136	coefficients, the computed new camera matrix, and newImageSize should be passed to
joeverbout	0:ea44dc9ed014	1137	initUndistortRectifyMap to produce the maps for remap .
joeverbout	0:ea44dc9ed014	1138	*/
joeverbout	0:ea44dc9ed014	1139	CV_EXPORTS_W Mat getOptimalNewCameraMatrix( InputArray cameraMatrix, InputArray distCoeffs,
joeverbout	0:ea44dc9ed014	1140	Size imageSize, double alpha, Size newImgSize = Size(),
joeverbout	0:ea44dc9ed014	1141	CV_OUT Rect* validPixROI = 0,
joeverbout	0:ea44dc9ed014	1142	bool centerPrincipalPoint = false);
joeverbout	0:ea44dc9ed014	1143
joeverbout	0:ea44dc9ed014	1144	/** @brief Converts points from Euclidean to homogeneous space.
joeverbout	0:ea44dc9ed014	1145
joeverbout	0:ea44dc9ed014	1146	@param src Input vector of N-dimensional points.
joeverbout	0:ea44dc9ed014	1147	@param dst Output vector of N+1-dimensional points.
joeverbout	0:ea44dc9ed014	1148
joeverbout	0:ea44dc9ed014	1149	The function converts points from Euclidean to homogeneous space by appending 1's to the tuple of
joeverbout	0:ea44dc9ed014	1150	point coordinates. That is, each point (x1, x2, ..., xn) is converted to (x1, x2, ..., xn, 1).
joeverbout	0:ea44dc9ed014	1151	*/
joeverbout	0:ea44dc9ed014	1152	CV_EXPORTS_W void convertPointsToHomogeneous( InputArray src, OutputArray dst );
joeverbout	0:ea44dc9ed014	1153
joeverbout	0:ea44dc9ed014	1154	/** @brief Converts points from homogeneous to Euclidean space.
joeverbout	0:ea44dc9ed014	1155
joeverbout	0:ea44dc9ed014	1156	@param src Input vector of N-dimensional points.
joeverbout	0:ea44dc9ed014	1157	@param dst Output vector of N-1-dimensional points.
joeverbout	0:ea44dc9ed014	1158
joeverbout	0:ea44dc9ed014	1159	The function converts points homogeneous to Euclidean space using perspective projection. That is,
joeverbout	0:ea44dc9ed014	1160	each point (x1, x2, ... x(n-1), xn) is converted to (x1/xn, x2/xn, ..., x(n-1)/xn). When xn=0, the
joeverbout	0:ea44dc9ed014	1161	output point coordinates will be (0,0,0,...).
joeverbout	0:ea44dc9ed014	1162	*/
joeverbout	0:ea44dc9ed014	1163	CV_EXPORTS_W void convertPointsFromHomogeneous( InputArray src, OutputArray dst );
joeverbout	0:ea44dc9ed014	1164
joeverbout	0:ea44dc9ed014	1165	/** @brief Converts points to/from homogeneous coordinates.
joeverbout	0:ea44dc9ed014	1166
joeverbout	0:ea44dc9ed014	1167	@param src Input array or vector of 2D, 3D, or 4D points.
joeverbout	0:ea44dc9ed014	1168	@param dst Output vector of 2D, 3D, or 4D points.
joeverbout	0:ea44dc9ed014	1169
joeverbout	0:ea44dc9ed014	1170	The function converts 2D or 3D points from/to homogeneous coordinates by calling either
joeverbout	0:ea44dc9ed014	1171	convertPointsToHomogeneous or convertPointsFromHomogeneous.
joeverbout	0:ea44dc9ed014	1172
joeverbout	0:ea44dc9ed014	1173	@note The function is obsolete. Use one of the previous two functions instead.
joeverbout	0:ea44dc9ed014	1174	*/
joeverbout	0:ea44dc9ed014	1175	CV_EXPORTS void convertPointsHomogeneous( InputArray src, OutputArray dst );
joeverbout	0:ea44dc9ed014	1176
joeverbout	0:ea44dc9ed014	1177	/** @brief Calculates a fundamental matrix from the corresponding points in two images.
joeverbout	0:ea44dc9ed014	1178
joeverbout	0:ea44dc9ed014	1179	@param points1 Array of N points from the first image. The point coordinates should be
joeverbout	0:ea44dc9ed014	1180	floating-point (single or double precision).
joeverbout	0:ea44dc9ed014	1181	@param points2 Array of the second image points of the same size and format as points1 .
joeverbout	0:ea44dc9ed014	1182	@param method Method for computing a fundamental matrix.
joeverbout	0:ea44dc9ed014	1183	- CV_FM_7POINT for a 7-point algorithm. \f$N = 7\f$
joeverbout	0:ea44dc9ed014	1184	- CV_FM_8POINT for an 8-point algorithm. \f$N \ge 8\f$
joeverbout	0:ea44dc9ed014	1185	- CV_FM_RANSAC for the RANSAC algorithm. \f$N \ge 8\f$
joeverbout	0:ea44dc9ed014	1186	- CV_FM_LMEDS for the LMedS algorithm. \f$N \ge 8\f$
joeverbout	0:ea44dc9ed014	1187	@param param1 Parameter used for RANSAC. It is the maximum distance from a point to an epipolar
joeverbout	0:ea44dc9ed014	1188	line in pixels, beyond which the point is considered an outlier and is not used for computing the
joeverbout	0:ea44dc9ed014	1189	final fundamental matrix. It can be set to something like 1-3, depending on the accuracy of the
joeverbout	0:ea44dc9ed014	1190	point localization, image resolution, and the image noise.
joeverbout	0:ea44dc9ed014	1191	@param param2 Parameter used for the RANSAC or LMedS methods only. It specifies a desirable level
joeverbout	0:ea44dc9ed014	1192	of confidence (probability) that the estimated matrix is correct.
joeverbout	0:ea44dc9ed014	1193	@param mask
joeverbout	0:ea44dc9ed014	1194
joeverbout	0:ea44dc9ed014	1195	The epipolar geometry is described by the following equation:
joeverbout	0:ea44dc9ed014	1196
joeverbout	0:ea44dc9ed014	1197	\f[[p_2; 1]^T F [p_1; 1] = 0\f]
joeverbout	0:ea44dc9ed014	1198
joeverbout	0:ea44dc9ed014	1199	where \f$F\f$ is a fundamental matrix, \f$p_1\f$ and \f$p_2\f$ are corresponding points in the first and the
joeverbout	0:ea44dc9ed014	1200	second images, respectively.
joeverbout	0:ea44dc9ed014	1201
joeverbout	0:ea44dc9ed014	1202	The function calculates the fundamental matrix using one of four methods listed above and returns
joeverbout	0:ea44dc9ed014	1203	the found fundamental matrix. Normally just one matrix is found. But in case of the 7-point
joeverbout	0:ea44dc9ed014	1204	algorithm, the function may return up to 3 solutions ( \f$9 \times 3\f$ matrix that stores all 3
joeverbout	0:ea44dc9ed014	1205	matrices sequentially).
joeverbout	0:ea44dc9ed014	1206
joeverbout	0:ea44dc9ed014	1207	The calculated fundamental matrix may be passed further to computeCorrespondEpilines that finds the
joeverbout	0:ea44dc9ed014	1208	epipolar lines corresponding to the specified points. It can also be passed to
joeverbout	0:ea44dc9ed014	1209	stereoRectifyUncalibrated to compute the rectification transformation. :
joeverbout	0:ea44dc9ed014	1210	@code
joeverbout	0:ea44dc9ed014	1211	// Example. Estimation of fundamental matrix using the RANSAC algorithm
joeverbout	0:ea44dc9ed014	1212	int point_count = 100;
joeverbout	0:ea44dc9ed014	1213	vector<Point2f> points1(point_count);
joeverbout	0:ea44dc9ed014	1214	vector<Point2f> points2(point_count);
joeverbout	0:ea44dc9ed014	1215
joeverbout	0:ea44dc9ed014	1216	// initialize the points here ...
joeverbout	0:ea44dc9ed014	1217	for( int i = 0; i < point_count; i++ )
joeverbout	0:ea44dc9ed014	1218	{
joeverbout	0:ea44dc9ed014	1219	points1[i] = ...;
joeverbout	0:ea44dc9ed014	1220	points2[i] = ...;
joeverbout	0:ea44dc9ed014	1221	}
joeverbout	0:ea44dc9ed014	1222
joeverbout	0:ea44dc9ed014	1223	Mat fundamental_matrix =
joeverbout	0:ea44dc9ed014	1224	findFundamentalMat(points1, points2, FM_RANSAC, 3, 0.99);
joeverbout	0:ea44dc9ed014	1225	@endcode
joeverbout	0:ea44dc9ed014	1226	*/
joeverbout	0:ea44dc9ed014	1227	CV_EXPORTS_W Mat findFundamentalMat( InputArray points1, InputArray points2,
joeverbout	0:ea44dc9ed014	1228	int method = FM_RANSAC,
joeverbout	0:ea44dc9ed014	1229	double param1 = 3., double param2 = 0.99,
joeverbout	0:ea44dc9ed014	1230	OutputArray mask = noArray() );
joeverbout	0:ea44dc9ed014	1231
joeverbout	0:ea44dc9ed014	1232	/** @overload */
joeverbout	0:ea44dc9ed014	1233	CV_EXPORTS Mat findFundamentalMat( InputArray points1, InputArray points2,
joeverbout	0:ea44dc9ed014	1234	OutputArray mask, int method = FM_RANSAC,
joeverbout	0:ea44dc9ed014	1235	double param1 = 3., double param2 = 0.99 );
joeverbout	0:ea44dc9ed014	1236
joeverbout	0:ea44dc9ed014	1237	/** @brief Calculates an essential matrix from the corresponding points in two images.
joeverbout	0:ea44dc9ed014	1238
joeverbout	0:ea44dc9ed014	1239	@param points1 Array of N (N \>= 5) 2D points from the first image. The point coordinates should
joeverbout	0:ea44dc9ed014	1240	be floating-point (single or double precision).
joeverbout	0:ea44dc9ed014	1241	@param points2 Array of the second image points of the same size and format as points1 .
joeverbout	0:ea44dc9ed014	1242	@param cameraMatrix Camera matrix \f$K = \vecthreethree{f_x}{0}{c_x}{0}{f_y}{c_y}{0}{0}{1}\f$ .
joeverbout	0:ea44dc9ed014	1243	Note that this function assumes that points1 and points2 are feature points from cameras with the
joeverbout	0:ea44dc9ed014	1244	same camera matrix.
joeverbout	0:ea44dc9ed014	1245	@param method Method for computing a fundamental matrix.
joeverbout	0:ea44dc9ed014	1246	- RANSAC for the RANSAC algorithm.
joeverbout	0:ea44dc9ed014	1247	- MEDS for the LMedS algorithm.
joeverbout	0:ea44dc9ed014	1248	@param threshold Parameter used for RANSAC. It is the maximum distance from a point to an epipolar
joeverbout	0:ea44dc9ed014	1249	line in pixels, beyond which the point is considered an outlier and is not used for computing the
joeverbout	0:ea44dc9ed014	1250	final fundamental matrix. It can be set to something like 1-3, depending on the accuracy of the
joeverbout	0:ea44dc9ed014	1251	point localization, image resolution, and the image noise.
joeverbout	0:ea44dc9ed014	1252	@param prob Parameter used for the RANSAC or LMedS methods only. It specifies a desirable level of
joeverbout	0:ea44dc9ed014	1253	confidence (probability) that the estimated matrix is correct.
joeverbout	0:ea44dc9ed014	1254	@param mask Output array of N elements, every element of which is set to 0 for outliers and to 1
joeverbout	0:ea44dc9ed014	1255	for the other points. The array is computed only in the RANSAC and LMedS methods.
joeverbout	0:ea44dc9ed014	1256
joeverbout	0:ea44dc9ed014	1257	This function estimates essential matrix based on the five-point algorithm solver in @cite Nister03 .
joeverbout	0:ea44dc9ed014	1258	@cite SteweniusCFS is also a related. The epipolar geometry is described by the following equation:
joeverbout	0:ea44dc9ed014	1259
joeverbout	0:ea44dc9ed014	1260	\f[[p_2; 1]^T K^{-T} E K^{-1} [p_1; 1] = 0\f]
joeverbout	0:ea44dc9ed014	1261
joeverbout	0:ea44dc9ed014	1262	where \f$E\f$ is an essential matrix, \f$p_1\f$ and \f$p_2\f$ are corresponding points in the first and the
joeverbout	0:ea44dc9ed014	1263	second images, respectively. The result of this function may be passed further to
joeverbout	0:ea44dc9ed014	1264	decomposeEssentialMat or recoverPose to recover the relative pose between cameras.
joeverbout	0:ea44dc9ed014	1265	*/
joeverbout	0:ea44dc9ed014	1266	CV_EXPORTS_W Mat findEssentialMat( InputArray points1, InputArray points2,
joeverbout	0:ea44dc9ed014	1267	InputArray cameraMatrix, int method = RANSAC,
joeverbout	0:ea44dc9ed014	1268	double prob = 0.999, double threshold = 1.0,
joeverbout	0:ea44dc9ed014	1269	OutputArray mask = noArray() );
joeverbout	0:ea44dc9ed014	1270
joeverbout	0:ea44dc9ed014	1271	/** @overload
joeverbout	0:ea44dc9ed014	1272	@param points1 Array of N (N \>= 5) 2D points from the first image. The point coordinates should
joeverbout	0:ea44dc9ed014	1273	be floating-point (single or double precision).
joeverbout	0:ea44dc9ed014	1274	@param points2 Array of the second image points of the same size and format as points1 .
joeverbout	0:ea44dc9ed014	1275	@param focal focal length of the camera. Note that this function assumes that points1 and points2
joeverbout	0:ea44dc9ed014	1276	are feature points from cameras with same focal length and principle point.
joeverbout	0:ea44dc9ed014	1277	@param pp principle point of the camera.
joeverbout	0:ea44dc9ed014	1278	@param method Method for computing a fundamental matrix.
joeverbout	0:ea44dc9ed014	1279	- RANSAC for the RANSAC algorithm.
joeverbout	0:ea44dc9ed014	1280	- LMEDS for the LMedS algorithm.
joeverbout	0:ea44dc9ed014	1281	@param threshold Parameter used for RANSAC. It is the maximum distance from a point to an epipolar
joeverbout	0:ea44dc9ed014	1282	line in pixels, beyond which the point is considered an outlier and is not used for computing the
joeverbout	0:ea44dc9ed014	1283	final fundamental matrix. It can be set to something like 1-3, depending on the accuracy of the
joeverbout	0:ea44dc9ed014	1284	point localization, image resolution, and the image noise.
joeverbout	0:ea44dc9ed014	1285	@param prob Parameter used for the RANSAC or LMedS methods only. It specifies a desirable level of
joeverbout	0:ea44dc9ed014	1286	confidence (probability) that the estimated matrix is correct.
joeverbout	0:ea44dc9ed014	1287	@param mask Output array of N elements, every element of which is set to 0 for outliers and to 1
joeverbout	0:ea44dc9ed014	1288	for the other points. The array is computed only in the RANSAC and LMedS methods.
joeverbout	0:ea44dc9ed014	1289
joeverbout	0:ea44dc9ed014	1290	This function differs from the one above that it computes camera matrix from focal length and
joeverbout	0:ea44dc9ed014	1291	principal point:
joeverbout	0:ea44dc9ed014	1292
joeverbout	0:ea44dc9ed014	1293	\f[K =
joeverbout	0:ea44dc9ed014	1294	\begin{bmatrix}
joeverbout	0:ea44dc9ed014	1295	f & 0 & x_{pp} \\
joeverbout	0:ea44dc9ed014	1296	0 & f & y_{pp} \\
joeverbout	0:ea44dc9ed014	1297	0 & 0 & 1
joeverbout	0:ea44dc9ed014	1298	\end{bmatrix}\f]
joeverbout	0:ea44dc9ed014	1299	*/
joeverbout	0:ea44dc9ed014	1300	CV_EXPORTS_W Mat findEssentialMat( InputArray points1, InputArray points2,
joeverbout	0:ea44dc9ed014	1301	double focal = 1.0, Point2d pp = Point2d(0, 0),
joeverbout	0:ea44dc9ed014	1302	int method = RANSAC, double prob = 0.999,
joeverbout	0:ea44dc9ed014	1303	double threshold = 1.0, OutputArray mask = noArray() );
joeverbout	0:ea44dc9ed014	1304
joeverbout	0:ea44dc9ed014	1305	/** @brief Decompose an essential matrix to possible rotations and translation.
joeverbout	0:ea44dc9ed014	1306
joeverbout	0:ea44dc9ed014	1307	@param E The input essential matrix.
joeverbout	0:ea44dc9ed014	1308	@param R1 One possible rotation matrix.
joeverbout	0:ea44dc9ed014	1309	@param R2 Another possible rotation matrix.
joeverbout	0:ea44dc9ed014	1310	@param t One possible translation.
joeverbout	0:ea44dc9ed014	1311
joeverbout	0:ea44dc9ed014	1312	This function decompose an essential matrix E using svd decomposition @cite HartleyZ00 . Generally 4
joeverbout	0:ea44dc9ed014	1313	possible poses exists for a given E. They are \f$[R_1, t]\f$, \f$[R_1, -t]\f$, \f$[R_2, t]\f$, \f$[R_2, -t]\f$. By
joeverbout	0:ea44dc9ed014	1314	decomposing E, you can only get the direction of the translation, so the function returns unit t.
joeverbout	0:ea44dc9ed014	1315	*/
joeverbout	0:ea44dc9ed014	1316	CV_EXPORTS_W void decomposeEssentialMat( InputArray E, OutputArray R1, OutputArray R2, OutputArray t );
joeverbout	0:ea44dc9ed014	1317
joeverbout	0:ea44dc9ed014	1318	/** @brief Recover relative camera rotation and translation from an estimated essential matrix and the
joeverbout	0:ea44dc9ed014	1319	corresponding points in two images, using cheirality check. Returns the number of inliers which pass
joeverbout	0:ea44dc9ed014	1320	the check.
joeverbout	0:ea44dc9ed014	1321
joeverbout	0:ea44dc9ed014	1322	@param E The input essential matrix.
joeverbout	0:ea44dc9ed014	1323	@param points1 Array of N 2D points from the first image. The point coordinates should be
joeverbout	0:ea44dc9ed014	1324	floating-point (single or double precision).
joeverbout	0:ea44dc9ed014	1325	@param points2 Array of the second image points of the same size and format as points1 .
joeverbout	0:ea44dc9ed014	1326	@param cameraMatrix Camera matrix \f$K = \vecthreethree{f_x}{0}{c_x}{0}{f_y}{c_y}{0}{0}{1}\f$ .
joeverbout	0:ea44dc9ed014	1327	Note that this function assumes that points1 and points2 are feature points from cameras with the
joeverbout	0:ea44dc9ed014	1328	same camera matrix.
joeverbout	0:ea44dc9ed014	1329	@param R Recovered relative rotation.
joeverbout	0:ea44dc9ed014	1330	@param t Recoverd relative translation.
joeverbout	0:ea44dc9ed014	1331	@param mask Input/output mask for inliers in points1 and points2.
joeverbout	0:ea44dc9ed014	1332	: If it is not empty, then it marks inliers in points1 and points2 for then given essential
joeverbout	0:ea44dc9ed014	1333	matrix E. Only these inliers will be used to recover pose. In the output mask only inliers
joeverbout	0:ea44dc9ed014	1334	which pass the cheirality check.
joeverbout	0:ea44dc9ed014	1335	This function decomposes an essential matrix using decomposeEssentialMat and then verifies possible
joeverbout	0:ea44dc9ed014	1336	pose hypotheses by doing cheirality check. The cheirality check basically means that the
joeverbout	0:ea44dc9ed014	1337	triangulated 3D points should have positive depth. Some details can be found in @cite Nister03 .
joeverbout	0:ea44dc9ed014	1338
joeverbout	0:ea44dc9ed014	1339	This function can be used to process output E and mask from findEssentialMat. In this scenario,
joeverbout	0:ea44dc9ed014	1340	points1 and points2 are the same input for findEssentialMat. :
joeverbout	0:ea44dc9ed014	1341	@code
joeverbout	0:ea44dc9ed014	1342	// Example. Estimation of fundamental matrix using the RANSAC algorithm
joeverbout	0:ea44dc9ed014	1343	int point_count = 100;
joeverbout	0:ea44dc9ed014	1344	vector<Point2f> points1(point_count);
joeverbout	0:ea44dc9ed014	1345	vector<Point2f> points2(point_count);
joeverbout	0:ea44dc9ed014	1346
joeverbout	0:ea44dc9ed014	1347	// initialize the points here ...
joeverbout	0:ea44dc9ed014	1348	for( int i = 0; i < point_count; i++ )
joeverbout	0:ea44dc9ed014	1349	{
joeverbout	0:ea44dc9ed014	1350	points1[i] = ...;
joeverbout	0:ea44dc9ed014	1351	points2[i] = ...;
joeverbout	0:ea44dc9ed014	1352	}
joeverbout	0:ea44dc9ed014	1353
joeverbout	0:ea44dc9ed014	1354	// cametra matrix with both focal lengths = 1, and principal point = (0, 0)
joeverbout	0:ea44dc9ed014	1355	Mat cameraMatrix = Mat::eye(3, 3, CV_64F);
joeverbout	0:ea44dc9ed014	1356
joeverbout	0:ea44dc9ed014	1357	Mat E, R, t, mask;
joeverbout	0:ea44dc9ed014	1358
joeverbout	0:ea44dc9ed014	1359	E = findEssentialMat(points1, points2, cameraMatrix, RANSAC, 0.999, 1.0, mask);
joeverbout	0:ea44dc9ed014	1360	recoverPose(E, points1, points2, cameraMatrix, R, t, mask);
joeverbout	0:ea44dc9ed014	1361	@endcode
joeverbout	0:ea44dc9ed014	1362	*/
joeverbout	0:ea44dc9ed014	1363	CV_EXPORTS_W int recoverPose( InputArray E, InputArray points1, InputArray points2,
joeverbout	0:ea44dc9ed014	1364	InputArray cameraMatrix, OutputArray R, OutputArray t,
joeverbout	0:ea44dc9ed014	1365	InputOutputArray mask = noArray() );
joeverbout	0:ea44dc9ed014	1366
joeverbout	0:ea44dc9ed014	1367	/** @overload
joeverbout	0:ea44dc9ed014	1368	@param E The input essential matrix.
joeverbout	0:ea44dc9ed014	1369	@param points1 Array of N 2D points from the first image. The point coordinates should be
joeverbout	0:ea44dc9ed014	1370	floating-point (single or double precision).
joeverbout	0:ea44dc9ed014	1371	@param points2 Array of the second image points of the same size and format as points1 .
joeverbout	0:ea44dc9ed014	1372	@param R Recovered relative rotation.
joeverbout	0:ea44dc9ed014	1373	@param t Recoverd relative translation.
joeverbout	0:ea44dc9ed014	1374	@param focal Focal length of the camera. Note that this function assumes that points1 and points2
joeverbout	0:ea44dc9ed014	1375	are feature points from cameras with same focal length and principle point.
joeverbout	0:ea44dc9ed014	1376	@param pp Principle point of the camera.
joeverbout	0:ea44dc9ed014	1377	@param mask Input/output mask for inliers in points1 and points2.
joeverbout	0:ea44dc9ed014	1378	: If it is not empty, then it marks inliers in points1 and points2 for then given essential
joeverbout	0:ea44dc9ed014	1379	matrix E. Only these inliers will be used to recover pose. In the output mask only inliers
joeverbout	0:ea44dc9ed014	1380	which pass the cheirality check.
joeverbout	0:ea44dc9ed014	1381
joeverbout	0:ea44dc9ed014	1382	This function differs from the one above that it computes camera matrix from focal length and
joeverbout	0:ea44dc9ed014	1383	principal point:
joeverbout	0:ea44dc9ed014	1384
joeverbout	0:ea44dc9ed014	1385	\f[K =
joeverbout	0:ea44dc9ed014	1386	\begin{bmatrix}
joeverbout	0:ea44dc9ed014	1387	f & 0 & x_{pp} \\
joeverbout	0:ea44dc9ed014	1388	0 & f & y_{pp} \\
joeverbout	0:ea44dc9ed014	1389	0 & 0 & 1
joeverbout	0:ea44dc9ed014	1390	\end{bmatrix}\f]
joeverbout	0:ea44dc9ed014	1391	*/
joeverbout	0:ea44dc9ed014	1392	CV_EXPORTS_W int recoverPose( InputArray E, InputArray points1, InputArray points2,
joeverbout	0:ea44dc9ed014	1393	OutputArray R, OutputArray t,
joeverbout	0:ea44dc9ed014	1394	double focal = 1.0, Point2d pp = Point2d(0, 0),
joeverbout	0:ea44dc9ed014	1395	InputOutputArray mask = noArray() );
joeverbout	0:ea44dc9ed014	1396
joeverbout	0:ea44dc9ed014	1397	/** @brief For points in an image of a stereo pair, computes the corresponding epilines in the other image.
joeverbout	0:ea44dc9ed014	1398
joeverbout	0:ea44dc9ed014	1399	@param points Input points. \f$N \times 1\f$ or \f$1 \times N\f$ matrix of type CV_32FC2 or
joeverbout	0:ea44dc9ed014	1400	vector\<Point2f\> .
joeverbout	0:ea44dc9ed014	1401	@param whichImage Index of the image (1 or 2) that contains the points .
joeverbout	0:ea44dc9ed014	1402	@param F Fundamental matrix that can be estimated using findFundamentalMat or stereoRectify .
joeverbout	0:ea44dc9ed014	1403	@param lines Output vector of the epipolar lines corresponding to the points in the other image.
joeverbout	0:ea44dc9ed014	1404	Each line \f$ax + by + c=0\f$ is encoded by 3 numbers \f$(a, b, c)\f$ .
joeverbout	0:ea44dc9ed014	1405
joeverbout	0:ea44dc9ed014	1406	For every point in one of the two images of a stereo pair, the function finds the equation of the
joeverbout	0:ea44dc9ed014	1407	corresponding epipolar line in the other image.
joeverbout	0:ea44dc9ed014	1408
joeverbout	0:ea44dc9ed014	1409	From the fundamental matrix definition (see findFundamentalMat ), line \f$l^{(2)}_i\f$ in the second
joeverbout	0:ea44dc9ed014	1410	image for the point \f$p^{(1)}_i\f$ in the first image (when whichImage=1 ) is computed as:
joeverbout	0:ea44dc9ed014	1411
joeverbout	0:ea44dc9ed014	1412	\f[l^{(2)}_i = F p^{(1)}_i\f]
joeverbout	0:ea44dc9ed014	1413
joeverbout	0:ea44dc9ed014	1414	And vice versa, when whichImage=2, \f$l^{(1)}_i\f$ is computed from \f$p^{(2)}_i\f$ as:
joeverbout	0:ea44dc9ed014	1415
joeverbout	0:ea44dc9ed014	1416	\f[l^{(1)}_i = F^T p^{(2)}_i\f]
joeverbout	0:ea44dc9ed014	1417
joeverbout	0:ea44dc9ed014	1418	Line coefficients are defined up to a scale. They are normalized so that \f$a_i^2+b_i^2=1\f$ .
joeverbout	0:ea44dc9ed014	1419	*/
joeverbout	0:ea44dc9ed014	1420	CV_EXPORTS_W void computeCorrespondEpilines( InputArray points, int whichImage,
joeverbout	0:ea44dc9ed014	1421	InputArray F, OutputArray lines );
joeverbout	0:ea44dc9ed014	1422
joeverbout	0:ea44dc9ed014	1423	/** @brief Reconstructs points by triangulation.
joeverbout	0:ea44dc9ed014	1424
joeverbout	0:ea44dc9ed014	1425	@param projMatr1 3x4 projection matrix of the first camera.
joeverbout	0:ea44dc9ed014	1426	@param projMatr2 3x4 projection matrix of the second camera.
joeverbout	0:ea44dc9ed014	1427	@param projPoints1 2xN array of feature points in the first image. In case of c++ version it can
joeverbout	0:ea44dc9ed014	1428	be also a vector of feature points or two-channel matrix of size 1xN or Nx1.
joeverbout	0:ea44dc9ed014	1429	@param projPoints2 2xN array of corresponding points in the second image. In case of c++ version
joeverbout	0:ea44dc9ed014	1430	it can be also a vector of feature points or two-channel matrix of size 1xN or Nx1.
joeverbout	0:ea44dc9ed014	1431	@param points4D 4xN array of reconstructed points in homogeneous coordinates.
joeverbout	0:ea44dc9ed014	1432
joeverbout	0:ea44dc9ed014	1433	The function reconstructs 3-dimensional points (in homogeneous coordinates) by using their
joeverbout	0:ea44dc9ed014	1434	observations with a stereo camera. Projections matrices can be obtained from stereoRectify.
joeverbout	0:ea44dc9ed014	1435
joeverbout	0:ea44dc9ed014	1436	@note
joeverbout	0:ea44dc9ed014	1437	Keep in mind that all input data should be of float type in order for this function to work.
joeverbout	0:ea44dc9ed014	1438
joeverbout	0:ea44dc9ed014	1439	@sa
joeverbout	0:ea44dc9ed014	1440	reprojectImageTo3D
joeverbout	0:ea44dc9ed014	1441	*/
joeverbout	0:ea44dc9ed014	1442	CV_EXPORTS_W void triangulatePoints( InputArray projMatr1, InputArray projMatr2,
joeverbout	0:ea44dc9ed014	1443	InputArray projPoints1, InputArray projPoints2,
joeverbout	0:ea44dc9ed014	1444	OutputArray points4D );
joeverbout	0:ea44dc9ed014	1445
joeverbout	0:ea44dc9ed014	1446	/** @brief Refines coordinates of corresponding points.
joeverbout	0:ea44dc9ed014	1447
joeverbout	0:ea44dc9ed014	1448	@param F 3x3 fundamental matrix.
joeverbout	0:ea44dc9ed014	1449	@param points1 1xN array containing the first set of points.
joeverbout	0:ea44dc9ed014	1450	@param points2 1xN array containing the second set of points.
joeverbout	0:ea44dc9ed014	1451	@param newPoints1 The optimized points1.
joeverbout	0:ea44dc9ed014	1452	@param newPoints2 The optimized points2.
joeverbout	0:ea44dc9ed014	1453
joeverbout	0:ea44dc9ed014	1454	The function implements the Optimal Triangulation Method (see Multiple View Geometry for details).
joeverbout	0:ea44dc9ed014	1455	For each given point correspondence points1[i] \<-\> points2[i], and a fundamental matrix F, it
joeverbout	0:ea44dc9ed014	1456	computes the corrected correspondences newPoints1[i] \<-\> newPoints2[i] that minimize the geometric
joeverbout	0:ea44dc9ed014	1457	error \f$d(points1[i], newPoints1[i])^2 + d(points2[i],newPoints2[i])^2\f$ (where \f$d(a,b)\f$ is the
joeverbout	0:ea44dc9ed014	1458	geometric distance between points \f$a\f$ and \f$b\f$ ) subject to the epipolar constraint
joeverbout	0:ea44dc9ed014	1459	\f$newPoints2^T * F * newPoints1 = 0\f$ .
joeverbout	0:ea44dc9ed014	1460	*/
joeverbout	0:ea44dc9ed014	1461	CV_EXPORTS_W void correctMatches( InputArray F, InputArray points1, InputArray points2,
joeverbout	0:ea44dc9ed014	1462	OutputArray newPoints1, OutputArray newPoints2 );
joeverbout	0:ea44dc9ed014	1463
joeverbout	0:ea44dc9ed014	1464	/** @brief Filters off small noise blobs (speckles) in the disparity map
joeverbout	0:ea44dc9ed014	1465
joeverbout	0:ea44dc9ed014	1466	@param img The input 16-bit signed disparity image
joeverbout	0:ea44dc9ed014	1467	@param newVal The disparity value used to paint-off the speckles
joeverbout	0:ea44dc9ed014	1468	@param maxSpeckleSize The maximum speckle size to consider it a speckle. Larger blobs are not
joeverbout	0:ea44dc9ed014	1469	affected by the algorithm
joeverbout	0:ea44dc9ed014	1470	@param maxDiff Maximum difference between neighbor disparity pixels to put them into the same
joeverbout	0:ea44dc9ed014	1471	blob. Note that since StereoBM, StereoSGBM and may be other algorithms return a fixed-point
joeverbout	0:ea44dc9ed014	1472	disparity map, where disparity values are multiplied by 16, this scale factor should be taken into
joeverbout	0:ea44dc9ed014	1473	account when specifying this parameter value.
joeverbout	0:ea44dc9ed014	1474	@param buf The optional temporary buffer to avoid memory allocation within the function.
joeverbout	0:ea44dc9ed014	1475	*/
joeverbout	0:ea44dc9ed014	1476	CV_EXPORTS_W void filterSpeckles( InputOutputArray img, double newVal,
joeverbout	0:ea44dc9ed014	1477	int maxSpeckleSize, double maxDiff,
joeverbout	0:ea44dc9ed014	1478	InputOutputArray buf = noArray() );
joeverbout	0:ea44dc9ed014	1479
joeverbout	0:ea44dc9ed014	1480	//! computes valid disparity ROI from the valid ROIs of the rectified images (that are returned by cv::stereoRectify())
joeverbout	0:ea44dc9ed014	1481	CV_EXPORTS_W Rect getValidDisparityROI( Rect roi1, Rect roi2,
joeverbout	0:ea44dc9ed014	1482	int minDisparity, int numberOfDisparities,
joeverbout	0:ea44dc9ed014	1483	int SADWindowSize );
joeverbout	0:ea44dc9ed014	1484
joeverbout	0:ea44dc9ed014	1485	//! validates disparity using the left-right check. The matrix "cost" should be computed by the stereo correspondence algorithm
joeverbout	0:ea44dc9ed014	1486	CV_EXPORTS_W void validateDisparity( InputOutputArray disparity, InputArray cost,
joeverbout	0:ea44dc9ed014	1487	int minDisparity, int numberOfDisparities,
joeverbout	0:ea44dc9ed014	1488	int disp12MaxDisp = 1 );
joeverbout	0:ea44dc9ed014	1489
joeverbout	0:ea44dc9ed014	1490	/** @brief Reprojects a disparity image to 3D space.
joeverbout	0:ea44dc9ed014	1491
joeverbout	0:ea44dc9ed014	1492	@param disparity Input single-channel 8-bit unsigned, 16-bit signed, 32-bit signed or 32-bit
joeverbout	0:ea44dc9ed014	1493	floating-point disparity image. If 16-bit signed format is used, the values are assumed to have no
joeverbout	0:ea44dc9ed014	1494	fractional bits.
joeverbout	0:ea44dc9ed014	1495	@param _3dImage Output 3-channel floating-point image of the same size as disparity . Each
joeverbout	0:ea44dc9ed014	1496	element of _3dImage(x,y) contains 3D coordinates of the point (x,y) computed from the disparity
joeverbout	0:ea44dc9ed014	1497	map.
joeverbout	0:ea44dc9ed014	1498	@param Q \f$4 \times 4\f$ perspective transformation matrix that can be obtained with stereoRectify.
joeverbout	0:ea44dc9ed014	1499	@param handleMissingValues Indicates, whether the function should handle missing values (i.e.
joeverbout	0:ea44dc9ed014	1500	points where the disparity was not computed). If handleMissingValues=true, then pixels with the
joeverbout	0:ea44dc9ed014	1501	minimal disparity that corresponds to the outliers (see StereoMatcher::compute ) are transformed
joeverbout	0:ea44dc9ed014	1502	to 3D points with a very large Z value (currently set to 10000).
joeverbout	0:ea44dc9ed014	1503	@param ddepth The optional output array depth. If it is -1, the output image will have CV_32F
joeverbout	0:ea44dc9ed014	1504	depth. ddepth can also be set to CV_16S, CV_32S or CV_32F.
joeverbout	0:ea44dc9ed014	1505
joeverbout	0:ea44dc9ed014	1506	The function transforms a single-channel disparity map to a 3-channel image representing a 3D
joeverbout	0:ea44dc9ed014	1507	surface. That is, for each pixel (x,y) andthe corresponding disparity d=disparity(x,y) , it
joeverbout	0:ea44dc9ed014	1508	computes:
joeverbout	0:ea44dc9ed014	1509
joeverbout	0:ea44dc9ed014	1510	\f[\begin{array}{l} [X \; Y \; Z \; W]^T = \texttt{Q} *[x \; y \; \texttt{disparity} (x,y) \; 1]^T \\ \texttt{\_3dImage} (x,y) = (X/W, \; Y/W, \; Z/W) \end{array}\f]
joeverbout	0:ea44dc9ed014	1511
joeverbout	0:ea44dc9ed014	1512	The matrix Q can be an arbitrary \f$4 \times 4\f$ matrix (for example, the one computed by
joeverbout	0:ea44dc9ed014	1513	stereoRectify). To reproject a sparse set of points {(x,y,d),...} to 3D space, use
joeverbout	0:ea44dc9ed014	1514	perspectiveTransform .
joeverbout	0:ea44dc9ed014	1515	*/
joeverbout	0:ea44dc9ed014	1516	CV_EXPORTS_W void reprojectImageTo3D( InputArray disparity,
joeverbout	0:ea44dc9ed014	1517	OutputArray _3dImage, InputArray Q,
joeverbout	0:ea44dc9ed014	1518	bool handleMissingValues = false,
joeverbout	0:ea44dc9ed014	1519	int ddepth = -1 );
joeverbout	0:ea44dc9ed014	1520
joeverbout	0:ea44dc9ed014	1521	/** @brief Calculates the Sampson Distance between two points.
joeverbout	0:ea44dc9ed014	1522
joeverbout	0:ea44dc9ed014	1523	The function sampsonDistance calculates and returns the first order approximation of the geometric error as:
joeverbout	0:ea44dc9ed014	1524	\f[sd( \texttt{pt1} , \texttt{pt2} )= \frac{(\texttt{pt2}^t \cdot \texttt{F} \cdot \texttt{pt1})^2}{(\texttt{F} \cdot \texttt{pt1})(0) + (\texttt{F} \cdot \texttt{pt1})(1) + (\texttt{F}^t \cdot \texttt{pt2})(0) + (\texttt{F}^t \cdot \texttt{pt2})(1)}\f]
joeverbout	0:ea44dc9ed014	1525	The fundamental matrix may be calculated using the cv::findFundamentalMat function. See HZ 11.4.3 for details.
joeverbout	0:ea44dc9ed014	1526	@param pt1 first homogeneous 2d point
joeverbout	0:ea44dc9ed014	1527	@param pt2 second homogeneous 2d point
joeverbout	0:ea44dc9ed014	1528	@param F fundamental matrix
joeverbout	0:ea44dc9ed014	1529	*/
joeverbout	0:ea44dc9ed014	1530	CV_EXPORTS_W double sampsonDistance(InputArray pt1, InputArray pt2, InputArray F);
joeverbout	0:ea44dc9ed014	1531
joeverbout	0:ea44dc9ed014	1532	/** @brief Computes an optimal affine transformation between two 3D point sets.
joeverbout	0:ea44dc9ed014	1533
joeverbout	0:ea44dc9ed014	1534	@param src First input 3D point set.
joeverbout	0:ea44dc9ed014	1535	@param dst Second input 3D point set.
joeverbout	0:ea44dc9ed014	1536	@param out Output 3D affine transformation matrix \f$3 \times 4\f$ .
joeverbout	0:ea44dc9ed014	1537	@param inliers Output vector indicating which points are inliers.
joeverbout	0:ea44dc9ed014	1538	@param ransacThreshold Maximum reprojection error in the RANSAC algorithm to consider a point as
joeverbout	0:ea44dc9ed014	1539	an inlier.
joeverbout	0:ea44dc9ed014	1540	@param confidence Confidence level, between 0 and 1, for the estimated transformation. Anything
joeverbout	0:ea44dc9ed014	1541	between 0.95 and 0.99 is usually good enough. Values too close to 1 can slow down the estimation
joeverbout	0:ea44dc9ed014	1542	significantly. Values lower than 0.8-0.9 can result in an incorrectly estimated transformation.
joeverbout	0:ea44dc9ed014	1543
joeverbout	0:ea44dc9ed014	1544	The function estimates an optimal 3D affine transformation between two 3D point sets using the
joeverbout	0:ea44dc9ed014	1545	RANSAC algorithm.
joeverbout	0:ea44dc9ed014	1546	*/
joeverbout	0:ea44dc9ed014	1547	CV_EXPORTS_W int estimateAffine3D(InputArray src, InputArray dst,
joeverbout	0:ea44dc9ed014	1548	OutputArray out, OutputArray inliers,
joeverbout	0:ea44dc9ed014	1549	double ransacThreshold = 3, double confidence = 0.99);
joeverbout	0:ea44dc9ed014	1550
joeverbout	0:ea44dc9ed014	1551	/** @brief Decompose a homography matrix to rotation(s), translation(s) and plane normal(s).
joeverbout	0:ea44dc9ed014	1552
joeverbout	0:ea44dc9ed014	1553	@param H The input homography matrix between two images.
joeverbout	0:ea44dc9ed014	1554	@param K The input intrinsic camera calibration matrix.
joeverbout	0:ea44dc9ed014	1555	@param rotations Array of rotation matrices.
joeverbout	0:ea44dc9ed014	1556	@param translations Array of translation matrices.
joeverbout	0:ea44dc9ed014	1557	@param normals Array of plane normal matrices.
joeverbout	0:ea44dc9ed014	1558
joeverbout	0:ea44dc9ed014	1559	This function extracts relative camera motion between two views observing a planar object from the
joeverbout	0:ea44dc9ed014	1560	homography H induced by the plane. The intrinsic camera matrix K must also be provided. The function
joeverbout	0:ea44dc9ed014	1561	may return up to four mathematical solution sets. At least two of the solutions may further be
joeverbout	0:ea44dc9ed014	1562	invalidated if point correspondences are available by applying positive depth constraint (all points
joeverbout	0:ea44dc9ed014	1563	must be in front of the camera). The decomposition method is described in detail in @cite Malis .
joeverbout	0:ea44dc9ed014	1564	*/
joeverbout	0:ea44dc9ed014	1565	CV_EXPORTS_W int decomposeHomographyMat(InputArray H,
joeverbout	0:ea44dc9ed014	1566	InputArray K,
joeverbout	0:ea44dc9ed014	1567	OutputArrayOfArrays rotations,
joeverbout	0:ea44dc9ed014	1568	OutputArrayOfArrays translations,
joeverbout	0:ea44dc9ed014	1569	OutputArrayOfArrays normals);
joeverbout	0:ea44dc9ed014	1570
joeverbout	0:ea44dc9ed014	1571	/** @brief The base class for stereo correspondence algorithms.
joeverbout	0:ea44dc9ed014	1572	*/
joeverbout	0:ea44dc9ed014	1573	class CV_EXPORTS_W StereoMatcher : public Algorithm
joeverbout	0:ea44dc9ed014	1574	{
joeverbout	0:ea44dc9ed014	1575	public:
joeverbout	0:ea44dc9ed014	1576	enum { DISP_SHIFT = 4,
joeverbout	0:ea44dc9ed014	1577	DISP_SCALE = (1 << DISP_SHIFT)
joeverbout	0:ea44dc9ed014	1578	};
joeverbout	0:ea44dc9ed014	1579
joeverbout	0:ea44dc9ed014	1580	/** @brief Computes disparity map for the specified stereo pair
joeverbout	0:ea44dc9ed014	1581
joeverbout	0:ea44dc9ed014	1582	@param left Left 8-bit single-channel image.
joeverbout	0:ea44dc9ed014	1583	@param right Right image of the same size and the same type as the left one.
joeverbout	0:ea44dc9ed014	1584	@param disparity Output disparity map. It has the same size as the input images. Some algorithms,
joeverbout	0:ea44dc9ed014	1585	like StereoBM or StereoSGBM compute 16-bit fixed-point disparity map (where each disparity value
joeverbout	0:ea44dc9ed014	1586	has 4 fractional bits), whereas other algorithms output 32-bit floating-point disparity map.
joeverbout	0:ea44dc9ed014	1587	*/
joeverbout	0:ea44dc9ed014	1588	CV_WRAP virtual void compute( InputArray left, InputArray right,
joeverbout	0:ea44dc9ed014	1589	OutputArray disparity ) = 0;
joeverbout	0:ea44dc9ed014	1590
joeverbout	0:ea44dc9ed014	1591	CV_WRAP virtual int getMinDisparity() const = 0;
joeverbout	0:ea44dc9ed014	1592	CV_WRAP virtual void setMinDisparity(int minDisparity) = 0;
joeverbout	0:ea44dc9ed014	1593
joeverbout	0:ea44dc9ed014	1594	CV_WRAP virtual int getNumDisparities() const = 0;
joeverbout	0:ea44dc9ed014	1595	CV_WRAP virtual void setNumDisparities(int numDisparities) = 0;
joeverbout	0:ea44dc9ed014	1596
joeverbout	0:ea44dc9ed014	1597	CV_WRAP virtual int getBlockSize() const = 0;
joeverbout	0:ea44dc9ed014	1598	CV_WRAP virtual void setBlockSize(int blockSize) = 0;
joeverbout	0:ea44dc9ed014	1599
joeverbout	0:ea44dc9ed014	1600	CV_WRAP virtual int getSpeckleWindowSize() const = 0;
joeverbout	0:ea44dc9ed014	1601	CV_WRAP virtual void setSpeckleWindowSize(int speckleWindowSize) = 0;
joeverbout	0:ea44dc9ed014	1602
joeverbout	0:ea44dc9ed014	1603	CV_WRAP virtual int getSpeckleRange() const = 0;
joeverbout	0:ea44dc9ed014	1604	CV_WRAP virtual void setSpeckleRange(int speckleRange) = 0;
joeverbout	0:ea44dc9ed014	1605
joeverbout	0:ea44dc9ed014	1606	CV_WRAP virtual int getDisp12MaxDiff() const = 0;
joeverbout	0:ea44dc9ed014	1607	CV_WRAP virtual void setDisp12MaxDiff(int disp12MaxDiff) = 0;
joeverbout	0:ea44dc9ed014	1608	};
joeverbout	0:ea44dc9ed014	1609
joeverbout	0:ea44dc9ed014	1610
joeverbout	0:ea44dc9ed014	1611	/** @brief Class for computing stereo correspondence using the block matching algorithm, introduced and
joeverbout	0:ea44dc9ed014	1612	contributed to OpenCV by K. Konolige.
joeverbout	0:ea44dc9ed014	1613	*/
joeverbout	0:ea44dc9ed014	1614	class CV_EXPORTS_W StereoBM : public StereoMatcher
joeverbout	0:ea44dc9ed014	1615	{
joeverbout	0:ea44dc9ed014	1616	public:
joeverbout	0:ea44dc9ed014	1617	enum { PREFILTER_NORMALIZED_RESPONSE = 0,
joeverbout	0:ea44dc9ed014	1618	PREFILTER_XSOBEL = 1
joeverbout	0:ea44dc9ed014	1619	};
joeverbout	0:ea44dc9ed014	1620
joeverbout	0:ea44dc9ed014	1621	CV_WRAP virtual int getPreFilterType() const = 0;
joeverbout	0:ea44dc9ed014	1622	CV_WRAP virtual void setPreFilterType(int preFilterType) = 0;
joeverbout	0:ea44dc9ed014	1623
joeverbout	0:ea44dc9ed014	1624	CV_WRAP virtual int getPreFilterSize() const = 0;
joeverbout	0:ea44dc9ed014	1625	CV_WRAP virtual void setPreFilterSize(int preFilterSize) = 0;
joeverbout	0:ea44dc9ed014	1626
joeverbout	0:ea44dc9ed014	1627	CV_WRAP virtual int getPreFilterCap() const = 0;
joeverbout	0:ea44dc9ed014	1628	CV_WRAP virtual void setPreFilterCap(int preFilterCap) = 0;
joeverbout	0:ea44dc9ed014	1629
joeverbout	0:ea44dc9ed014	1630	CV_WRAP virtual int getTextureThreshold() const = 0;
joeverbout	0:ea44dc9ed014	1631	CV_WRAP virtual void setTextureThreshold(int textureThreshold) = 0;
joeverbout	0:ea44dc9ed014	1632
joeverbout	0:ea44dc9ed014	1633	CV_WRAP virtual int getUniquenessRatio() const = 0;
joeverbout	0:ea44dc9ed014	1634	CV_WRAP virtual void setUniquenessRatio(int uniquenessRatio) = 0;
joeverbout	0:ea44dc9ed014	1635
joeverbout	0:ea44dc9ed014	1636	CV_WRAP virtual int getSmallerBlockSize() const = 0;
joeverbout	0:ea44dc9ed014	1637	CV_WRAP virtual void setSmallerBlockSize(int blockSize) = 0;
joeverbout	0:ea44dc9ed014	1638
joeverbout	0:ea44dc9ed014	1639	CV_WRAP virtual Rect getROI1() const = 0;
joeverbout	0:ea44dc9ed014	1640	CV_WRAP virtual void setROI1(Rect roi1) = 0;
joeverbout	0:ea44dc9ed014	1641
joeverbout	0:ea44dc9ed014	1642	CV_WRAP virtual Rect getROI2() const = 0;
joeverbout	0:ea44dc9ed014	1643	CV_WRAP virtual void setROI2(Rect roi2) = 0;
joeverbout	0:ea44dc9ed014	1644
joeverbout	0:ea44dc9ed014	1645	/** @brief Creates StereoBM object
joeverbout	0:ea44dc9ed014	1646
joeverbout	0:ea44dc9ed014	1647	@param numDisparities the disparity search range. For each pixel algorithm will find the best
joeverbout	0:ea44dc9ed014	1648	disparity from 0 (default minimum disparity) to numDisparities. The search range can then be
joeverbout	0:ea44dc9ed014	1649	shifted by changing the minimum disparity.
joeverbout	0:ea44dc9ed014	1650	@param blockSize the linear size of the blocks compared by the algorithm. The size should be odd
joeverbout	0:ea44dc9ed014	1651	(as the block is centered at the current pixel). Larger block size implies smoother, though less
joeverbout	0:ea44dc9ed014	1652	accurate disparity map. Smaller block size gives more detailed disparity map, but there is higher
joeverbout	0:ea44dc9ed014	1653	chance for algorithm to find a wrong correspondence.
joeverbout	0:ea44dc9ed014	1654
joeverbout	0:ea44dc9ed014	1655	The function create StereoBM object. You can then call StereoBM::compute() to compute disparity for
joeverbout	0:ea44dc9ed014	1656	a specific stereo pair.
joeverbout	0:ea44dc9ed014	1657	*/
joeverbout	0:ea44dc9ed014	1658	CV_WRAP static Ptr<StereoBM> create(int numDisparities = 0, int blockSize = 21);
joeverbout	0:ea44dc9ed014	1659	};
joeverbout	0:ea44dc9ed014	1660
joeverbout	0:ea44dc9ed014	1661	/** @brief The class implements the modified H. Hirschmuller algorithm @cite HH08 that differs from the original
joeverbout	0:ea44dc9ed014	1662	one as follows:
joeverbout	0:ea44dc9ed014	1663
joeverbout	0:ea44dc9ed014	1664	- By default, the algorithm is single-pass, which means that you consider only 5 directions
joeverbout	0:ea44dc9ed014	1665	instead of 8. Set mode=StereoSGBM::MODE_HH in createStereoSGBM to run the full variant of the
joeverbout	0:ea44dc9ed014	1666	algorithm but beware that it may consume a lot of memory.
joeverbout	0:ea44dc9ed014	1667	- The algorithm matches blocks, not individual pixels. Though, setting blockSize=1 reduces the
joeverbout	0:ea44dc9ed014	1668	blocks to single pixels.
joeverbout	0:ea44dc9ed014	1669	- Mutual information cost function is not implemented. Instead, a simpler Birchfield-Tomasi
joeverbout	0:ea44dc9ed014	1670	sub-pixel metric from @cite BT98 is used. Though, the color images are supported as well.
joeverbout	0:ea44dc9ed014	1671	- Some pre- and post- processing steps from K. Konolige algorithm StereoBM are included, for
joeverbout	0:ea44dc9ed014	1672	example: pre-filtering (StereoBM::PREFILTER_XSOBEL type) and post-filtering (uniqueness
joeverbout	0:ea44dc9ed014	1673	check, quadratic interpolation and speckle filtering).
joeverbout	0:ea44dc9ed014	1674
joeverbout	0:ea44dc9ed014	1675	@note
joeverbout	0:ea44dc9ed014	1676	- (Python) An example illustrating the use of the StereoSGBM matching algorithm can be found
joeverbout	0:ea44dc9ed014	1677	at opencv_source_code/samples/python/stereo_match.py
joeverbout	0:ea44dc9ed014	1678	*/
joeverbout	0:ea44dc9ed014	1679	class CV_EXPORTS_W StereoSGBM : public StereoMatcher
joeverbout	0:ea44dc9ed014	1680	{
joeverbout	0:ea44dc9ed014	1681	public:
joeverbout	0:ea44dc9ed014	1682	enum
joeverbout	0:ea44dc9ed014	1683	{
joeverbout	0:ea44dc9ed014	1684	MODE_SGBM = 0,
joeverbout	0:ea44dc9ed014	1685	MODE_HH = 1,
joeverbout	0:ea44dc9ed014	1686	MODE_SGBM_3WAY = 2
joeverbout	0:ea44dc9ed014	1687	};
joeverbout	0:ea44dc9ed014	1688
joeverbout	0:ea44dc9ed014	1689	CV_WRAP virtual int getPreFilterCap() const = 0;
joeverbout	0:ea44dc9ed014	1690	CV_WRAP virtual void setPreFilterCap(int preFilterCap) = 0;
joeverbout	0:ea44dc9ed014	1691
joeverbout	0:ea44dc9ed014	1692	CV_WRAP virtual int getUniquenessRatio() const = 0;
joeverbout	0:ea44dc9ed014	1693	CV_WRAP virtual void setUniquenessRatio(int uniquenessRatio) = 0;
joeverbout	0:ea44dc9ed014	1694
joeverbout	0:ea44dc9ed014	1695	CV_WRAP virtual int getP1() const = 0;
joeverbout	0:ea44dc9ed014	1696	CV_WRAP virtual void setP1(int P1) = 0;
joeverbout	0:ea44dc9ed014	1697
joeverbout	0:ea44dc9ed014	1698	CV_WRAP virtual int getP2() const = 0;
joeverbout	0:ea44dc9ed014	1699	CV_WRAP virtual void setP2(int P2) = 0;
joeverbout	0:ea44dc9ed014	1700
joeverbout	0:ea44dc9ed014	1701	CV_WRAP virtual int getMode() const = 0;
joeverbout	0:ea44dc9ed014	1702	CV_WRAP virtual void setMode(int mode) = 0;
joeverbout	0:ea44dc9ed014	1703
joeverbout	0:ea44dc9ed014	1704	/** @brief Creates StereoSGBM object
joeverbout	0:ea44dc9ed014	1705
joeverbout	0:ea44dc9ed014	1706	@param minDisparity Minimum possible disparity value. Normally, it is zero but sometimes
joeverbout	0:ea44dc9ed014	1707	rectification algorithms can shift images, so this parameter needs to be adjusted accordingly.
joeverbout	0:ea44dc9ed014	1708	@param numDisparities Maximum disparity minus minimum disparity. The value is always greater than
joeverbout	0:ea44dc9ed014	1709	zero. In the current implementation, this parameter must be divisible by 16.
joeverbout	0:ea44dc9ed014	1710	@param blockSize Matched block size. It must be an odd number \>=1 . Normally, it should be
joeverbout	0:ea44dc9ed014	1711	somewhere in the 3..11 range.
joeverbout	0:ea44dc9ed014	1712	@param P1 The first parameter controlling the disparity smoothness. See below.
joeverbout	0:ea44dc9ed014	1713	@param P2 The second parameter controlling the disparity smoothness. The larger the values are,
joeverbout	0:ea44dc9ed014	1714	the smoother the disparity is. P1 is the penalty on the disparity change by plus or minus 1
joeverbout	0:ea44dc9ed014	1715	between neighbor pixels. P2 is the penalty on the disparity change by more than 1 between neighbor
joeverbout	0:ea44dc9ed014	1716	pixels. The algorithm requires P2 \> P1 . See stereo_match.cpp sample where some reasonably good
joeverbout	0:ea44dc9ed014	1717	P1 and P2 values are shown (like 8\number_of_image_channels\SADWindowSize\*SADWindowSize and
joeverbout	0:ea44dc9ed014	1718	32\number_of_image_channels\SADWindowSize\*SADWindowSize , respectively).
joeverbout	0:ea44dc9ed014	1719	@param disp12MaxDiff Maximum allowed difference (in integer pixel units) in the left-right
joeverbout	0:ea44dc9ed014	1720	disparity check. Set it to a non-positive value to disable the check.
joeverbout	0:ea44dc9ed014	1721	@param preFilterCap Truncation value for the prefiltered image pixels. The algorithm first
joeverbout	0:ea44dc9ed014	1722	computes x-derivative at each pixel and clips its value by [-preFilterCap, preFilterCap] interval.
joeverbout	0:ea44dc9ed014	1723	The result values are passed to the Birchfield-Tomasi pixel cost function.
joeverbout	0:ea44dc9ed014	1724	@param uniquenessRatio Margin in percentage by which the best (minimum) computed cost function
joeverbout	0:ea44dc9ed014	1725	value should "win" the second best value to consider the found match correct. Normally, a value
joeverbout	0:ea44dc9ed014	1726	within the 5-15 range is good enough.
joeverbout	0:ea44dc9ed014	1727	@param speckleWindowSize Maximum size of smooth disparity regions to consider their noise speckles
joeverbout	0:ea44dc9ed014	1728	and invalidate. Set it to 0 to disable speckle filtering. Otherwise, set it somewhere in the
joeverbout	0:ea44dc9ed014	1729	50-200 range.
joeverbout	0:ea44dc9ed014	1730	@param speckleRange Maximum disparity variation within each connected component. If you do speckle
joeverbout	0:ea44dc9ed014	1731	filtering, set the parameter to a positive value, it will be implicitly multiplied by 16.
joeverbout	0:ea44dc9ed014	1732	Normally, 1 or 2 is good enough.
joeverbout	0:ea44dc9ed014	1733	@param mode Set it to StereoSGBM::MODE_HH to run the full-scale two-pass dynamic programming
joeverbout	0:ea44dc9ed014	1734	algorithm. It will consume O(W\H\numDisparities) bytes, which is large for 640x480 stereo and
joeverbout	0:ea44dc9ed014	1735	huge for HD-size pictures. By default, it is set to false .
joeverbout	0:ea44dc9ed014	1736
joeverbout	0:ea44dc9ed014	1737	The first constructor initializes StereoSGBM with all the default parameters. So, you only have to
joeverbout	0:ea44dc9ed014	1738	set StereoSGBM::numDisparities at minimum. The second constructor enables you to set each parameter
joeverbout	0:ea44dc9ed014	1739	to a custom value.
joeverbout	0:ea44dc9ed014	1740	*/
joeverbout	0:ea44dc9ed014	1741	CV_WRAP static Ptr<StereoSGBM> create(int minDisparity, int numDisparities, int blockSize,
joeverbout	0:ea44dc9ed014	1742	int P1 = 0, int P2 = 0, int disp12MaxDiff = 0,
joeverbout	0:ea44dc9ed014	1743	int preFilterCap = 0, int uniquenessRatio = 0,
joeverbout	0:ea44dc9ed014	1744	int speckleWindowSize = 0, int speckleRange = 0,
joeverbout	0:ea44dc9ed014	1745	int mode = StereoSGBM::MODE_SGBM);
joeverbout	0:ea44dc9ed014	1746	};
joeverbout	0:ea44dc9ed014	1747
joeverbout	0:ea44dc9ed014	1748	//! @} calib3d
joeverbout	0:ea44dc9ed014	1749
joeverbout	0:ea44dc9ed014	1750	/** @brief The methods in this namespace use a so-called fisheye camera model.
joeverbout	0:ea44dc9ed014	1751	@ingroup calib3d_fisheye
joeverbout	0:ea44dc9ed014	1752	*/
joeverbout	0:ea44dc9ed014	1753	namespace fisheye
joeverbout	0:ea44dc9ed014	1754	{
joeverbout	0:ea44dc9ed014	1755	//! @addtogroup calib3d_fisheye
joeverbout	0:ea44dc9ed014	1756	//! @{
joeverbout	0:ea44dc9ed014	1757
joeverbout	0:ea44dc9ed014	1758	enum{
joeverbout	0:ea44dc9ed014	1759	CALIB_USE_INTRINSIC_GUESS = 1,
joeverbout	0:ea44dc9ed014	1760	CALIB_RECOMPUTE_EXTRINSIC = 2,
joeverbout	0:ea44dc9ed014	1761	CALIB_CHECK_COND = 4,
joeverbout	0:ea44dc9ed014	1762	CALIB_FIX_SKEW = 8,
joeverbout	0:ea44dc9ed014	1763	CALIB_FIX_K1 = 16,
joeverbout	0:ea44dc9ed014	1764	CALIB_FIX_K2 = 32,
joeverbout	0:ea44dc9ed014	1765	CALIB_FIX_K3 = 64,
joeverbout	0:ea44dc9ed014	1766	CALIB_FIX_K4 = 128,
joeverbout	0:ea44dc9ed014	1767	CALIB_FIX_INTRINSIC = 256
joeverbout	0:ea44dc9ed014	1768	};
joeverbout	0:ea44dc9ed014	1769
joeverbout	0:ea44dc9ed014	1770	/** @brief Projects points using fisheye model
joeverbout	0:ea44dc9ed014	1771
joeverbout	0:ea44dc9ed014	1772	@param objectPoints Array of object points, 1xN/Nx1 3-channel (or vector\<Point3f\> ), where N is
joeverbout	0:ea44dc9ed014	1773	the number of points in the view.
joeverbout	0:ea44dc9ed014	1774	@param imagePoints Output array of image points, 2xN/Nx2 1-channel or 1xN/Nx1 2-channel, or
joeverbout	0:ea44dc9ed014	1775	vector\<Point2f\>.
joeverbout	0:ea44dc9ed014	1776	@param affine
joeverbout	0:ea44dc9ed014	1777	@param K Camera matrix \f$K = \vecthreethree{f_x}{0}{c_x}{0}{f_y}{c_y}{0}{0}{_1}\f$.
joeverbout	0:ea44dc9ed014	1778	@param D Input vector of distortion coefficients \f$(k_1, k_2, k_3, k_4)\f$.
joeverbout	0:ea44dc9ed014	1779	@param alpha The skew coefficient.
joeverbout	0:ea44dc9ed014	1780	@param jacobian Optional output 2Nx15 jacobian matrix of derivatives of image points with respect
joeverbout	0:ea44dc9ed014	1781	to components of the focal lengths, coordinates of the principal point, distortion coefficients,
joeverbout	0:ea44dc9ed014	1782	rotation vector, translation vector, and the skew. In the old interface different components of
joeverbout	0:ea44dc9ed014	1783	the jacobian are returned via different output parameters.
joeverbout	0:ea44dc9ed014	1784
joeverbout	0:ea44dc9ed014	1785	The function computes projections of 3D points to the image plane given intrinsic and extrinsic
joeverbout	0:ea44dc9ed014	1786	camera parameters. Optionally, the function computes Jacobians - matrices of partial derivatives of
joeverbout	0:ea44dc9ed014	1787	image points coordinates (as functions of all the input parameters) with respect to the particular
joeverbout	0:ea44dc9ed014	1788	parameters, intrinsic and/or extrinsic.
joeverbout	0:ea44dc9ed014	1789	*/
joeverbout	0:ea44dc9ed014	1790	CV_EXPORTS void projectPoints(InputArray objectPoints, OutputArray imagePoints, const Affine3d& affine,
joeverbout	0:ea44dc9ed014	1791	InputArray K, InputArray D, double alpha = 0, OutputArray jacobian = noArray());
joeverbout	0:ea44dc9ed014	1792
joeverbout	0:ea44dc9ed014	1793	/** @overload */
joeverbout	0:ea44dc9ed014	1794	CV_EXPORTS_W void projectPoints(InputArray objectPoints, OutputArray imagePoints, InputArray rvec, InputArray tvec,
joeverbout	0:ea44dc9ed014	1795	InputArray K, InputArray D, double alpha = 0, OutputArray jacobian = noArray());
joeverbout	0:ea44dc9ed014	1796
joeverbout	0:ea44dc9ed014	1797	/** @brief Distorts 2D points using fisheye model.
joeverbout	0:ea44dc9ed014	1798
joeverbout	0:ea44dc9ed014	1799	@param undistorted Array of object points, 1xN/Nx1 2-channel (or vector\<Point2f\> ), where N is
joeverbout	0:ea44dc9ed014	1800	the number of points in the view.
joeverbout	0:ea44dc9ed014	1801	@param K Camera matrix \f$K = \vecthreethree{f_x}{0}{c_x}{0}{f_y}{c_y}{0}{0}{_1}\f$.
joeverbout	0:ea44dc9ed014	1802	@param D Input vector of distortion coefficients \f$(k_1, k_2, k_3, k_4)\f$.
joeverbout	0:ea44dc9ed014	1803	@param alpha The skew coefficient.
joeverbout	0:ea44dc9ed014	1804	@param distorted Output array of image points, 1xN/Nx1 2-channel, or vector\<Point2f\> .
joeverbout	0:ea44dc9ed014	1805	*/
joeverbout	0:ea44dc9ed014	1806	CV_EXPORTS_W void distortPoints(InputArray undistorted, OutputArray distorted, InputArray K, InputArray D, double alpha = 0);
joeverbout	0:ea44dc9ed014	1807
joeverbout	0:ea44dc9ed014	1808	/** @brief Undistorts 2D points using fisheye model
joeverbout	0:ea44dc9ed014	1809
joeverbout	0:ea44dc9ed014	1810	@param distorted Array of object points, 1xN/Nx1 2-channel (or vector\<Point2f\> ), where N is the
joeverbout	0:ea44dc9ed014	1811	number of points in the view.
joeverbout	0:ea44dc9ed014	1812	@param K Camera matrix \f$K = \vecthreethree{f_x}{0}{c_x}{0}{f_y}{c_y}{0}{0}{_1}\f$.
joeverbout	0:ea44dc9ed014	1813	@param D Input vector of distortion coefficients \f$(k_1, k_2, k_3, k_4)\f$.
joeverbout	0:ea44dc9ed014	1814	@param R Rectification transformation in the object space: 3x3 1-channel, or vector: 3x1/1x3
joeverbout	0:ea44dc9ed014	1815	1-channel or 1x1 3-channel
joeverbout	0:ea44dc9ed014	1816	@param P New camera matrix (3x3) or new projection matrix (3x4)
joeverbout	0:ea44dc9ed014	1817	@param undistorted Output array of image points, 1xN/Nx1 2-channel, or vector\<Point2f\> .
joeverbout	0:ea44dc9ed014	1818	*/
joeverbout	0:ea44dc9ed014	1819	CV_EXPORTS_W void undistortPoints(InputArray distorted, OutputArray undistorted,
joeverbout	0:ea44dc9ed014	1820	InputArray K, InputArray D, InputArray R = noArray(), InputArray P = noArray());
joeverbout	0:ea44dc9ed014	1821
joeverbout	0:ea44dc9ed014	1822	/** @brief Computes undistortion and rectification maps for image transform by cv::remap(). If D is empty zero
joeverbout	0:ea44dc9ed014	1823	distortion is used, if R or P is empty identity matrixes are used.
joeverbout	0:ea44dc9ed014	1824
joeverbout	0:ea44dc9ed014	1825	@param K Camera matrix \f$K = \vecthreethree{f_x}{0}{c_x}{0}{f_y}{c_y}{0}{0}{_1}\f$.
joeverbout	0:ea44dc9ed014	1826	@param D Input vector of distortion coefficients \f$(k_1, k_2, k_3, k_4)\f$.
joeverbout	0:ea44dc9ed014	1827	@param R Rectification transformation in the object space: 3x3 1-channel, or vector: 3x1/1x3
joeverbout	0:ea44dc9ed014	1828	1-channel or 1x1 3-channel
joeverbout	0:ea44dc9ed014	1829	@param P New camera matrix (3x3) or new projection matrix (3x4)
joeverbout	0:ea44dc9ed014	1830	@param size Undistorted image size.
joeverbout	0:ea44dc9ed014	1831	@param m1type Type of the first output map that can be CV_32FC1 or CV_16SC2 . See convertMaps()
joeverbout	0:ea44dc9ed014	1832	for details.
joeverbout	0:ea44dc9ed014	1833	@param map1 The first output map.
joeverbout	0:ea44dc9ed014	1834	@param map2 The second output map.
joeverbout	0:ea44dc9ed014	1835	*/
joeverbout	0:ea44dc9ed014	1836	CV_EXPORTS_W void initUndistortRectifyMap(InputArray K, InputArray D, InputArray R, InputArray P,
joeverbout	0:ea44dc9ed014	1837	const cv::Size& size, int m1type, OutputArray map1, OutputArray map2);
joeverbout	0:ea44dc9ed014	1838
joeverbout	0:ea44dc9ed014	1839	/** @brief Transforms an image to compensate for fisheye lens distortion.
joeverbout	0:ea44dc9ed014	1840
joeverbout	0:ea44dc9ed014	1841	@param distorted image with fisheye lens distortion.
joeverbout	0:ea44dc9ed014	1842	@param undistorted Output image with compensated fisheye lens distortion.
joeverbout	0:ea44dc9ed014	1843	@param K Camera matrix \f$K = \vecthreethree{f_x}{0}{c_x}{0}{f_y}{c_y}{0}{0}{_1}\f$.
joeverbout	0:ea44dc9ed014	1844	@param D Input vector of distortion coefficients \f$(k_1, k_2, k_3, k_4)\f$.
joeverbout	0:ea44dc9ed014	1845	@param Knew Camera matrix of the distorted image. By default, it is the identity matrix but you
joeverbout	0:ea44dc9ed014	1846	may additionally scale and shift the result by using a different matrix.
joeverbout	0:ea44dc9ed014	1847	@param new_size
joeverbout	0:ea44dc9ed014	1848
joeverbout	0:ea44dc9ed014	1849	The function transforms an image to compensate radial and tangential lens distortion.
joeverbout	0:ea44dc9ed014	1850
joeverbout	0:ea44dc9ed014	1851	The function is simply a combination of fisheye::initUndistortRectifyMap (with unity R ) and remap
joeverbout	0:ea44dc9ed014	1852	(with bilinear interpolation). See the former function for details of the transformation being
joeverbout	0:ea44dc9ed014	1853	performed.
joeverbout	0:ea44dc9ed014	1854
joeverbout	0:ea44dc9ed014	1855	See below the results of undistortImage.
joeverbout	0:ea44dc9ed014	1856	- a\) result of undistort of perspective camera model (all possible coefficients (k_1, k_2, k_3,
joeverbout	0:ea44dc9ed014	1857	k_4, k_5, k_6) of distortion were optimized under calibration)
joeverbout	0:ea44dc9ed014	1858	- b\) result of fisheye::undistortImage of fisheye camera model (all possible coefficients (k_1, k_2,
joeverbout	0:ea44dc9ed014	1859	k_3, k_4) of fisheye distortion were optimized under calibration)
joeverbout	0:ea44dc9ed014	1860	- c\) original image was captured with fisheye lens
joeverbout	0:ea44dc9ed014	1861
joeverbout	0:ea44dc9ed014	1862	Pictures a) and b) almost the same. But if we consider points of image located far from the center
joeverbout	0:ea44dc9ed014	1863	of image, we can notice that on image a) these points are distorted.
joeverbout	0:ea44dc9ed014	1864
joeverbout	0:ea44dc9ed014	1865	![image](pics/fisheye_undistorted.jpg)
joeverbout	0:ea44dc9ed014	1866	*/
joeverbout	0:ea44dc9ed014	1867	CV_EXPORTS_W void undistortImage(InputArray distorted, OutputArray undistorted,
joeverbout	0:ea44dc9ed014	1868	InputArray K, InputArray D, InputArray Knew = cv::noArray(), const Size& new_size = Size());
joeverbout	0:ea44dc9ed014	1869
joeverbout	0:ea44dc9ed014	1870	/** @brief Estimates new camera matrix for undistortion or rectification.
joeverbout	0:ea44dc9ed014	1871
joeverbout	0:ea44dc9ed014	1872	@param K Camera matrix \f$K = \vecthreethree{f_x}{0}{c_x}{0}{f_y}{c_y}{0}{0}{_1}\f$.
joeverbout	0:ea44dc9ed014	1873	@param image_size
joeverbout	0:ea44dc9ed014	1874	@param D Input vector of distortion coefficients \f$(k_1, k_2, k_3, k_4)\f$.
joeverbout	0:ea44dc9ed014	1875	@param R Rectification transformation in the object space: 3x3 1-channel, or vector: 3x1/1x3
joeverbout	0:ea44dc9ed014	1876	1-channel or 1x1 3-channel
joeverbout	0:ea44dc9ed014	1877	@param P New camera matrix (3x3) or new projection matrix (3x4)
joeverbout	0:ea44dc9ed014	1878	@param balance Sets the new focal length in range between the min focal length and the max focal
joeverbout	0:ea44dc9ed014	1879	length. Balance is in range of [0, 1].
joeverbout	0:ea44dc9ed014	1880	@param new_size
joeverbout	0:ea44dc9ed014	1881	@param fov_scale Divisor for new focal length.
joeverbout	0:ea44dc9ed014	1882	*/
joeverbout	0:ea44dc9ed014	1883	CV_EXPORTS_W void estimateNewCameraMatrixForUndistortRectify(InputArray K, InputArray D, const Size &image_size, InputArray R,
joeverbout	0:ea44dc9ed014	1884	OutputArray P, double balance = 0.0, const Size& new_size = Size(), double fov_scale = 1.0);
joeverbout	0:ea44dc9ed014	1885
joeverbout	0:ea44dc9ed014	1886	/** @brief Performs camera calibaration
joeverbout	0:ea44dc9ed014	1887
joeverbout	0:ea44dc9ed014	1888	@param objectPoints vector of vectors of calibration pattern points in the calibration pattern
joeverbout	0:ea44dc9ed014	1889	coordinate space.
joeverbout	0:ea44dc9ed014	1890	@param imagePoints vector of vectors of the projections of calibration pattern points.
joeverbout	0:ea44dc9ed014	1891	imagePoints.size() and objectPoints.size() and imagePoints[i].size() must be equal to
joeverbout	0:ea44dc9ed014	1892	objectPoints[i].size() for each i.
joeverbout	0:ea44dc9ed014	1893	@param image_size Size of the image used only to initialize the intrinsic camera matrix.
joeverbout	0:ea44dc9ed014	1894	@param K Output 3x3 floating-point camera matrix
joeverbout	0:ea44dc9ed014	1895	\f$A = \vecthreethree{f_x}{0}{c_x}{0}{f_y}{c_y}{0}{0}{1}\f$ . If
joeverbout	0:ea44dc9ed014	1896	fisheye::CALIB_USE_INTRINSIC_GUESS/ is specified, some or all of fx, fy, cx, cy must be
joeverbout	0:ea44dc9ed014	1897	initialized before calling the function.
joeverbout	0:ea44dc9ed014	1898	@param D Output vector of distortion coefficients \f$(k_1, k_2, k_3, k_4)\f$.
joeverbout	0:ea44dc9ed014	1899	@param rvecs Output vector of rotation vectors (see Rodrigues ) estimated for each pattern view.
joeverbout	0:ea44dc9ed014	1900	That is, each k-th rotation vector together with the corresponding k-th translation vector (see
joeverbout	0:ea44dc9ed014	1901	the next output parameter description) brings the calibration pattern from the model coordinate
joeverbout	0:ea44dc9ed014	1902	space (in which object points are specified) to the world coordinate space, that is, a real
joeverbout	0:ea44dc9ed014	1903	position of the calibration pattern in the k-th pattern view (k=0.. M -1).
joeverbout	0:ea44dc9ed014	1904	@param tvecs Output vector of translation vectors estimated for each pattern view.
joeverbout	0:ea44dc9ed014	1905	@param flags Different flags that may be zero or a combination of the following values:
joeverbout	0:ea44dc9ed014	1906	- fisheye::CALIB_USE_INTRINSIC_GUESS cameraMatrix contains valid initial values of
joeverbout	0:ea44dc9ed014	1907	fx, fy, cx, cy that are optimized further. Otherwise, (cx, cy) is initially set to the image
joeverbout	0:ea44dc9ed014	1908	center ( imageSize is used), and focal distances are computed in a least-squares fashion.
joeverbout	0:ea44dc9ed014	1909	- fisheye::CALIB_RECOMPUTE_EXTRINSIC Extrinsic will be recomputed after each iteration
joeverbout	0:ea44dc9ed014	1910	of intrinsic optimization.
joeverbout	0:ea44dc9ed014	1911	- fisheye::CALIB_CHECK_COND The functions will check validity of condition number.
joeverbout	0:ea44dc9ed014	1912	- fisheye::CALIB_FIX_SKEW Skew coefficient (alpha) is set to zero and stay zero.
joeverbout	0:ea44dc9ed014	1913	- fisheye::CALIB_FIX_K1..4 Selected distortion coefficients are set to zeros and stay
joeverbout	0:ea44dc9ed014	1914	zero.
joeverbout	0:ea44dc9ed014	1915	@param criteria Termination criteria for the iterative optimization algorithm.
joeverbout	0:ea44dc9ed014	1916	*/
joeverbout	0:ea44dc9ed014	1917	CV_EXPORTS_W double calibrate(InputArrayOfArrays objectPoints, InputArrayOfArrays imagePoints, const Size& image_size,
joeverbout	0:ea44dc9ed014	1918	InputOutputArray K, InputOutputArray D, OutputArrayOfArrays rvecs, OutputArrayOfArrays tvecs, int flags = 0,
joeverbout	0:ea44dc9ed014	1919	TermCriteria criteria = TermCriteria(TermCriteria::COUNT + TermCriteria::EPS, 100, DBL_EPSILON));
joeverbout	0:ea44dc9ed014	1920
joeverbout	0:ea44dc9ed014	1921	/** @brief Stereo rectification for fisheye camera model
joeverbout	0:ea44dc9ed014	1922
joeverbout	0:ea44dc9ed014	1923	@param K1 First camera matrix.
joeverbout	0:ea44dc9ed014	1924	@param D1 First camera distortion parameters.
joeverbout	0:ea44dc9ed014	1925	@param K2 Second camera matrix.
joeverbout	0:ea44dc9ed014	1926	@param D2 Second camera distortion parameters.
joeverbout	0:ea44dc9ed014	1927	@param imageSize Size of the image used for stereo calibration.
joeverbout	0:ea44dc9ed014	1928	@param R Rotation matrix between the coordinate systems of the first and the second
joeverbout	0:ea44dc9ed014	1929	cameras.
joeverbout	0:ea44dc9ed014	1930	@param tvec Translation vector between coordinate systems of the cameras.
joeverbout	0:ea44dc9ed014	1931	@param R1 Output 3x3 rectification transform (rotation matrix) for the first camera.
joeverbout	0:ea44dc9ed014	1932	@param R2 Output 3x3 rectification transform (rotation matrix) for the second camera.
joeverbout	0:ea44dc9ed014	1933	@param P1 Output 3x4 projection matrix in the new (rectified) coordinate systems for the first
joeverbout	0:ea44dc9ed014	1934	camera.
joeverbout	0:ea44dc9ed014	1935	@param P2 Output 3x4 projection matrix in the new (rectified) coordinate systems for the second
joeverbout	0:ea44dc9ed014	1936	camera.
joeverbout	0:ea44dc9ed014	1937	@param Q Output \f$4 \times 4\f$ disparity-to-depth mapping matrix (see reprojectImageTo3D ).
joeverbout	0:ea44dc9ed014	1938	@param flags Operation flags that may be zero or CV_CALIB_ZERO_DISPARITY . If the flag is set,
joeverbout	0:ea44dc9ed014	1939	the function makes the principal points of each camera have the same pixel coordinates in the
joeverbout	0:ea44dc9ed014	1940	rectified views. And if the flag is not set, the function may still shift the images in the
joeverbout	0:ea44dc9ed014	1941	horizontal or vertical direction (depending on the orientation of epipolar lines) to maximize the
joeverbout	0:ea44dc9ed014	1942	useful image area.
joeverbout	0:ea44dc9ed014	1943	@param newImageSize New image resolution after rectification. The same size should be passed to
joeverbout	0:ea44dc9ed014	1944	initUndistortRectifyMap (see the stereo_calib.cpp sample in OpenCV samples directory). When (0,0)
joeverbout	0:ea44dc9ed014	1945	is passed (default), it is set to the original imageSize . Setting it to larger value can help you
joeverbout	0:ea44dc9ed014	1946	preserve details in the original image, especially when there is a big radial distortion.
joeverbout	0:ea44dc9ed014	1947	@param balance Sets the new focal length in range between the min focal length and the max focal
joeverbout	0:ea44dc9ed014	1948	length. Balance is in range of [0, 1].
joeverbout	0:ea44dc9ed014	1949	@param fov_scale Divisor for new focal length.
joeverbout	0:ea44dc9ed014	1950	*/
joeverbout	0:ea44dc9ed014	1951	CV_EXPORTS_W void stereoRectify(InputArray K1, InputArray D1, InputArray K2, InputArray D2, const Size &imageSize, InputArray R, InputArray tvec,
joeverbout	0:ea44dc9ed014	1952	OutputArray R1, OutputArray R2, OutputArray P1, OutputArray P2, OutputArray Q, int flags, const Size &newImageSize = Size(),
joeverbout	0:ea44dc9ed014	1953	double balance = 0.0, double fov_scale = 1.0);
joeverbout	0:ea44dc9ed014	1954
joeverbout	0:ea44dc9ed014	1955	/** @brief Performs stereo calibration
joeverbout	0:ea44dc9ed014	1956
joeverbout	0:ea44dc9ed014	1957	@param objectPoints Vector of vectors of the calibration pattern points.
joeverbout	0:ea44dc9ed014	1958	@param imagePoints1 Vector of vectors of the projections of the calibration pattern points,
joeverbout	0:ea44dc9ed014	1959	observed by the first camera.
joeverbout	0:ea44dc9ed014	1960	@param imagePoints2 Vector of vectors of the projections of the calibration pattern points,
joeverbout	0:ea44dc9ed014	1961	observed by the second camera.
joeverbout	0:ea44dc9ed014	1962	@param K1 Input/output first camera matrix:
joeverbout	0:ea44dc9ed014	1963	\f$\vecthreethree{f_x^{(j)}}{0}{c_x^{(j)}}{0}{f_y^{(j)}}{c_y^{(j)}}{0}{0}{1}\f$ , \f$j = 0,\, 1\f$ . If
joeverbout	0:ea44dc9ed014	1964	any of fisheye::CALIB_USE_INTRINSIC_GUESS , fisheye::CV_CALIB_FIX_INTRINSIC are specified,
joeverbout	0:ea44dc9ed014	1965	some or all of the matrix components must be initialized.
joeverbout	0:ea44dc9ed014	1966	@param D1 Input/output vector of distortion coefficients \f$(k_1, k_2, k_3, k_4)\f$ of 4 elements.
joeverbout	0:ea44dc9ed014	1967	@param K2 Input/output second camera matrix. The parameter is similar to K1 .
joeverbout	0:ea44dc9ed014	1968	@param D2 Input/output lens distortion coefficients for the second camera. The parameter is
joeverbout	0:ea44dc9ed014	1969	similar to D1 .
joeverbout	0:ea44dc9ed014	1970	@param imageSize Size of the image used only to initialize intrinsic camera matrix.
joeverbout	0:ea44dc9ed014	1971	@param R Output rotation matrix between the 1st and the 2nd camera coordinate systems.
joeverbout	0:ea44dc9ed014	1972	@param T Output translation vector between the coordinate systems of the cameras.
joeverbout	0:ea44dc9ed014	1973	@param flags Different flags that may be zero or a combination of the following values:
joeverbout	0:ea44dc9ed014	1974	- fisheye::CV_CALIB_FIX_INTRINSIC Fix K1, K2? and D1, D2? so that only R, T matrices
joeverbout	0:ea44dc9ed014	1975	are estimated.
joeverbout	0:ea44dc9ed014	1976	- fisheye::CALIB_USE_INTRINSIC_GUESS K1, K2 contains valid initial values of
joeverbout	0:ea44dc9ed014	1977	fx, fy, cx, cy that are optimized further. Otherwise, (cx, cy) is initially set to the image
joeverbout	0:ea44dc9ed014	1978	center (imageSize is used), and focal distances are computed in a least-squares fashion.
joeverbout	0:ea44dc9ed014	1979	- fisheye::CALIB_RECOMPUTE_EXTRINSIC Extrinsic will be recomputed after each iteration
joeverbout	0:ea44dc9ed014	1980	of intrinsic optimization.
joeverbout	0:ea44dc9ed014	1981	- fisheye::CALIB_CHECK_COND The functions will check validity of condition number.
joeverbout	0:ea44dc9ed014	1982	- fisheye::CALIB_FIX_SKEW Skew coefficient (alpha) is set to zero and stay zero.
joeverbout	0:ea44dc9ed014	1983	- fisheye::CALIB_FIX_K1..4 Selected distortion coefficients are set to zeros and stay
joeverbout	0:ea44dc9ed014	1984	zero.
joeverbout	0:ea44dc9ed014	1985	@param criteria Termination criteria for the iterative optimization algorithm.
joeverbout	0:ea44dc9ed014	1986	*/
joeverbout	0:ea44dc9ed014	1987	CV_EXPORTS_W double stereoCalibrate(InputArrayOfArrays objectPoints, InputArrayOfArrays imagePoints1, InputArrayOfArrays imagePoints2,
joeverbout	0:ea44dc9ed014	1988	InputOutputArray K1, InputOutputArray D1, InputOutputArray K2, InputOutputArray D2, Size imageSize,
joeverbout	0:ea44dc9ed014	1989	OutputArray R, OutputArray T, int flags = fisheye::CALIB_FIX_INTRINSIC,
joeverbout	0:ea44dc9ed014	1990	TermCriteria criteria = TermCriteria(TermCriteria::COUNT + TermCriteria::EPS, 100, DBL_EPSILON));
joeverbout	0:ea44dc9ed014	1991
joeverbout	0:ea44dc9ed014	1992	//! @} calib3d_fisheye
joeverbout	0:ea44dc9ed014	1993	}
joeverbout	0:ea44dc9ed014	1994
joeverbout	0:ea44dc9ed014	1995	} // cv
joeverbout	0:ea44dc9ed014	1996
joeverbout	0:ea44dc9ed014	1997	#ifndef DISABLE_OPENCV_24_COMPATIBILITY
joeverbout	0:ea44dc9ed014	1998	#include "opencv2/calib3d/calib3d_c.h"
joeverbout	0:ea44dc9ed014	1999	#endif
joeverbout	0:ea44dc9ed014	2000
joeverbout	0:ea44dc9ed014	2001	#endif
joeverbout	0:ea44dc9ed014	2002

Repository toolbox

Export to desktop IDE

Build repository

Repository details

Type:	Program
Mbed OS support:	Mbed 2 deprecated
Created:	31 Mar 2016
Imports:	152
Forks:	0
Commits:	1
Dependents:	0
Dependencies:	1
Followers:	2

opencv2/calib3d.hpp@0:ea44dc9ed014, 2016-03-31 (annotated)

Who changed what in which revision?

Repository toolbox

Repository details

Important Information for this Arm website

Access Warning