1 /*M/////////////////////////////////////////////////////////////////////////////////////// 2 // 3 // IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING. 4 // 5 // By downloading, copying, installing or using the software you agree to this license. 6 // If you do not agree to this license, do not download, install, 7 // copy or use the software. 8 // 9 // 10 // License Agreement 11 // For Open Source Computer Vision Library 12 // 13 // Copyright (C) 2000-2008, Intel Corporation, all rights reserved. 14 // Copyright (C) 2009, Willow Garage Inc., all rights reserved. 15 // Copyright (C) 2013, OpenCV Foundation, all rights reserved. 16 // Third party copyrights are property of their respective owners. 17 // 18 // Redistribution and use in source and binary forms, with or without modification, 19 // are permitted provided that the following conditions are met: 20 // 21 // * Redistribution's of source code must retain the above copyright notice, 22 // this list of conditions and the following disclaimer. 23 // 24 // * Redistribution's in binary form must reproduce the above copyright notice, 25 // this list of conditions and the following disclaimer in the documentation 26 // and/or other materials provided with the distribution. 27 // 28 // * The name of the copyright holders may not be used to endorse or promote products 29 // derived from this software without specific prior written permission. 30 // 31 // This software is provided by the copyright holders and contributors "as is" and 32 // any express or implied warranties, including, but not limited to, the implied 33 // warranties of merchantability and fitness for a particular purpose are disclaimed. 34 // In no event shall the Intel Corporation or contributors be liable for any direct, 35 // indirect, incidental, special, exemplary, or consequential damages 36 // (including, but not limited to, procurement of substitute goods or services; 37 // loss of use, data, or profits; or business interruption) however caused 38 // and on any theory of liability, whether in contract, strict liability, 39 // or tort (including negligence or otherwise) arising in any way out of 40 // the use of this software, even if advised of the possibility of such damage. 41 // 42 //M*/ 43 44 #ifndef __OPENCV_TRACKING_HPP__ 45 #define __OPENCV_TRACKING_HPP__ 46 47 #include "opencv2/core.hpp" 48 #include "opencv2/imgproc.hpp" 49 50 namespace cv 51 { 52 53 //! @addtogroup video_track 54 //! @{ 55 56 enum { OPTFLOW_USE_INITIAL_FLOW = 4, 57 OPTFLOW_LK_GET_MIN_EIGENVALS = 8, 58 OPTFLOW_FARNEBACK_GAUSSIAN = 256 59 }; 60 61 /** @brief Finds an object center, size, and orientation. 62 63 @param probImage Back projection of the object histogram. See calcBackProject. 64 @param window Initial search window. 65 @param criteria Stop criteria for the underlying meanShift. 66 returns 67 (in old interfaces) Number of iterations CAMSHIFT took to converge 68 The function implements the CAMSHIFT object tracking algorithm @cite Bradski98 . First, it finds an 69 object center using meanShift and then adjusts the window size and finds the optimal rotation. The 70 function returns the rotated rectangle structure that includes the object position, size, and 71 orientation. The next position of the search window can be obtained with RotatedRect::boundingRect() 72 73 See the OpenCV sample camshiftdemo.c that tracks colored objects. 74 75 @note 76 - (Python) A sample explaining the camshift tracking algorithm can be found at 77 opencv_source_code/samples/python2/camshift.py 78 */ 79 CV_EXPORTS_W RotatedRect CamShift( InputArray probImage, CV_IN_OUT Rect& window, 80 TermCriteria criteria ); 81 82 /** @brief Finds an object on a back projection image. 83 84 @param probImage Back projection of the object histogram. See calcBackProject for details. 85 @param window Initial search window. 86 @param criteria Stop criteria for the iterative search algorithm. 87 returns 88 : Number of iterations CAMSHIFT took to converge. 89 The function implements the iterative object search algorithm. It takes the input back projection of 90 an object and the initial position. The mass center in window of the back projection image is 91 computed and the search window center shifts to the mass center. The procedure is repeated until the 92 specified number of iterations criteria.maxCount is done or until the window center shifts by less 93 than criteria.epsilon. The algorithm is used inside CamShift and, unlike CamShift , the search 94 window size or orientation do not change during the search. You can simply pass the output of 95 calcBackProject to this function. But better results can be obtained if you pre-filter the back 96 projection and remove the noise. For example, you can do this by retrieving connected components 97 with findContours , throwing away contours with small area ( contourArea ), and rendering the 98 remaining contours with drawContours. 99 100 @note 101 - A mean-shift tracking sample can be found at opencv_source_code/samples/cpp/camshiftdemo.cpp 102 */ 103 CV_EXPORTS_W int meanShift( InputArray probImage, CV_IN_OUT Rect& window, TermCriteria criteria ); 104 105 /** @brief Constructs the image pyramid which can be passed to calcOpticalFlowPyrLK. 106 107 @param img 8-bit input image. 108 @param pyramid output pyramid. 109 @param winSize window size of optical flow algorithm. Must be not less than winSize argument of 110 calcOpticalFlowPyrLK. It is needed to calculate required padding for pyramid levels. 111 @param maxLevel 0-based maximal pyramid level number. 112 @param withDerivatives set to precompute gradients for the every pyramid level. If pyramid is 113 constructed without the gradients then calcOpticalFlowPyrLK will calculate them internally. 114 @param pyrBorder the border mode for pyramid layers. 115 @param derivBorder the border mode for gradients. 116 @param tryReuseInputImage put ROI of input image into the pyramid if possible. You can pass false 117 to force data copying. 118 @return number of levels in constructed pyramid. Can be less than maxLevel. 119 */ 120 CV_EXPORTS_W int buildOpticalFlowPyramid( InputArray img, OutputArrayOfArrays pyramid, 121 Size winSize, int maxLevel, bool withDerivatives = true, 122 int pyrBorder = BORDER_REFLECT_101, 123 int derivBorder = BORDER_CONSTANT, 124 bool tryReuseInputImage = true ); 125 126 /** @brief Calculates an optical flow for a sparse feature set using the iterative Lucas-Kanade method with 127 pyramids. 128 129 @param prevImg first 8-bit input image or pyramid constructed by buildOpticalFlowPyramid. 130 @param nextImg second input image or pyramid of the same size and the same type as prevImg. 131 @param prevPts vector of 2D points for which the flow needs to be found; point coordinates must be 132 single-precision floating-point numbers. 133 @param nextPts output vector of 2D points (with single-precision floating-point coordinates) 134 containing the calculated new positions of input features in the second image; when 135 OPTFLOW_USE_INITIAL_FLOW flag is passed, the vector must have the same size as in the input. 136 @param status output status vector (of unsigned chars); each element of the vector is set to 1 if 137 the flow for the corresponding features has been found, otherwise, it is set to 0. 138 @param err output vector of errors; each element of the vector is set to an error for the 139 corresponding feature, type of the error measure can be set in flags parameter; if the flow wasn't 140 found then the error is not defined (use the status parameter to find such cases). 141 @param winSize size of the search window at each pyramid level. 142 @param maxLevel 0-based maximal pyramid level number; if set to 0, pyramids are not used (single 143 level), if set to 1, two levels are used, and so on; if pyramids are passed to input then 144 algorithm will use as many levels as pyramids have but no more than maxLevel. 145 @param criteria parameter, specifying the termination criteria of the iterative search algorithm 146 (after the specified maximum number of iterations criteria.maxCount or when the search window 147 moves by less than criteria.epsilon. 148 @param flags operation flags: 149 - **OPTFLOW_USE_INITIAL_FLOW** uses initial estimations, stored in nextPts; if the flag is 150 not set, then prevPts is copied to nextPts and is considered the initial estimate. 151 - **OPTFLOW_LK_GET_MIN_EIGENVALS** use minimum eigen values as an error measure (see 152 minEigThreshold description); if the flag is not set, then L1 distance between patches 153 around the original and a moved point, divided by number of pixels in a window, is used as a 154 error measure. 155 @param minEigThreshold the algorithm calculates the minimum eigen value of a 2x2 normal matrix of 156 optical flow equations (this matrix is called a spatial gradient matrix in @cite Bouguet00), divided 157 by number of pixels in a window; if this value is less than minEigThreshold, then a corresponding 158 feature is filtered out and its flow is not processed, so it allows to remove bad points and get a 159 performance boost. 160 161 The function implements a sparse iterative version of the Lucas-Kanade optical flow in pyramids. See 162 @cite Bouguet00 . The function is parallelized with the TBB library. 163 164 @note 165 166 - An example using the Lucas-Kanade optical flow algorithm can be found at 167 opencv_source_code/samples/cpp/lkdemo.cpp 168 - (Python) An example using the Lucas-Kanade optical flow algorithm can be found at 169 opencv_source_code/samples/python2/lk_track.py 170 - (Python) An example using the Lucas-Kanade tracker for homography matching can be found at 171 opencv_source_code/samples/python2/lk_homography.py 172 */ 173 CV_EXPORTS_W void calcOpticalFlowPyrLK( InputArray prevImg, InputArray nextImg, 174 InputArray prevPts, InputOutputArray nextPts, 175 OutputArray status, OutputArray err, 176 Size winSize = Size(21,21), int maxLevel = 3, 177 TermCriteria criteria = TermCriteria(TermCriteria::COUNT+TermCriteria::EPS, 30, 0.01), 178 int flags = 0, double minEigThreshold = 1e-4 ); 179 180 /** @brief Computes a dense optical flow using the Gunnar Farneback's algorithm. 181 182 @param prev first 8-bit single-channel input image. 183 @param next second input image of the same size and the same type as prev. 184 @param flow computed flow image that has the same size as prev and type CV_32FC2. 185 @param pyr_scale parameter, specifying the image scale (\<1) to build pyramids for each image; 186 pyr_scale=0.5 means a classical pyramid, where each next layer is twice smaller than the previous 187 one. 188 @param levels number of pyramid layers including the initial image; levels=1 means that no extra 189 layers are created and only the original images are used. 190 @param winsize averaging window size; larger values increase the algorithm robustness to image 191 noise and give more chances for fast motion detection, but yield more blurred motion field. 192 @param iterations number of iterations the algorithm does at each pyramid level. 193 @param poly_n size of the pixel neighborhood used to find polynomial expansion in each pixel; 194 larger values mean that the image will be approximated with smoother surfaces, yielding more 195 robust algorithm and more blurred motion field, typically poly_n =5 or 7. 196 @param poly_sigma standard deviation of the Gaussian that is used to smooth derivatives used as a 197 basis for the polynomial expansion; for poly_n=5, you can set poly_sigma=1.1, for poly_n=7, a 198 good value would be poly_sigma=1.5. 199 @param flags operation flags that can be a combination of the following: 200 - **OPTFLOW_USE_INITIAL_FLOW** uses the input flow as an initial flow approximation. 201 - **OPTFLOW_FARNEBACK_GAUSSIAN** uses the Gaussian \f$\texttt{winsize}\times\texttt{winsize}\f$ 202 filter instead of a box filter of the same size for optical flow estimation; usually, this 203 option gives z more accurate flow than with a box filter, at the cost of lower speed; 204 normally, winsize for a Gaussian window should be set to a larger value to achieve the same 205 level of robustness. 206 207 The function finds an optical flow for each prev pixel using the @cite Farneback2003 algorithm so that 208 209 \f[\texttt{prev} (y,x) \sim \texttt{next} ( y + \texttt{flow} (y,x)[1], x + \texttt{flow} (y,x)[0])\f] 210 211 @note 212 213 - An example using the optical flow algorithm described by Gunnar Farneback can be found at 214 opencv_source_code/samples/cpp/fback.cpp 215 - (Python) An example using the optical flow algorithm described by Gunnar Farneback can be 216 found at opencv_source_code/samples/python2/opt_flow.py 217 */ 218 CV_EXPORTS_W void calcOpticalFlowFarneback( InputArray prev, InputArray next, InputOutputArray flow, 219 double pyr_scale, int levels, int winsize, 220 int iterations, int poly_n, double poly_sigma, 221 int flags ); 222 223 /** @brief Computes an optimal affine transformation between two 2D point sets. 224 225 @param src First input 2D point set stored in std::vector or Mat, or an image stored in Mat. 226 @param dst Second input 2D point set of the same size and the same type as A, or another image. 227 @param fullAffine If true, the function finds an optimal affine transformation with no additional 228 restrictions (6 degrees of freedom). Otherwise, the class of transformations to choose from is 229 limited to combinations of translation, rotation, and uniform scaling (5 degrees of freedom). 230 231 The function finds an optimal affine transform *[A|b]* (a 2 x 3 floating-point matrix) that 232 approximates best the affine transformation between: 233 234 * Two point sets 235 * Two raster images. In this case, the function first finds some features in the src image and 236 finds the corresponding features in dst image. After that, the problem is reduced to the first 237 case. 238 In case of point sets, the problem is formulated as follows: you need to find a 2x2 matrix *A* and 239 2x1 vector *b* so that: 240 241 \f[[A^*|b^*] = arg \min _{[A|b]} \sum _i \| \texttt{dst}[i] - A { \texttt{src}[i]}^T - b \| ^2\f] 242 where src[i] and dst[i] are the i-th points in src and dst, respectively 243 \f$[A|b]\f$ can be either arbitrary (when fullAffine=true ) or have a form of 244 \f[\begin{bmatrix} a_{11} & a_{12} & b_1 \\ -a_{12} & a_{11} & b_2 \end{bmatrix}\f] 245 when fullAffine=false. 246 247 @sa 248 getAffineTransform, getPerspectiveTransform, findHomography 249 */ 250 CV_EXPORTS_W Mat estimateRigidTransform( InputArray src, InputArray dst, bool fullAffine ); 251 252 253 enum 254 { 255 MOTION_TRANSLATION = 0, 256 MOTION_EUCLIDEAN = 1, 257 MOTION_AFFINE = 2, 258 MOTION_HOMOGRAPHY = 3 259 }; 260 261 /** @brief Finds the geometric transform (warp) between two images in terms of the ECC criterion @cite EP08 . 262 263 @param templateImage single-channel template image; CV_8U or CV_32F array. 264 @param inputImage single-channel input image which should be warped with the final warpMatrix in 265 order to provide an image similar to templateImage, same type as temlateImage. 266 @param warpMatrix floating-point \f$2\times 3\f$ or \f$3\times 3\f$ mapping matrix (warp). 267 @param motionType parameter, specifying the type of motion: 268 - **MOTION_TRANSLATION** sets a translational motion model; warpMatrix is \f$2\times 3\f$ with 269 the first \f$2\times 2\f$ part being the unity matrix and the rest two parameters being 270 estimated. 271 - **MOTION_EUCLIDEAN** sets a Euclidean (rigid) transformation as motion model; three 272 parameters are estimated; warpMatrix is \f$2\times 3\f$. 273 - **MOTION_AFFINE** sets an affine motion model (DEFAULT); six parameters are estimated; 274 warpMatrix is \f$2\times 3\f$. 275 - **MOTION_HOMOGRAPHY** sets a homography as a motion model; eight parameters are 276 estimated;\`warpMatrix\` is \f$3\times 3\f$. 277 @param criteria parameter, specifying the termination criteria of the ECC algorithm; 278 criteria.epsilon defines the threshold of the increment in the correlation coefficient between two 279 iterations (a negative criteria.epsilon makes criteria.maxcount the only termination criterion). 280 Default values are shown in the declaration above. 281 @param inputMask An optional mask to indicate valid values of inputImage. 282 283 The function estimates the optimum transformation (warpMatrix) with respect to ECC criterion 284 (@cite EP08), that is 285 286 \f[\texttt{warpMatrix} = \texttt{warpMatrix} = \arg\max_{W} \texttt{ECC}(\texttt{templateImage}(x,y),\texttt{inputImage}(x',y'))\f] 287 288 where 289 290 \f[\begin{bmatrix} x' \\ y' \end{bmatrix} = W \cdot \begin{bmatrix} x \\ y \\ 1 \end{bmatrix}\f] 291 292 (the equation holds with homogeneous coordinates for homography). It returns the final enhanced 293 correlation coefficient, that is the correlation coefficient between the template image and the 294 final warped input image. When a \f$3\times 3\f$ matrix is given with motionType =0, 1 or 2, the third 295 row is ignored. 296 297 Unlike findHomography and estimateRigidTransform, the function findTransformECC implements an 298 area-based alignment that builds on intensity similarities. In essence, the function updates the 299 initial transformation that roughly aligns the images. If this information is missing, the identity 300 warp (unity matrix) should be given as input. Note that if images undergo strong 301 displacements/rotations, an initial transformation that roughly aligns the images is necessary 302 (e.g., a simple euclidean/similarity transform that allows for the images showing the same image 303 content approximately). Use inverse warping in the second image to take an image close to the first 304 one, i.e. use the flag WARP_INVERSE_MAP with warpAffine or warpPerspective. See also the OpenCV 305 sample image_alignment.cpp that demonstrates the use of the function. Note that the function throws 306 an exception if algorithm does not converges. 307 308 @sa 309 estimateRigidTransform, findHomography 310 */ 311 CV_EXPORTS_W double findTransformECC( InputArray templateImage, InputArray inputImage, 312 InputOutputArray warpMatrix, int motionType = MOTION_AFFINE, 313 TermCriteria criteria = TermCriteria(TermCriteria::COUNT+TermCriteria::EPS, 50, 0.001), 314 InputArray inputMask = noArray()); 315 316 /** @brief Kalman filter class. 317 318 The class implements a standard Kalman filter <http://en.wikipedia.org/wiki/Kalman_filter>, 319 @cite Welch95 . However, you can modify transitionMatrix, controlMatrix, and measurementMatrix to get 320 an extended Kalman filter functionality. See the OpenCV sample kalman.cpp. 321 322 @note 323 324 - An example using the standard Kalman filter can be found at 325 opencv_source_code/samples/cpp/kalman.cpp 326 */ 327 class CV_EXPORTS_W KalmanFilter 328 { 329 public: 330 /** @brief The constructors. 331 332 @note In C API when CvKalman\* kalmanFilter structure is not needed anymore, it should be released 333 with cvReleaseKalman(&kalmanFilter) 334 */ 335 CV_WRAP KalmanFilter(); 336 /** @overload 337 @param dynamParams Dimensionality of the state. 338 @param measureParams Dimensionality of the measurement. 339 @param controlParams Dimensionality of the control vector. 340 @param type Type of the created matrices that should be CV_32F or CV_64F. 341 */ 342 CV_WRAP KalmanFilter( int dynamParams, int measureParams, int controlParams = 0, int type = CV_32F ); 343 344 /** @brief Re-initializes Kalman filter. The previous content is destroyed. 345 346 @param dynamParams Dimensionality of the state. 347 @param measureParams Dimensionality of the measurement. 348 @param controlParams Dimensionality of the control vector. 349 @param type Type of the created matrices that should be CV_32F or CV_64F. 350 */ 351 void init( int dynamParams, int measureParams, int controlParams = 0, int type = CV_32F ); 352 353 /** @brief Computes a predicted state. 354 355 @param control The optional input control 356 */ 357 CV_WRAP const Mat& predict( const Mat& control = Mat() ); 358 359 /** @brief Updates the predicted state from the measurement. 360 361 @param measurement The measured system parameters 362 */ 363 CV_WRAP const Mat& correct( const Mat& measurement ); 364 365 CV_PROP_RW Mat statePre; //!< predicted state (x'(k)): x(k)=A*x(k-1)+B*u(k) 366 CV_PROP_RW Mat statePost; //!< corrected state (x(k)): x(k)=x'(k)+K(k)*(z(k)-H*x'(k)) 367 CV_PROP_RW Mat transitionMatrix; //!< state transition matrix (A) 368 CV_PROP_RW Mat controlMatrix; //!< control matrix (B) (not used if there is no control) 369 CV_PROP_RW Mat measurementMatrix; //!< measurement matrix (H) 370 CV_PROP_RW Mat processNoiseCov; //!< process noise covariance matrix (Q) 371 CV_PROP_RW Mat measurementNoiseCov;//!< measurement noise covariance matrix (R) 372 CV_PROP_RW Mat errorCovPre; //!< priori error estimate covariance matrix (P'(k)): P'(k)=A*P(k-1)*At + Q)*/ 373 CV_PROP_RW Mat gain; //!< Kalman gain matrix (K(k)): K(k)=P'(k)*Ht*inv(H*P'(k)*Ht+R) 374 CV_PROP_RW Mat errorCovPost; //!< posteriori error estimate covariance matrix (P(k)): P(k)=(I-K(k)*H)*P'(k) 375 376 // temporary matrices 377 Mat temp1; 378 Mat temp2; 379 Mat temp3; 380 Mat temp4; 381 Mat temp5; 382 }; 383 384 385 class CV_EXPORTS_W DenseOpticalFlow : public Algorithm 386 { 387 public: 388 /** @brief Calculates an optical flow. 389 390 @param I0 first 8-bit single-channel input image. 391 @param I1 second input image of the same size and the same type as prev. 392 @param flow computed flow image that has the same size as prev and type CV_32FC2. 393 */ 394 CV_WRAP virtual void calc( InputArray I0, InputArray I1, InputOutputArray flow ) = 0; 395 /** @brief Releases all inner buffers. 396 */ 397 CV_WRAP virtual void collectGarbage() = 0; 398 }; 399 400 /** @brief "Dual TV L1" Optical Flow Algorithm. 401 402 The class implements the "Dual TV L1" optical flow algorithm described in @cite Zach2007 and 403 @cite Javier2012 . 404 Here are important members of the class that control the algorithm, which you can set after 405 constructing the class instance: 406 407 - member double tau 408 Time step of the numerical scheme. 409 410 - member double lambda 411 Weight parameter for the data term, attachment parameter. This is the most relevant 412 parameter, which determines the smoothness of the output. The smaller this parameter is, 413 the smoother the solutions we obtain. It depends on the range of motions of the images, so 414 its value should be adapted to each image sequence. 415 416 - member double theta 417 Weight parameter for (u - v)\^2, tightness parameter. It serves as a link between the 418 attachment and the regularization terms. In theory, it should have a small value in order 419 to maintain both parts in correspondence. The method is stable for a large range of values 420 of this parameter. 421 422 - member int nscales 423 Number of scales used to create the pyramid of images. 424 425 - member int warps 426 Number of warpings per scale. Represents the number of times that I1(x+u0) and grad( 427 I1(x+u0) ) are computed per scale. This is a parameter that assures the stability of the 428 method. It also affects the running time, so it is a compromise between speed and 429 accuracy. 430 431 - member double epsilon 432 Stopping criterion threshold used in the numerical scheme, which is a trade-off between 433 precision and running time. A small value will yield more accurate solutions at the 434 expense of a slower convergence. 435 436 - member int iterations 437 Stopping criterion iterations number used in the numerical scheme. 438 439 C. Zach, T. Pock and H. Bischof, "A Duality Based Approach for Realtime TV-L1 Optical Flow". 440 Javier Sanchez, Enric Meinhardt-Llopis and Gabriele Facciolo. "TV-L1 Optical Flow Estimation". 441 */ 442 class CV_EXPORTS_W DualTVL1OpticalFlow : public DenseOpticalFlow 443 { 444 public: 445 //! @brief Time step of the numerical scheme 446 /** @see setTau */ 447 virtual double getTau() const = 0; 448 /** @copybrief getTau @see getTau */ 449 virtual void setTau(double val) = 0; 450 //! @brief Weight parameter for the data term, attachment parameter 451 /** @see setLambda */ 452 virtual double getLambda() const = 0; 453 /** @copybrief getLambda @see getLambda */ 454 virtual void setLambda(double val) = 0; 455 //! @brief Weight parameter for (u - v)^2, tightness parameter 456 /** @see setTheta */ 457 virtual double getTheta() const = 0; 458 /** @copybrief getTheta @see getTheta */ 459 virtual void setTheta(double val) = 0; 460 //! @brief coefficient for additional illumination variation term 461 /** @see setGamma */ 462 virtual double getGamma() const = 0; 463 /** @copybrief getGamma @see getGamma */ 464 virtual void setGamma(double val) = 0; 465 //! @brief Number of scales used to create the pyramid of images 466 /** @see setScalesNumber */ 467 virtual int getScalesNumber() const = 0; 468 /** @copybrief getScalesNumber @see getScalesNumber */ 469 virtual void setScalesNumber(int val) = 0; 470 //! @brief Number of warpings per scale 471 /** @see setWarpingsNumber */ 472 virtual int getWarpingsNumber() const = 0; 473 /** @copybrief getWarpingsNumber @see getWarpingsNumber */ 474 virtual void setWarpingsNumber(int val) = 0; 475 //! @brief Stopping criterion threshold used in the numerical scheme, which is a trade-off between precision and running time 476 /** @see setEpsilon */ 477 virtual double getEpsilon() const = 0; 478 /** @copybrief getEpsilon @see getEpsilon */ 479 virtual void setEpsilon(double val) = 0; 480 //! @brief Inner iterations (between outlier filtering) used in the numerical scheme 481 /** @see setInnerIterations */ 482 virtual int getInnerIterations() const = 0; 483 /** @copybrief getInnerIterations @see getInnerIterations */ 484 virtual void setInnerIterations(int val) = 0; 485 //! @brief Outer iterations (number of inner loops) used in the numerical scheme 486 /** @see setOuterIterations */ 487 virtual int getOuterIterations() const = 0; 488 /** @copybrief getOuterIterations @see getOuterIterations */ 489 virtual void setOuterIterations(int val) = 0; 490 //! @brief Use initial flow 491 /** @see setUseInitialFlow */ 492 virtual bool getUseInitialFlow() const = 0; 493 /** @copybrief getUseInitialFlow @see getUseInitialFlow */ 494 virtual void setUseInitialFlow(bool val) = 0; 495 //! @brief Step between scales (<1) 496 /** @see setScaleStep */ 497 virtual double getScaleStep() const = 0; 498 /** @copybrief getScaleStep @see getScaleStep */ 499 virtual void setScaleStep(double val) = 0; 500 //! @brief Median filter kernel size (1 = no filter) (3 or 5) 501 /** @see setMedianFiltering */ 502 virtual int getMedianFiltering() const = 0; 503 /** @copybrief getMedianFiltering @see getMedianFiltering */ 504 virtual void setMedianFiltering(int val) = 0; 505 }; 506 507 /** @brief Creates instance of cv::DenseOpticalFlow 508 */ 509 CV_EXPORTS_W Ptr<DualTVL1OpticalFlow> createOptFlow_DualTVL1(); 510 511 //! @} video_track 512 513 } // cv 514 515 #endif 516