1 \section{Motion Analysis and Object Tracking}
6 Adds a frame to an accumulator.
9 void cvAcc( \par const CvArr* image,\par CvArr* sum,\par const CvArr* mask=NULL );
11 \cvdefPy{Acc(image,sum,mask=NULL)-> None}
14 \cvarg{image}{Input image, 1- or 3-channel, 8-bit or 32-bit floating point. (each channel of multi-channel image is processed independently)}
15 \cvarg{sum}{Accumulator with the same number of channels as input image, 32-bit or 64-bit floating-point}
16 \cvarg{mask}{Optional operation mask}
19 The function adds the whole image \texttt{image} or its selected region to the accumulator \texttt{sum}:
21 \[ \texttt{sum}(x,y) \leftarrow \texttt{sum}(x,y) + \texttt{image}(x,y) \quad \text{if} \quad \texttt{mask}(x,y) \ne 0 \]
23 \cvCPyFunc{CalcGlobalOrientation}
24 Calculates the global motion orientation of some selected region.
27 double cvCalcGlobalOrientation( \par const CvArr* orientation,\par const CvArr* mask,\par const CvArr* mhi,\par double timestamp,\par double duration );
28 }\cvdefPy{CalcGlobalOrientation(orientation,mask,mhi,timestamp,duration)-> float}
31 \cvarg{orientation}{Motion gradient orientation image; calculated by the function \cvCPyCross{CalcMotionGradient}}
32 \cvarg{mask}{Mask image. It may be a conjunction of a valid gradient mask, obtained with \cvCPyCross{CalcMotionGradient} and the mask of the region, whose direction needs to be calculated}
33 \cvarg{mhi}{Motion history image}
34 \cvarg{timestamp}{Current time in milliseconds or other units, it is better to store time passed to \cvCPyCross{UpdateMotionHistory} before and reuse it here, because running \cvCPyCross{UpdateMotionHistory} and \cvCPyCross{CalcMotionGradient} on large images may take some time}
35 \cvarg{duration}{Maximal duration of motion track in milliseconds, the same as \cvCPyCross{UpdateMotionHistory}}
38 The function calculates the general
39 motion direction in the selected region and returns the angle between
40 0 degrees and 360 degrees . At first the function builds the orientation histogram
41 and finds the basic orientation as a coordinate of the histogram
42 maximum. After that the function calculates the shift relative to the
43 basic orientation as a weighted sum of all of the orientation vectors: the more
44 recent the motion, the greater the weight. The resultant angle is
45 a circular sum of the basic orientation and the shift.
47 \cvCPyFunc{CalcMotionGradient}
48 Calculates the gradient orientation of a motion history image.
51 void cvCalcMotionGradient( \par const CvArr* mhi,\par CvArr* mask,\par CvArr* orientation,\par double delta1,\par double delta2,\par int apertureSize=3 );
52 }\cvdefPy{CalcMotionGradient(mhi,mask,orientation,delta1,delta2,apertureSize=3)-> None}
55 \cvarg{mhi}{Motion history image}
56 \cvarg{mask}{Mask image; marks pixels where the motion gradient data is correct; output parameter}
57 \cvarg{orientation}{Motion gradient orientation image; contains angles from 0 to ~360 degrees }
58 \cvarg{delta1}{See below}
59 \cvarg{delta2}{See below}
60 \cvarg{apertureSize}{Aperture size of derivative operators used by the function: CV\_SCHARR, 1, 3, 5 or 7 (see \cvCPyCross{Sobel})}
63 The function calculates the derivatives $Dx$ and $Dy$ of \texttt{mhi} and then calculates gradient orientation as:
66 \texttt{orientation}(x,y)=\arctan{\frac{Dy(x,y)}{Dx(x,y)}}
69 where both $Dx(x,y)$ and $Dy(x,y)$ signs are taken into account (as in the \cvCPyCross{CartToPolar} function). After that \texttt{mask} is filled to indicate where the orientation is valid (see the \texttt{delta1} and \texttt{delta2} description).
71 The function finds the minimum ($m(x,y)$) and maximum ($M(x,y)$) mhi values over each pixel $(x,y)$ neighborhood and assumes the gradient is valid only if
73 \min(\texttt{delta1} , \texttt{delta2} ) \le M(x,y)-m(x,y) \le \max(\texttt{delta1} ,\texttt{delta2} ).
76 \cvCPyFunc{CalcOpticalFlowBM}
77 Calculates the optical flow for two images by using the block matching method.
80 void cvCalcOpticalFlowBM( \par const CvArr* prev,\par const CvArr* curr,\par CvSize blockSize,\par CvSize shiftSize,\par CvSize max\_range,\par int usePrevious,\par CvArr* velx,\par CvArr* vely );
81 }\cvdefPy{CalcOpticalFlowBM(prev,curr,blockSize,shiftSize,max\_range,usePrevious,velx,vely)-> None}
84 \cvarg{prev}{First image, 8-bit, single-channel}
85 \cvarg{curr}{Second image, 8-bit, single-channel}
86 \cvarg{blockSize}{Size of basic blocks that are compared}
87 \cvarg{shiftSize}{Block coordinate increments}
88 \cvarg{max\_range}{Size of the scanned neighborhood in pixels around the block}
89 \cvarg{usePrevious}{Uses the previous (input) velocity field}
90 \cvarg{velx}{Horizontal component of the optical flow of
92 \left\lfloor \frac{\texttt{prev->width} - \texttt{blockSize.width}}{\texttt{shiftSize.width}} \right\rfloor
94 \left\lfloor \frac{\texttt{prev->height} - \texttt{blockSize.height}}{\texttt{shiftSize.height}} \right\rfloor
96 size, 32-bit floating-point, single-channel}
97 \cvarg{vely}{Vertical component of the optical flow of the same size \texttt{velx}, 32-bit floating-point, single-channel}
100 The function calculates the optical
101 flow for overlapped blocks $\texttt{blockSize.width} \times \texttt{blockSize.height}$ pixels each, thus the velocity
102 fields are smaller than the original images. For every block in \texttt{prev} the functions tries to find a similar block in
103 \texttt{curr} in some neighborhood of the original block or shifted by (velx(x0,y0),vely(x0,y0)) block as has been calculated by previous
104 function call (if \texttt{usePrevious=1})
106 \cvCPyFunc{CalcOpticalFlowHS}
107 Calculates the optical flow for two images.
110 void cvCalcOpticalFlowHS( \par const CvArr* prev,\par const CvArr* curr,\par int usePrevious,\par CvArr* velx,\par CvArr* vely,\par double lambda,\par CvTermCriteria criteria );
111 }\cvdefPy{CalcOpticalFlowHS(prev,curr,usePrevious,velx,vely,lambda,criteria)-> None}
114 \cvarg{prev}{First image, 8-bit, single-channel}
115 \cvarg{curr}{Second image, 8-bit, single-channel}
116 \cvarg{usePrevious}{Uses the previous (input) velocity field}
117 \cvarg{velx}{Horizontal component of the optical flow of the same size as input images, 32-bit floating-point, single-channel}
118 \cvarg{vely}{Vertical component of the optical flow of the same size as input images, 32-bit floating-point, single-channel}
119 \cvarg{lambda}{Lagrangian multiplier}
120 \cvarg{criteria}{Criteria of termination of velocity computing}
123 The function computes the flow for every pixel of the first input image using the Horn and Schunck algorithm
126 \cvCPyFunc{CalcOpticalFlowLK}
127 Calculates the optical flow for two images.
130 void cvCalcOpticalFlowLK( \par const CvArr* prev,\par const CvArr* curr,\par CvSize winSize,\par CvArr* velx,\par CvArr* vely );
131 }\cvdefPy{CalcOpticalFlowLK(prev,curr,winSize,velx,vely)-> None}
134 \cvarg{prev}{First image, 8-bit, single-channel}
135 \cvarg{curr}{Second image, 8-bit, single-channel}
136 \cvarg{winSize}{Size of the averaging window used for grouping pixels}
137 \cvarg{velx}{Horizontal component of the optical flow of the same size as input images, 32-bit floating-point, single-channel}
138 \cvarg{vely}{Vertical component of the optical flow of the same size as input images, 32-bit floating-point, single-channel}
141 The function computes the flow for every pixel of the first input image using the Lucas and Kanade algorithm
144 \cvCPyFunc{CalcOpticalFlowPyrLK}
145 Calculates the optical flow for a sparse feature set using the iterative Lucas-Kanade method with pyramids.
148 void cvCalcOpticalFlowPyrLK( \par const CvArr* prev,\par const CvArr* curr,\par CvArr* prevPyr,\par CvArr* currPyr,\par const CvPoint2D32f* prevFeatures,\par CvPoint2D32f* currFeatures,\par int count,\par CvSize winSize,\par int level,\par char* status,\par float* track\_error,\par CvTermCriteria criteria,\par int flags );
151 CalcOpticalFlowPyrLK( prev, curr, prevPyr, currPyr, prevFeatures, winSize, level, criteria, flags, guesses = None) -> (currFeatures, status, track\_error)
155 \cvarg{prev}{First frame, at time \texttt{t}}
156 \cvarg{curr}{Second frame, at time \texttt{t + dt} }
157 \cvarg{prevPyr}{Buffer for the pyramid for the first frame. If the pointer is not \texttt{NULL} , the buffer must have a sufficient size to store the pyramid from level \texttt{1} to level \texttt{level} ; the total size of \texttt{(image\_width+8)*image\_height/3} bytes is sufficient}
158 \cvarg{currPyr}{Similar to \texttt{prevPyr}, used for the second frame}
159 \cvarg{prevFeatures}{Array of points for which the flow needs to be found}
160 \cvarg{currFeatures}{Array of 2D points containing the calculated new positions of the input features in the second image}
162 \cvarg{count}{Number of feature points}
164 \cvarg{winSize}{Size of the search window of each pyramid level}
165 \cvarg{level}{Maximal pyramid level number. If \texttt{0} , pyramids are not used (single level), if \texttt{1} , two levels are used, etc}
166 \cvarg{status}{Array. Every element of the array is set to \texttt{1} if the flow for the corresponding feature has been found, \texttt{0} otherwise}
167 \cvarg{track\_error}{Array of double numbers containing the difference between patches around the original and moved points. Optional parameter; can be \texttt{NULL}}
168 \cvarg{criteria}{Specifies when the iteration process of finding the flow for each point on each pyramid level should be stopped}
169 \cvarg{flags}{Miscellaneous flags:
171 \cvarg{CV\_LKFLOWPyr\_A\_READY}{pyramid for the first frame is precalculated before the call}
172 \cvarg{CV\_LKFLOWPyr\_B\_READY}{ pyramid for the second frame is precalculated before the call}
173 \cvC{\cvarg{CV\_LKFLOW\_INITIAL\_GUESSES}{array B contains initial coordinates of features before the function call}}
175 \cvPy{\cvarg{guesses}{optional array of estimated coordinates of features in second frame, with same length as \texttt{prevFeatures}}}
178 The function implements the sparse iterative version of the Lucas-Kanade optical flow in pyramids
180 . It calculates the coordinates of the feature points on the current video
181 frame given their coordinates on the previous frame. The function finds
182 the coordinates with sub-pixel accuracy.
184 Both parameters \texttt{prevPyr} and \texttt{currPyr} comply with the
185 following rules: if the image pointer is 0, the function allocates the
186 buffer internally, calculates the pyramid, and releases the buffer after
187 processing. Otherwise, the function calculates the pyramid and stores
188 it in the buffer unless the flag \texttt{CV\_LKFLOWPyr\_A[B]\_READY}
189 is set. The image should be large enough to fit the Gaussian pyramid
190 data. After the function call both pyramids are calculated and the
191 readiness flag for the corresponding image can be set in the next call
192 (i.e., typically, for all the image pairs except the very first one
193 \texttt{CV\_LKFLOWPyr\_A\_READY} is set).
197 Finds the object center, size, and orientation.
200 int cvCamShift( \par const CvArr* prob\_image,\par CvRect window,\par CvTermCriteria criteria,\par CvConnectedComp* comp,\par CvBox2D* box=NULL );
202 \cvdefPy{CamShift(prob\_image,window,criteria)-> (int, comp, box)}
205 \cvarg{prob\_image}{Back projection of object histogram (see \cvCPyCross{CalcBackProject})}
206 \cvarg{window}{Initial search window}
207 \cvarg{criteria}{Criteria applied to determine when the window search should be finished}
208 \cvarg{comp}{Resultant structure that contains the converged search window coordinates (\texttt{comp->rect} field) and the sum of all of the pixels inside the window (\texttt{comp->area} field)}
210 \cvarg{box}{Circumscribed box for the object. If not \texttt{NULL}, it contains object size and orientation}
212 \cvarg{box}{Circumscribed box for the object.}
216 The function implements the CAMSHIFT object tracking algrorithm
218 First, it finds an object center using \cvCPyCross{MeanShift} and, after that, calculates the object size and orientation. The function returns number of iterations made within \cvCPyCross{MeanShift}.
220 The \texttt{CamShiftTracker} class declared in cv.hpp implements the color object tracker that uses the function.
223 \subsection{CvConDensation}
227 typedef struct CvConDensation
229 int MP; //Dimension of measurement vector
230 int DP; // Dimension of state vector
231 float* DynamMatr; // Matrix of the linear Dynamics system
232 float* State; // Vector of State
233 int SamplesNum; // Number of the Samples
234 float** flSamples; // array of the Sample Vectors
235 float** flNewSamples; // temporary array of the Sample Vectors
236 float* flConfidence; // Confidence for each Sample
237 float* flCumulative; // Cumulative confidence
238 float* Temp; // Temporary vector
239 float* RandomSample; // RandomVector to update sample set
240 CvRandState* RandS; // Array of structures to generate random vectors
244 The structure \texttt{CvConDensation} stores the CONditional DENSity propagATION tracker state. The information about the algorithm can be found at \url{http://www.dai.ed.ac.uk/CVonline/LOCAL\_COPIES/ISARD1/condensation.html}.
246 \cvCPyFunc{CreateConDensation}
247 Allocates the ConDensation filter structure.
250 CvConDensation* cvCreateConDensation( \par int dynam\_params,\par int measure\_params,\par int sample\_count );
254 \cvarg{dynam\_params}{Dimension of the state vector}
255 \cvarg{measure\_params}{Dimension of the measurement vector}
256 \cvarg{sample\_count}{Number of samples}
259 The function creates a \texttt{CvConDensation} structure and returns a pointer to the structure.
261 \cvCPyFunc{ConDensInitSampleSet}
262 Initializes the sample set for the ConDensation algorithm.
265 void cvConDensInitSampleSet( CvConDensation* condens, \par CvMat* lower\_bound, \par CvMat* upper\_bound );
269 \cvarg{condens}{Pointer to a structure to be initialized}
270 \cvarg{lower\_bound}{Vector of the lower boundary for each dimension}
271 \cvarg{upper\_bound}{Vector of the upper boundary for each dimension}
274 The function fills the samples arrays in the structure \texttt{condens} with values within the specified ranges.
277 \cvclass{CvKalman}\label{CvKalman}
282 typedef struct CvKalman
284 int MP; /* number of measurement vector dimensions */
285 int DP; /* number of state vector dimensions */
286 int CP; /* number of control vector dimensions */
288 /* backward compatibility fields */
290 float* PosterState; /* =state_pre->data.fl */
291 float* PriorState; /* =state_post->data.fl */
292 float* DynamMatr; /* =transition_matrix->data.fl */
293 float* MeasurementMatr; /* =measurement_matrix->data.fl */
294 float* MNCovariance; /* =measurement_noise_cov->data.fl */
295 float* PNCovariance; /* =process_noise_cov->data.fl */
296 float* KalmGainMatr; /* =gain->data.fl */
297 float* PriorErrorCovariance;/* =error_cov_pre->data.fl */
298 float* PosterErrorCovariance;/* =error_cov_post->data.fl */
299 float* Temp1; /* temp1->data.fl */
300 float* Temp2; /* temp2->data.fl */
303 CvMat* state_pre; /* predicted state (x'(k)):
304 x(k)=A*x(k-1)+B*u(k) */
305 CvMat* state_post; /* corrected state (x(k)):
306 x(k)=x'(k)+K(k)*(z(k)-H*x'(k)) */
307 CvMat* transition_matrix; /* state transition matrix (A) */
308 CvMat* control_matrix; /* control matrix (B)
309 (it is not used if there is no control)*/
310 CvMat* measurement_matrix; /* measurement matrix (H) */
311 CvMat* process_noise_cov; /* process noise covariance matrix (Q) */
312 CvMat* measurement_noise_cov; /* measurement noise covariance matrix (R) */
313 CvMat* error_cov_pre; /* priori error estimate covariance matrix (P'(k)):
314 P'(k)=A*P(k-1)*At + Q*/
315 CvMat* gain; /* Kalman gain matrix (K(k)):
316 K(k)=P'(k)*Ht*inv(H*P'(k)*Ht+R)*/
317 CvMat* error_cov_post; /* posteriori error estimate covariance matrix (P(k)):
318 P(k)=(I-K(k)*H)*P'(k) */
319 CvMat* temp1; /* temporary matrices */
329 \cvarg{MP}{number of measurement vector dimensions}
330 \cvarg{DP}{number of state vector dimensions}
331 \cvarg{CP}{number of control vector dimensions}
332 \cvarg{state\_pre}{predicted state (x'(k)): x(k)=A*x(k-1)+B*u(k)}
333 \cvarg{state\_post}{corrected state (x(k)): x(k)=x'(k)+K(k)*(z(k)-H*x'(k))}
334 \cvarg{transition\_matrix}{state transition matrix (A)}
335 \cvarg{control\_matrix}{control matrix (B) (it is not used if there is no control)}
336 \cvarg{measurement\_matrix}{measurement matrix (H)}
337 \cvarg{process\_noise\_cov}{process noise covariance matrix (Q)}
338 \cvarg{measurement\_noise\_cov}{measurement noise covariance matrix (R)}
339 \cvarg{error\_cov\_pre}{priori error estimate covariance matrix (P'(k)): P'(k)=A*P(k-1)*At + Q}
340 \cvarg{gain}{Kalman gain matrix (K(k)): K(k)=P'(k)*Ht*inv(H*P'(k)*Ht+R)}
341 \cvarg{error\_cov\_post}{posteriori error estimate covariance matrix (P(k)): P(k)=(I-K(k)*H)*P'(k)}
345 The structure \texttt{CvKalman} is used to keep the Kalman filter
346 state. It is created by the \cvCPyCross{CreateKalman} function, updated
347 by the \cvCPyCross{KalmanPredict} and \cvCPyCross{KalmanCorrect} functions
348 and released by the \cvCPyCross{ReleaseKalman} function. Normally, the
349 structure is used for the standard Kalman filter (notation and the
350 formulas below are borrowed from the excellent Kalman tutorial
355 x_k=A \cdot x_{k-1}+B \cdot u_k+w_k\\
364 x_k\;(x_{k-1})& \text{state of the system at the moment \emph{k} (\emph{k-1})}\\
365 z_k & \text{measurement of the system state at the moment \emph{k}}\\
366 u_k & \text{external control applied at the moment \emph{k}}
370 $w_k$ and $v_k$ are normally-distributed process and measurement noise, respectively:
381 $Q$ process noise covariance matrix, constant or variable,
383 $R$ measurement noise covariance matrix, constant or variable
385 In the case of the standard Kalman filter, all of the matrices: A, B, H, Q and R are initialized once after the \cvCPyCross{CvKalman} structure is allocated via \cvCPyCross{CreateKalman}. However, the same structure and the same functions may be used to simulate the extended Kalman filter by linearizing the extended Kalman filter equation in the current system state neighborhood, in this case A, B, H (and, probably, Q and R) should be updated on every step.
387 \cvCPyFunc{CreateKalman}
388 Allocates the Kalman filter structure.
391 CvKalman* cvCreateKalman( \par int dynam\_params,\par int measure\_params,\par int control\_params=0 );
395 CreateKalman(dynam\_params, measure\_params, control\_params=0) -> CvKalman
399 \cvarg{dynam\_params}{dimensionality of the state vector}
400 \cvarg{measure\_params}{dimensionality of the measurement vector}
401 \cvarg{control\_params}{dimensionality of the control vector}
404 The function allocates \cvCPyCross{CvKalman} and all its matrices and initializes them somehow.
407 \cvCPyFunc{KalmanCorrect}
408 Adjusts the model state.
411 const CvMat* cvKalmanCorrect( CvKalman* kalman, const CvMat* measurement );
414 KalmanCorrect(kalman, measurement) -> cvmat
419 \cvarg{kalman}{Pointer to the structure to be updated}
421 \cvarg{kalman}{Kalman filter object returned by \cvCPyCross{CreateKalman}}
423 \cvarg{measurement}{CvMat containing the measurement vector}
426 The function adjusts the stochastic model state on the basis of the given measurement of the model state:
430 K_k=P'_k \cdot H^T \cdot (H \cdot P'_k \cdot H^T+R)^{-1}\\
431 x_k=x'_k+K_k \cdot (z_k-H \cdot x'_k)\\
432 P_k=(I-K_k \cdot H) \cdot P'_k
438 \begin{tabular}{l p{4 in}}
440 $z_k$ & given measurement (\texttt{mesurement} parameter)\\ \hline
441 $K_k$ & Kalman "gain" matrix.\\ \hline
444 The function stores the adjusted state at \texttt{kalman->state\_post} and returns it on output.
447 Example. Using Kalman filter to track a rotating point
453 int main(int argc, char** argv)
456 const float A[] = { 1, 1, 0, 1 };
458 IplImage* img = cvCreateImage( cvSize(500,500), 8, 3 );
459 CvKalman* kalman = cvCreateKalman( 2, 1, 0 );
460 /* state is (phi, delta_phi) - angle and angle increment */
461 CvMat* state = cvCreateMat( 2, 1, CV_32FC1 );
462 CvMat* process_noise = cvCreateMat( 2, 1, CV_32FC1 );
463 /* only phi (angle) is measured */
464 CvMat* measurement = cvCreateMat( 1, 1, CV_32FC1 );
468 cvRandInit( &rng, 0, 1, -1, CV_RAND_UNI );
470 cvZero( measurement );
471 cvNamedWindow( "Kalman", 1 );
475 cvRandSetRange( &rng, 0, 0.1, 0 );
476 rng.disttype = CV_RAND_NORMAL;
478 cvRand( &rng, state );
480 memcpy( kalman->transition_matrix->data.fl, A, sizeof(A));
481 cvSetIdentity( kalman->measurement_matrix, cvRealScalar(1) );
482 cvSetIdentity( kalman->process_noise_cov, cvRealScalar(1e-5) );
483 cvSetIdentity( kalman->measurement_noise_cov, cvRealScalar(1e-1) );
484 cvSetIdentity( kalman->error_cov_post, cvRealScalar(1));
485 /* choose random initial state */
486 cvRand( &rng, kalman->state_post );
488 rng.disttype = CV_RAND_NORMAL;
492 #define calc_point(angle) \
493 cvPoint( cvRound(img->width/2 + img->width/3*cos(angle)), \
494 cvRound(img->height/2 - img->width/3*sin(angle)))
496 float state_angle = state->data.fl[0];
497 CvPoint state_pt = calc_point(state_angle);
499 /* predict point position */
500 const CvMat* prediction = cvKalmanPredict( kalman, 0 );
501 float predict_angle = prediction->data.fl[0];
502 CvPoint predict_pt = calc_point(predict_angle);
503 float measurement_angle;
504 CvPoint measurement_pt;
506 cvRandSetRange( &rng,
508 sqrt(kalman->measurement_noise_cov->data.fl[0]),
510 cvRand( &rng, measurement );
512 /* generate measurement */
513 cvMatMulAdd( kalman->measurement_matrix, state, measurement, measurement );
515 measurement_angle = measurement->data.fl[0];
516 measurement_pt = calc_point(measurement_angle);
519 #define draw_cross( center, color, d ) \
520 cvLine( img, cvPoint( center.x - d, center.y - d ), \
521 cvPoint( center.x + d, center.y + d ), \
523 cvLine( img, cvPoint( center.x + d, center.y - d ), \
524 cvPoint( center.x - d, center.y + d ), \
528 draw_cross( state_pt, CV_RGB(255,255,255), 3 );
529 draw_cross( measurement_pt, CV_RGB(255,0,0), 3 );
530 draw_cross( predict_pt, CV_RGB(0,255,0), 3 );
531 cvLine( img, state_pt, predict_pt, CV_RGB(255,255,0), 3, 0 );
533 /* adjust Kalman filter state */
534 cvKalmanCorrect( kalman, measurement );
536 cvRandSetRange( &rng,
538 sqrt(kalman->process_noise_cov->data.fl[0]),
540 cvRand( &rng, process_noise );
541 cvMatMulAdd( kalman->transition_matrix,
546 cvShowImage( "Kalman", img );
547 code = cvWaitKey( 100 );
549 if( code > 0 ) /* break current simulation by pressing a key */
552 if( code == 27 ) /* exit by ESCAPE */
561 \cvCPyFunc{KalmanPredict}
562 Estimates the subsequent model state.
565 const CvMat* cvKalmanPredict( \par CvKalman* kalman, \par const CvMat* control=NULL );
568 KalmanPredict(kalman, control=None) -> cvmat
573 \cvarg{kalman}{Kalman filter state}
575 \cvarg{kalman}{Kalman filter object returned by \cvCPyCross{CreateKalman}}
577 \cvarg{control}{Control vector $u_k$, should be NULL iff there is no external control (\texttt{control\_params} =0)}
580 The function estimates the subsequent stochastic model state by its current state and stores it at \texttt{kalman->state\_pre}:
584 x'_k=A \cdot x_{k-1}+B \cdot u_k\\
585 P'_k=A \cdot P_{k-1}+A^T + Q
591 \begin{tabular}{l p{5 in}}
593 $x'_k$ & is predicted state \texttt{kalman->state\_pre},\\ \hline
594 $x_{k-1}$ & is corrected state on the previous step \texttt{kalman->state\_post}
595 (should be initialized somehow in the beginning, zero vector by default),\\ \hline
596 $u_k$ & is external control (\texttt{control} parameter),\\ \hline
597 $P'_k$ & is priori error covariance matrix \texttt{kalman->error\_cov\_pre}\\ \hline
598 $P_{k-1}$ & is posteriori error covariance matrix on the previous step \texttt{kalman->error\_cov\_post}
599 (should be initialized somehow in the beginning, identity matrix by default),
602 The function returns the estimated state.
604 \subsection{KalmanUpdateByMeasurement}
606 Synonym for \cross{KalmanCorrect}
608 \subsection{KalmanUpdateByTime}
610 Synonym for \cross{KalmanPredict}
612 \cvCPyFunc{MeanShift}
613 Finds the object center on back projection.
616 int cvMeanShift( \par const CvArr* prob\_image,\par CvRect window,\par CvTermCriteria criteria,\par CvConnectedComp* comp );
617 }\cvdefPy{MeanShift(prob\_image,window,criteria)-> comp}
620 \cvarg{prob\_image}{Back projection of the object histogram (see \cvCPyCross{CalcBackProject})}
621 \cvarg{window}{Initial search window}
622 \cvarg{criteria}{Criteria applied to determine when the window search should be finished}
623 \cvarg{comp}{Resultant structure that contains the converged search window coordinates (\texttt{comp->rect} field) and the sum of all of the pixels inside the window (\texttt{comp->area} field)}
626 The function iterates to find the object center
627 given its back projection and initial position of search window. The
628 iterations are made until the search window center moves by less than
629 the given value and/or until the function has done the maximum number
630 of iterations. The function returns the number of iterations made.
632 \cvCPyFunc{MultiplyAcc}
633 Adds the product of two input images to the accumulator.
636 void cvMultiplyAcc( \par const CvArr* image1,\par const CvArr* image2,\par CvArr* acc,\par const CvArr* mask=NULL );
638 \cvdefPy{MultiplyAcc(image1,image2,acc,mask=NULL)-> None}
641 \cvarg{image1}{First input image, 1- or 3-channel, 8-bit or 32-bit floating point (each channel of multi-channel image is processed independently)}
642 \cvarg{image2}{Second input image, the same format as the first one}
643 \cvarg{acc}{Accumulator with the same number of channels as input images, 32-bit or 64-bit floating-point}
644 \cvarg{mask}{Optional operation mask}
647 The function adds the product of 2 images or their selected regions to the accumulator \texttt{acc}:
649 \[ \texttt{acc}(x,y) \leftarrow \texttt{acc}(x,y) + \texttt{image1}(x,y) \cdot \texttt{image2}(x,y) \quad \text{if} \quad \texttt{mask}(x,y) \ne 0 \]
652 \cvCPyFunc{ReleaseConDensation}
653 Deallocates the ConDensation filter structure.
656 void cvReleaseConDensation( CvConDensation** condens );
660 \cvarg{condens}{Pointer to the pointer to the structure to be released}
663 The function releases the structure \texttt{condens}) and frees all memory previously allocated for the structure.
669 \cvCPyFunc{ReleaseKalman}
670 Deallocates the Kalman filter structure.
673 void cvReleaseKalman( \par CvKalman** kalman );
677 \cvarg{kalman}{double pointer to the Kalman filter structure}
680 The function releases the structure \cvCPyCross{CvKalman} and all of the underlying matrices.
684 \cvCPyFunc{RunningAvg}
685 Updates the running average.
688 void cvRunningAvg( \par const CvArr* image,\par CvArr* acc,\par double alpha,\par const CvArr* mask=NULL );
690 \cvdefPy{RunningAvg(image,acc,alpha,mask=NULL)-> None}
693 \cvarg{image}{Input image, 1- or 3-channel, 8-bit or 32-bit floating point (each channel of multi-channel image is processed independently)}
694 \cvarg{acc}{Accumulator with the same number of channels as input image, 32-bit or 64-bit floating-point}
695 \cvarg{alpha}{Weight of input image}
696 \cvarg{mask}{Optional operation mask}
699 The function calculates the weighted sum of the input image
700 \texttt{image} and the accumulator \texttt{acc} so that \texttt{acc}
701 becomes a running average of frame sequence:
703 \[ \texttt{acc}(x,y) \leftarrow (1-\alpha) \cdot \texttt{acc}(x,y) + \alpha \cdot \texttt{image}(x,y) \quad \text{if} \quad \texttt{mask}(x,y) \ne 0 \]
705 where $\alpha$ regulates the update speed (how fast the accumulator forgets about previous frames).
708 \cvCPyFunc{SegmentMotion}
709 Segments a whole motion into separate moving parts.
712 CvSeq* cvSegmentMotion( \par const CvArr* mhi,\par CvArr* seg\_mask,\par CvMemStorage* storage,\par double timestamp,\par double seg\_thresh );
713 }\cvdefPy{SegmentMotion(mhi,seg\_mask,storage,timestamp,seg\_thresh)-> None}
716 \cvarg{mhi}{Motion history image}
717 \cvarg{seg\_mask}{Image where the mask found should be stored, single-channel, 32-bit floating-point}
718 \cvarg{storage}{Memory storage that will contain a sequence of motion connected components}
719 \cvarg{timestamp}{Current time in milliseconds or other units}
720 \cvarg{seg\_thresh}{Segmentation threshold; recommended to be equal to the interval between motion history "steps" or greater}
723 The function finds all of the motion segments and
724 marks them in \texttt{seg\_mask} with individual values (1,2,...). It
725 also returns a sequence of \cvCPyCross{CvConnectedComp}
726 structures, one for each motion component. After that the
727 motion direction for every component can be calculated with
728 \cvCPyCross{CalcGlobalOrientation} using the extracted mask of the particular
729 component \cvCPyCross{Cmp}.
731 \cvCPyFunc{SnakeImage}
732 Changes the contour position to minimize its energy.
735 void cvSnakeImage( \par const IplImage* image,\par CvPoint* points,\par int length,\par float* alpha,\par float* beta,\par float* gamma,\par int coeff\_usage,\par CvSize win,\par CvTermCriteria criteria,\par int calc\_gradient=1 );
736 }\cvdefPy{SnakeImage(image,points,alpha,beta,gamma,coeff\_usage,win,criteria,calc\_gradient=1)-> None}
739 \cvarg{image}{The source image or external energy field}
740 \cvarg{points}{Contour points (snake)}
742 \cvarg{length}{Number of points in the contour}
744 \cvarg{alpha}{Weight[s] of continuity energy, single float or array of \texttt{length} floats, one for each contour point}
745 \cvarg{beta}{Weight[s] of curvature energy, similar to \texttt{alpha}}
746 \cvarg{gamma}{Weight[s] of image energy, similar to \texttt{alpha}}
747 \cvarg{coeff\_usage}{Different uses of the previous three parameters:
749 \cvarg{CV\_VALUE}{indicates that each of \texttt{alpha, beta, gamma} is a pointer to a single value to be used for all points;}
750 \cvarg{CV\_ARRAY}{indicates that each of \texttt{alpha, beta, gamma} is a pointer to an array of coefficients different for all the points of the snake. All the arrays must have the size equal to the contour size.}
752 \cvarg{win}{Size of neighborhood of every point used to search the minimum, both \texttt{win.width} and \texttt{win.height} must be odd}
753 \cvarg{criteria}{Termination criteria}
754 \cvarg{calc\_gradient}{Gradient flag; if not 0, the function calculates the gradient magnitude for every image pixel and consideres it as the energy field, otherwise the input image itself is considered}
757 The function updates the snake in order to minimize its
758 total energy that is a sum of internal energy that depends on the contour
759 shape (the smoother contour is, the smaller internal energy is) and
760 external energy that depends on the energy field and reaches minimum at
761 the local energy extremums that correspond to the image edges in the case
762 of using an image gradient.
764 The parameter \texttt{criteria.epsilon} is used to define the minimal
765 number of points that must be moved during any iteration to keep the
766 iteration process running.
768 If at some iteration the number of moved points is less
769 than \texttt{criteria.epsilon} or the function performed
770 \texttt{criteria.max\_iter} iterations, the function terminates.
772 \cvCPyFunc{SquareAcc}
773 Adds the square of the source image to the accumulator.
776 void cvSquareAcc( \par const CvArr* image,\par CvArr* sqsum,\par const CvArr* mask=NULL );
777 }\cvdefPy{SquareAcc(image,sqsum,mask=NULL)-> None}
780 \cvarg{image}{Input image, 1- or 3-channel, 8-bit or 32-bit floating point (each channel of multi-channel image is processed independently)}
781 \cvarg{sqsum}{Accumulator with the same number of channels as input image, 32-bit or 64-bit floating-point}
782 \cvarg{mask}{Optional operation mask}
785 The function adds the input image \texttt{image} or its selected region, raised to power 2, to the accumulator \texttt{sqsum}:
787 \[ \texttt{sqsum}(x,y) \leftarrow \texttt{sqsum}(x,y) + \texttt{image}(x,y)^2 \quad \text{if} \quad \texttt{mask}(x,y) \ne 0 \]
789 \cvCPyFunc{UpdateMotionHistory}
790 Updates the motion history image by a moving silhouette.
793 void cvUpdateMotionHistory( \par const CvArr* silhouette,\par CvArr* mhi,\par double timestamp,\par double duration );
794 }\cvdefPy{UpdateMotionHistory(silhouette,mhi,timestamp,duration)-> None}
797 \cvarg{silhouette}{Silhouette mask that has non-zero pixels where the motion occurs}
798 \cvarg{mhi}{Motion history image, that is updated by the function (single-channel, 32-bit floating-point)}
799 \cvarg{timestamp}{Current time in milliseconds or other units}
800 \cvarg{duration}{Maximal duration of the motion track in the same units as \texttt{timestamp}}
803 The function updates the motion history image as following:
806 \texttt{mhi}(x,y)=\forkthree
807 {\texttt{timestamp}}{if $\texttt{silhouette}(x,y) \ne 0$}
808 {0}{if $\texttt{silhouette}(x,y) = 0$ and $\texttt{mhi} < (\texttt{timestamp} - \texttt{duration})$}
809 {\texttt{mhi}(x,y)}{otherwise}
811 That is, MHI pixels where motion occurs are set to the current timestamp, while the pixels where motion happened far ago are cleared.
817 \cvCppFunc{accumulate}
818 Adds image to the accumulator.
820 \cvdefCpp{void accumulate( const Mat\& src, Mat\& dst, const Mat\& mask=Mat() );}
822 \cvarg{src}{The input image, 1- or 3-channel, 8-bit or 32-bit floating point}
823 \cvarg{dst}{The accumulator image with the same number of channels as input image, 32-bit or 64-bit floating-point}
824 \cvarg{mask}{Optional operation mask}
827 The function adds \texttt{src}, or some of its elements, to \texttt{dst}:
829 \[ \texttt{dst}(x,y) \leftarrow \texttt{dst}(x,y) + \texttt{src}(x,y) \quad \text{if} \quad \texttt{mask}(x,y) \ne 0 \]
831 The function supports multi-channel images; each channel is processed independently.
833 The functions \texttt{accumulate*} can be used, for example, to collect statistic of background of a scene, viewed by a still camera, for the further foreground-background segmentation.
835 See also: \cvCppCross{accumulateSquare}, \cvCppCross{accumulateProduct}, \cvCppCross{accumulateWeighted}
837 \cvCppFunc{accumulateSquare}
838 Adds the square of the source image to the accumulator.
840 \cvdefCpp{void accumulateSquare( const Mat\& src, Mat\& dst, \par const Mat\& mask=Mat() );}
842 \cvarg{src}{The input image, 1- or 3-channel, 8-bit or 32-bit floating point}
843 \cvarg{dst}{The accumulator image with the same number of channels as input image, 32-bit or 64-bit floating-point}
844 \cvarg{mask}{Optional operation mask}
847 The function adds the input image \texttt{src} or its selected region, raised to power 2, to the accumulator \texttt{dst}:
849 \[ \texttt{dst}(x,y) \leftarrow \texttt{dst}(x,y) + \texttt{src}(x,y)^2 \quad \text{if} \quad \texttt{mask}(x,y) \ne 0 \]
851 The function supports multi-channel images; each channel is processed independently.
853 See also: \cvCppCross{accumulateSquare}, \cvCppCross{accumulateProduct}, \cvCppCross{accumulateWeighted}
855 \cvCppFunc{accumulateProduct}
856 Adds the per-element product of two input images to the accumulator.
858 \cvdefCpp{void accumulateProduct( const Mat\& src1, const Mat\& src2,\par
859 Mat\& dst, const Mat\& mask=Mat() );}
861 \cvarg{src1}{The first input image, 1- or 3-channel, 8-bit or 32-bit floating point}
862 \cvarg{src2}{The second input image of the same type and the same size as \texttt{src1}}
863 \cvarg{dst}{Accumulator with the same number of channels as input images, 32-bit or 64-bit floating-point}
864 \cvarg{mask}{Optional operation mask}
867 The function adds the product of 2 images or their selected regions to the accumulator \texttt{dst}:
869 \[ \texttt{dst}(x,y) \leftarrow \texttt{dst}(x,y) + \texttt{src1}(x,y) \cdot \texttt{src2}(x,y) \quad \text{if} \quad \texttt{mask}(x,y) \ne 0 \]
871 The function supports multi-channel images; each channel is processed independently.
873 See also: \cvCppCross{accumulate}, \cvCppCross{accumulateSquare}, \cvCppCross{accumulateWeighted}
875 \cvCppFunc{accumulateWeighted}
876 Updates the running average.
878 \cvdefCpp{void accumulateWeighted( const Mat\& src, Mat\& dst,\par
879 double alpha, const Mat\& mask=Mat() );}
881 \cvarg{src}{The input image, 1- or 3-channel, 8-bit or 32-bit floating point}
882 \cvarg{dst}{The accumulator image with the same number of channels as input image, 32-bit or 64-bit floating-point}
883 \cvarg{alpha}{Weight of the input image}
884 \cvarg{mask}{Optional operation mask}
887 The function calculates the weighted sum of the input image
888 \texttt{src} and the accumulator \texttt{dst} so that \texttt{dst}
889 becomes a running average of frame sequence:
891 \[ \texttt{dst}(x,y) \leftarrow (1-\texttt{alpha}) \cdot \texttt{dst}(x,y) + \texttt{alpha} \cdot \texttt{src}(x,y) \quad \text{if} \quad \texttt{mask}(x,y) \ne 0 \]
893 that is, \texttt{alpha} regulates the update speed (how fast the accumulator "forgets" about earlier images).
894 The function supports multi-channel images; each channel is processed independently.
896 See also: \cvCppCross{accumulate}, \cvCppCross{accumulateSquare}, \cvCppCross{accumulateProduct}
898 \cvCppFunc{calcOpticalFlowPyrLK}
899 Calculates the optical flow for a sparse feature set using the iterative Lucas-Kanade method with pyramids
901 \cvdefCpp{void calcOpticalFlowPyrLK( const Mat\& prevImg, const Mat\& nextImg,\par
902 const vector<Point2f>\& prevPts, vector<Point2f>\& nextPts,\par
903 vector<uchar>\& status, vector<float>\& err, \par
904 Size winSize=Size(15,15), int maxLevel=3,\par
905 TermCriteria criteria=TermCriteria(\par
906 TermCriteria::COUNT+TermCriteria::EPS, 30, 0.01),\par
907 double derivLambda=0.5, int flags=0 );}
909 \cvarg{prevImg}{The first 8-bit single-channel or 3-channel input image}
910 \cvarg{nextImg}{The second input image of the same size and the same type as \texttt{prevImg}}
911 \cvarg{prevPts}{Vector of points for which the flow needs to be found}
912 \cvarg{nextPts}{The output vector of points containing the calculated new positions of the input features in the second image}
913 \cvarg{status}{The output status vector. Each element of the vector is set to 1 if the flow for the corresponding features has been found, 0 otherwise}
914 \cvarg{err}{The output vector that will contain the difference between patches around the original and moved points}
915 \cvarg{winSize}{Size of the search window at each pyramid level}
916 \cvarg{maxLevel}{0-based maximal pyramid level number. If 0, pyramids are not used (single level), if 1, two levels are used etc.}
917 \cvarg{criteria}{Specifies the termination criteria of the iterative search algorithm (after the specified maximum number of iterations \texttt{criteria.maxCount} or when the search window moves by less than \texttt{criteria.epsilon}}
918 \cvarg{derivLambda}{The relative weight of the spatial image derivatives impact to the optical flow estimation. If \texttt{derivLambda=0}, only the image intensity is used, if \texttt{derivLambda=1}, only derivatives are used. Any other values between 0 and 1 means that both derivatives and the image intensity are used (in the corresponding proportions).}
919 \cvarg{flags}{The operation flags:
921 \cvarg{OPTFLOW\_USE\_INITIAL\_FLOW}{use initial estimations stored in \texttt{nextPts}. If the flag is not set, then initially $\texttt{nextPts}\leftarrow\texttt{prevPts}$}
925 The function implements the sparse iterative version of the Lucas-Kanade optical flow in pyramids, see \cite{Bouguet00}.
927 \cvCppFunc{calcOpticalFlowFarneback}
928 Computes dense optical flow using Gunnar Farneback's algorithm
930 \cvdefCpp{void calcOpticalFlowFarneback( const Mat\& prevImg, const Mat\& nextImg,\par
931 Mat\& flow, double pyrScale, int levels, int winsize,\par
932 int iterations, int polyN, double polySigma, int flags );}
934 \cvarg{prevImg}{The first 8-bit single-channel input image}
935 \cvarg{nextImg}{The second input image of the same size and the same type as \texttt{prevImg}}
936 \cvarg{flow}{The computed flow image; will have the same size as \texttt{prevImg} and type \texttt{CV\_32FC2}}
937 \cvarg{pyrScale}{Specifies the image scale (<1) to build the pyramids for each image. \texttt{pyrScale=0.5} means the classical pyramid, where each next layer is twice smaller than the previous}
938 \cvarg{levels}{The number of pyramid layers, including the initial image. \texttt{levels=1} means that no extra layers are created and only the original images are used}
939 \cvarg{winsize}{The averaging window size; The larger values increase the algorithm robustness to image noise and give more chances for fast motion detection, but yield more blurred motion field}
940 \cvarg{iterations}{The number of iterations the algorithm does at each pyramid level}
941 \cvarg{polyN}{Size of the pixel neighborhood used to find polynomial expansion in each pixel. The larger values mean that the image will be approximated with smoother surfaces, yielding more robust algorithm and more blurred motion field. Typically, \texttt{polyN}=5 or 7}
942 \cvarg{polySigma}{Standard deviation of the Gaussian that is used to smooth derivatives that are used as a basis for the polynomial expansion. For \texttt{polyN=5} you can set \texttt{polySigma=1.1}, for \texttt{polyN=7} a good value would be \texttt{polySigma=1.5}}
943 \cvarg{flags}{The operation flags; can be a combination of the following:
945 \cvarg{OPTFLOW\_USE\_INITIAL\_FLOW}{Use the input \texttt{flow} as the initial flow approximation}
946 \cvarg{OPTFLOW\_FARNEBACK\_GAUSSIAN}{Use a Gaussian $\texttt{winsize}\times\texttt{winsize}$ filter instead of box filter of the same size for optical flow estimation. Usually, this option gives more accurate flow than with a box filter, at the cost of lower speed (and normally \texttt{winsize} for a Gaussian window should be set to a larger value to achieve the same level of robustness)}
950 The function finds optical flow for each \texttt{prevImg} pixel using the alorithm so that
952 \[\texttt{prevImg}(x,y) \sim \texttt{nextImg}(\texttt{flow}(x,y)[0], \texttt{flow}(x,y)[1])\]
955 \cvCppFunc{updateMotionHistory}
956 Updates the motion history image by a moving silhouette.
958 \cvdefCpp{void updateMotionHistory( const Mat\& silhouette, Mat\& mhi,\par
959 double timestamp, double duration );}
961 \cvarg{silhouette}{Silhouette mask that has non-zero pixels where the motion occurs}
962 \cvarg{mhi}{Motion history image, that is updated by the function (single-channel, 32-bit floating-point)}
963 \cvarg{timestamp}{Current time in milliseconds or other units}
964 \cvarg{duration}{Maximal duration of the motion track in the same units as \texttt{timestamp}}
967 The function updates the motion history image as following:
970 \texttt{mhi}(x,y)=\forkthree
971 {\texttt{timestamp}}{if $\texttt{silhouette}(x,y) \ne 0$}
972 {0}{if $\texttt{silhouette}(x,y) = 0$ and $\texttt{mhi} < (\texttt{timestamp} - \texttt{duration})$}
973 {\texttt{mhi}(x,y)}{otherwise}
975 That is, MHI pixels where motion occurs are set to the current \texttt{timestamp}, while the pixels where motion happened last time a long time ago are cleared.
977 The function, together with \cvCppCross{calcMotionGradient} and \cvCppCross{calcGlobalOrientation}, implements the motion templates technique, described in \cite{Davis97} and \cite{Bradski00}.
978 See also the OpenCV sample \texttt{motempl.c} that demonstrates the use of all the motion template functions.
980 \cvCppFunc{calcMotionGradient}
981 Calculates the gradient orientation of a motion history image.
983 \cvdefCpp{void calcMotionGradient( const Mat\& mhi, Mat\& mask,\par
984 Mat\& orientation,\par
985 double delta1, double delta2,\par
986 int apertureSize=3 );}
988 \cvarg{mhi}{Motion history single-channel floating-point image}
989 \cvarg{mask}{The output mask image; will have the type \texttt{CV\_8UC1} and the same size as \texttt{mhi}. Its non-zero elements will mark pixels where the motion gradient data is correct}
990 \cvarg{orientation}{The output motion gradient orientation image; will have the same type and the same size as \texttt{mhi}. Each pixel of it will the motion orientation in degrees, from 0 to 360.}
991 \cvarg{delta1, delta2}{The minimal and maximal allowed difference between \texttt{mhi} values within a pixel neighorhood. That is, the function finds the minimum ($m(x,y)$) and maximum ($M(x,y)$) \texttt{mhi} values over $3 \times 3$ neighborhood of each pixel and marks the motion orientation at $(x, y)$ as valid only if
993 \min(\texttt{delta1} , \texttt{delta2} ) \le M(x,y)-m(x,y) \le \max(\texttt{delta1} ,\texttt{delta2}).
995 \cvarg{apertureSize}{The aperture size of \cvCppCross{Sobel} operator}
998 The function calculates the gradient orientation at each pixel $(x, y)$ as:
1001 \texttt{orientation}(x,y)=\arctan{\frac{d\texttt{mhi}/dy}{d\texttt{mhi}/dx}}
1004 (in fact, \cvCppCross{fastArctan} and \cvCppCross{phase} are used, so that the computed angle is measured in degrees and covers the full range 0..360). Also, the \texttt{mask} is filled to indicate pixels where the computed angle is valid.
1006 \cvCppFunc{calcGlobalOrientation}
1007 Calculates the global motion orientation in some selected region.
1009 \cvdefCpp{double calcGlobalOrientation( const Mat\& orientation, const Mat\& mask,\par
1010 const Mat\& mhi, double timestamp,\par
1013 \cvarg{orientation}{Motion gradient orientation image, calculated by the function \cvCppCross{calcMotionGradient}}
1014 \cvarg{mask}{Mask image. It may be a conjunction of a valid gradient mask, also calculated by \cvCppCross{calcMotionGradient}, and the mask of the region, whose direction needs to be calculated}
1015 \cvarg{mhi}{The motion history image, calculated by \cvCppCross{updateMotionHistory}}
1016 \cvarg{timestamp}{The timestamp passed to \cvCppCross{updateMotionHistory}}
1017 \cvarg{duration}{Maximal duration of motion track in milliseconds, passed to \cvCppCross{updateMotionHistory}}
1020 The function calculates the average
1021 motion direction in the selected region and returns the angle between
1022 0 degrees and 360 degrees. The average direction is computed from
1023 the weighted orientation histogram, where a recent motion has larger
1024 weight and the motion occurred in the past has smaller weight, as recorded in \texttt{mhi}.
1026 \cvCppFunc{CamShift}
1027 Finds the object center, size, and orientation
1029 \cvdefCpp{RotatedRect CamShift( const Mat\& probImage, Rect\& window,\par
1030 TermCriteria criteria );}
1032 \cvarg{probImage}{Back projection of the object histogram; see \cvCppCross{calcBackProject}}
1033 \cvarg{window}{Initial search window}
1034 \cvarg{criteria}{Stop criteria for the underlying \cvCppCross{meanShift}}
1037 The function implements the CAMSHIFT object tracking algrorithm
1039 First, it finds an object center using \cvCppCross{meanShift} and then adjust the window size and finds the optimal rotation. The function returns the rotated rectangle structure that includes the object position, size and the orientation. The next position of the search window can be obtained with \texttt{RotatedRect::boundingRect()}.
1041 See the OpenCV sample \texttt{camshiftdemo.c} that tracks colored objects.
1043 \cvCppFunc{meanShift}
1044 Finds the object on a back projection image.
1046 \cvdefCpp{int meanShift( const Mat\& probImage, Rect\& window,\par
1047 TermCriteria criteria );}
1049 \cvarg{probImage}{Back projection of the object histogram; see \cvCppCross{calcBackProject}}
1050 \cvarg{window}{Initial search window}
1051 \cvarg{criteria}{The stop criteria for the iterative search algorithm}
1054 The function implements iterative object search algorithm. It takes the object back projection on input and the initial position. The mass center in \texttt{window} of the back projection image is computed and the search window center shifts to the mass center. The procedure is repeated until the specified number of iterations \texttt{criteria.maxCount} is done or until the window center shifts by less than \texttt{criteria.epsilon}. The algorithm is used inside \cvCppCross{CamShift} and, unlike \cvCppCross{CamShift}, the search window size or orientation do not change during the search. You can simply pass the output of \cvCppCross{calcBackProject} to this function, but better results can be obtained if you pre-filter the back projection and remove the noise (e.g. by retrieving connected components with \cvCppCross{findContours}, throwing away contours with small area (\cvCppCross{contourArea}) and rendering the remaining contours with \cvCppCross{drawContours})
1057 \cvclass{KalmanFilter}
1064 KalmanFilter();newline
1065 KalmanFilter(int dynamParams, int measureParams, int controlParams=0);newline
1066 void init(int dynamParams, int measureParams, int controlParams=0);newline
1067 // predicts statePre from statePost
1068 const Mat& predict(const Mat& control=Mat());newline
1069 // corrects statePre based on the input measurement vector
1070 // and stores the result to statePost.
1071 const Mat& correct(const Mat& measurement);newline
1073 Mat statePre; // predicted state (x'(k)):
1074 // x(k)=A*x(k-1)+B*u(k)
1075 Mat statePost; // corrected state (x(k)):
1076 // x(k)=x'(k)+K(k)*(z(k)-H*x'(k))
1077 Mat transitionMatrix; // state transition matrix (A)
1078 Mat controlMatrix; // control matrix (B)
1079 // (it is not used if there is no control)
1080 Mat measurementMatrix; // measurement matrix (H)
1081 Mat processNoiseCov; // process noise covariance matrix (Q)
1082 Mat measurementNoiseCov;// measurement noise covariance matrix (R)
1083 Mat errorCovPre; // priori error estimate covariance matrix (P'(k)):
1084 // P'(k)=A*P(k-1)*At + Q)*/
1085 Mat gain; // Kalman gain matrix (K(k)):
1086 // K(k)=P'(k)*Ht*inv(H*P'(k)*Ht+R)
1087 Mat errorCovPost; // posteriori error estimate covariance matrix (P(k)):
1088 // P(k)=(I-K(k)*H)*P'(k)
1093 The class implements standard Kalman filter \url{http://en.wikipedia.org/wiki/Kalman_filter}. However, you can modify \texttt{transitionMatrix}, \texttt{controlMatrix} and \texttt{measurementMatrix} to get the extended Kalman filter functionality. See the OpenCV sample \texttt{kalman.c}