PDL::OpenCV::Tracking - PDL bindings for OpenCV DISOpticalFlow, DenseOpticalFlow, FarnebackOpticalFlow, KalmanFilter, SparseOpticalFlow, SparsePyrLKOpticalFlow, Tracker, TrackerCSRT, TrackerDaSiamRPN, TrackerGOTURN, TrackerKCF, TrackerMIL, VariationalRefinement
use PDL::OpenCV::Tracking;
Signature: ([phys] probImage(l1,c1,r1); indx [io,phys] window(n2=4); TermCriteriaWrapper * criteria; [o] RotatedRectWrapper * res)
Finds an object center, size, and orientation.
$res = CamShift($probImage,$window,$criteria);
@cite Bradski98 . First, it finds an object center using meanShift and then adjusts the window size and finds the optimal rotation. The function returns the rotated rectangle structure that includes the object position, size, and CV_WRAP orientation. The next position of the search window can be obtained with RotatedRect::boundingRect() See the OpenCV sample camshiftdemo.c that tracks colored objects. @note - (Python) A sample explaining the camshift tracking algorithm can be found at opencv_source_code/samples/python/camshift.py
Parameters:
Back projection of the object histogram. See calcBackProject.
Initial search window.
Stop criteria for the underlying meanShift. returns (in old interfaces) Number of iterations CAMSHIFT took to converge The function implements the CAMSHIFT object tracking algorithm
CamShift ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Signature: ([phys] probImage(l1,c1,r1); indx [io,phys] window(n2=4); int [o,phys] res(); TermCriteriaWrapper * criteria)
Finds an object on a back projection image.
$res = meanShift($probImage,$window,$criteria);
Back projection of the object histogram. See calcBackProject for details.
Stop criteria for the iterative search algorithm. returns : Number of iterations CAMSHIFT took to converge. The function implements the iterative object search algorithm. It takes the input back projection of an object and the initial position. The mass center in 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 criteria.maxCount is done or until the window center shifts by less than criteria.epsilon. The algorithm is used inside CamShift and, unlike CamShift , the search window size or orientation do not change during the search. You can simply pass the output of calcBackProject to this function. But better results can be obtained if you pre-filter the back projection and remove the noise. For example, you can do this by retrieving connected components with findContours , throwing away contours with small area ( contourArea ), and rendering the remaining contours with drawContours.
meanShift ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Signature: ([phys] img(l1,c1,r1); indx [phys] winSize(n3=2); int [phys] maxLevel(); byte [phys] withDerivatives(); int [phys] pyrBorder(); int [phys] derivBorder(); byte [phys] tryReuseInputImage(); int [o,phys] res(); [o] vector_MatWrapper * pyramid)
Constructs the image pyramid which can be passed to calcOpticalFlowPyrLK.
($pyramid,$res) = buildOpticalFlowPyramid($img,$winSize,$maxLevel); # with defaults ($pyramid,$res) = buildOpticalFlowPyramid($img,$winSize,$maxLevel,$withDerivatives,$pyrBorder,$derivBorder,$tryReuseInputImage);
8-bit input image.
output pyramid.
window size of optical flow algorithm. Must be not less than winSize argument of calcOpticalFlowPyrLK. It is needed to calculate required padding for pyramid levels.
0-based maximal pyramid level number.
set to precompute gradients for the every pyramid level. If pyramid is constructed without the gradients then calcOpticalFlowPyrLK will calculate them internally.
the border mode for pyramid layers.
the border mode for gradients.
put ROI of input image into the pyramid if possible. You can pass false to force data copying.
Returns: number of levels in constructed pyramid. Can be less than maxLevel.
buildOpticalFlowPyramid ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Signature: ([phys] prevImg(l1,c1,r1); [phys] nextImg(l2,c2,r2); [phys] prevPts(l3,c3,r3); [io,phys] nextPts(l4,c4,r4); [o,phys] status(l5,c5,r5); [o,phys] err(l6,c6,r6); indx [phys] winSize(n7=2); int [phys] maxLevel(); int [phys] flags(); double [phys] minEigThreshold(); TermCriteriaWrapper * criteria)
Calculates an optical flow for a sparse feature set using the iterative Lucas-Kanade method with pyramids. NO BROADCASTING.
($status,$err) = calcOpticalFlowPyrLK($prevImg,$nextImg,$prevPts,$nextPts); # with defaults ($status,$err) = calcOpticalFlowPyrLK($prevImg,$nextImg,$prevPts,$nextPts,$winSize,$maxLevel,$criteria,$flags,$minEigThreshold);
@cite Bouguet00), divided by number of pixels in a window; if this value is less than minEigThreshold, then a corresponding feature is filtered out and its flow is not processed, so it allows to remove bad points and get a performance boost. The function implements a sparse iterative version of the Lucas-Kanade optical flow in pyramids. See @cite Bouguet00 . The function is parallelized with the TBB library. @note - An example using the Lucas-Kanade optical flow algorithm can be found at opencv_source_code/samples/cpp/lkdemo.cpp - (Python) An example using the Lucas-Kanade optical flow algorithm can be found at opencv_source_code/samples/python/lk_track.py - (Python) An example using the Lucas-Kanade tracker for homography matching can be found at opencv_source_code/samples/python/lk_homography.py
first 8-bit input image or pyramid constructed by buildOpticalFlowPyramid.
second input image or pyramid of the same size and the same type as prevImg.
vector of 2D points for which the flow needs to be found; point coordinates must be single-precision floating-point numbers.
output vector of 2D points (with single-precision floating-point coordinates) containing the calculated new positions of input features in the second image; when OPTFLOW_USE_INITIAL_FLOW flag is passed, the vector must have the same size as in the input.
output status vector (of unsigned chars); each element of the vector is set to 1 if the flow for the corresponding features has been found, otherwise, it is set to 0.
output vector of errors; each element of the vector is set to an error for the corresponding feature, type of the error measure can be set in flags parameter; if the flow wasn't found then the error is not defined (use the status parameter to find such cases).
size of the search window at each pyramid level.
0-based maximal pyramid level number; if set to 0, pyramids are not used (single level), if set to 1, two levels are used, and so on; if pyramids are passed to input then algorithm will use as many levels as pyramids have but no more than maxLevel.
parameter, specifying the termination criteria of the iterative search algorithm (after the specified maximum number of iterations criteria.maxCount or when the search window moves by less than criteria.epsilon.
operation flags: - **OPTFLOW_USE_INITIAL_FLOW** uses initial estimations, stored in nextPts; if the flag is not set, then prevPts is copied to nextPts and is considered the initial estimate. - **OPTFLOW_LK_GET_MIN_EIGENVALS** use minimum eigen values as an error measure (see minEigThreshold description); if the flag is not set, then L1 distance between patches around the original and a moved point, divided by number of pixels in a window, is used as a error measure.
the algorithm calculates the minimum eigen value of a 2x2 normal matrix of optical flow equations (this matrix is called a spatial gradient matrix in
calcOpticalFlowPyrLK ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Signature: ([phys] prev(l1,c1,r1); [phys] next(l2,c2,r2); [io,phys] flow(l3,c3,r3); double [phys] pyr_scale(); int [phys] levels(); int [phys] winsize(); int [phys] iterations(); int [phys] poly_n(); double [phys] poly_sigma(); int [phys] flags())
Computes a dense optical flow using the Gunnar Farneback's algorithm.
calcOpticalFlowFarneback($prev,$next,$flow,$pyr_scale,$levels,$winsize,$iterations,$poly_n,$poly_sigma,$flags);
\<1) to build pyramids for each image; pyr_scale=0.5 means a classical pyramid, where each next layer is twice smaller than the previous one. \texttt{winsize}\times\texttt{winsize}filter instead of a box filter of the same size for optical flow estimation; usually, this option gives z more accurate flow than with a box filter, at the cost of lower speed; normally, winsize for a Gaussian window should be set to a larger value to achieve the same level of robustness. The function finds an optical flow for each prev pixel using the @cite Farneback2003 algorithm so that \f[\texttt{prev} (y,x) \sim \texttt{next} ( y + \texttt{flow} (y,x)[1], x + \texttt{flow} (y,x)[0])\f] @note - An example using the optical flow algorithm described by Gunnar Farneback can be found at opencv_source_code/samples/cpp/fback.cpp - (Python) An example using the optical flow algorithm described by Gunnar Farneback can be found at opencv_source_code/samples/python/opt_flow.py
\texttt{winsize}\times\texttt{winsize}
first 8-bit single-channel input image.
second input image of the same size and the same type as prev.
computed flow image that has the same size as prev and type CV_32FC2.
parameter, specifying the image scale (
number of pyramid layers including the initial image; levels=1 means that no extra layers are created and only the original images are used.
averaging window size; larger values increase the algorithm robustness to image noise and give more chances for fast motion detection, but yield more blurred motion field.
number of iterations the algorithm does at each pyramid level.
size of the pixel neighborhood used to find polynomial expansion in each pixel; larger values mean that the image will be approximated with smoother surfaces, yielding more robust algorithm and more blurred motion field, typically poly_n =5 or 7.
standard deviation of the Gaussian that is used to smooth derivatives used as a basis for the polynomial expansion; for poly_n=5, you can set poly_sigma=1.1, for poly_n=7, a good value would be poly_sigma=1.5.
operation flags that can be a combination of the following: - **OPTFLOW_USE_INITIAL_FLOW** uses the input flow as an initial flow approximation. - **OPTFLOW_FARNEBACK_GAUSSIAN** uses the Gaussian
calcOpticalFlowFarneback ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Signature: ([phys] templateImage(l1,c1,r1); [phys] inputImage(l2,c2,r2); [phys] inputMask(l3,c3,r3); double [o,phys] res())
Computes the Enhanced Correlation Coefficient value between two images
$res = computeECC($templateImage,$inputImage); # with defaults $res = computeECC($templateImage,$inputImage,$inputMask);
@cite EP08 .
single-channel template image; CV_8U or CV_32F array.
single-channel input image to be warped to provide an image similar to templateImage, same type as templateImage.
An optional mask to indicate valid values of inputImage.
See also: findTransformECC
computeECC ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Signature: ([phys] templateImage(l1,c1,r1); [phys] inputImage(l2,c2,r2); [io,phys] warpMatrix(l3,c3,r3); int [phys] motionType(); [phys] inputMask(l6,c6,r6); int [phys] gaussFiltSize(); double [o,phys] res(); TermCriteriaWrapper * criteria)
Finds the geometric transform (warp) between two images in terms of the ECC criterion
$res = findTransformECC($templateImage,$inputImage,$warpMatrix,$motionType,$criteria,$inputMask,$gaussFiltSize);
@cite EP08 . 2\times 3or 3\times 3mapping matrix (warp). 2\times 3with the first 2\times 2part being the unity matrix and the rest two parameters being estimated. - **MOTION_EUCLIDEAN** sets a Euclidean (rigid) transformation as motion model; three parameters are estimated; warpMatrix is 2\times 3. - **MOTION_AFFINE** sets an affine motion model (DEFAULT); six parameters are estimated; warpMatrix is 2\times 3. - **MOTION_HOMOGRAPHY** sets a homography as a motion model; eight parameters are estimated;\`warpMatrix\` is 3\times 3. The function estimates the optimum transformation (warpMatrix) with respect to ECC criterion (@cite EP08), that is \f[\texttt{warpMatrix} = \arg\max_{W} \texttt{ECC}(\texttt{templateImage}(x,y),\texttt{inputImage}(x',y'))\f] where \f[\begin{bmatrix} x' \\ y' \end{bmatrix} = W \cdot \begin{bmatrix} x \\ y \\ 1 \end{bmatrix}\f] (the equation holds with homogeneous coordinates for homography). It returns the final enhanced correlation coefficient, that is the correlation coefficient between the template image and the final warped input image. When a 3\times 3matrix is given with motionType =0, 1 or 2, the third row is ignored. Unlike findHomography and estimateRigidTransform, the function findTransformECC implements an area-based alignment that builds on intensity similarities. In essence, the function updates the initial transformation that roughly aligns the images. If this information is missing, the identity warp (unity matrix) is used as an initialization. Note that if images undergo strong displacements/rotations, an initial transformation that roughly aligns the images is necessary (e.g., a simple euclidean/similarity transform that allows for the images showing the same image content approximately). Use inverse warping in the second image to take an image close to the first one, i.e. use the flag WARP_INVERSE_MAP with warpAffine or warpPerspective. See also the OpenCV sample image_alignment.cpp that demonstrates the use of the function. Note that the function throws an exception if algorithm does not converges.
2\times 3
3\times 3
2\times 2
single-channel input image which should be warped with the final warpMatrix in order to provide an image similar to templateImage, same type as templateImage.
floating-point
parameter, specifying the type of motion: - **MOTION_TRANSLATION** sets a translational motion model; warpMatrix is
parameter, specifying the termination criteria of the ECC algorithm; criteria.epsilon defines the threshold of the increment in the correlation coefficient between two iterations (a negative criteria.epsilon makes criteria.maxcount the only termination criterion). Default values are shown in the declaration above.
An optional value indicating size of gaussian blur filter; (DEFAULT: 5)
See also: computeECC, estimateAffine2D, estimateAffinePartial2D, findHomography
findTransformECC ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Signature: ([phys] templateImage(l1,c1,r1); [phys] inputImage(l2,c2,r2); [io,phys] warpMatrix(l3,c3,r3); int [phys] motionType(); [phys] inputMask(l6,c6,r6); double [o,phys] res(); TermCriteriaWrapper * criteria)
$res = findTransformECC2($templateImage,$inputImage,$warpMatrix); # with defaults $res = findTransformECC2($templateImage,$inputImage,$warpMatrix,$motionType,$criteria,$inputMask);
@overload
findTransformECC2 ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Signature: ([o,phys] res(l2,c2,r2); StringWrapper* path)
Read a .flo file NO BROADCASTING.
$res = readOpticalFlow($path);
The function readOpticalFlow loads a flow field from a file and returns it as a single matrix. Resulting Mat has a type CV_32FC2 - floating-point, 2-channel. First channel corresponds to the flow in the horizontal direction (u), second - vertical (v).
Path to the file to be loaded
readOpticalFlow ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Signature: ([phys] flow(l2,c2,r2); byte [o,phys] res(); StringWrapper* path)
Write a .flo to disk
$res = writeOpticalFlow($path,$flow);
The function stores a flow field in a file, returns true on success, false otherwise. The flow field must be a 2-channel, floating-point matrix (CV_32FC2). First channel corresponds to the flow in the horizontal direction (u), second - vertical (v).
Path to the file to be written
Flow field to be stored
writeOpticalFlow ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
DIS optical flow algorithm.
This class implements the Dense Inverse Search (DIS) optical flow algorithm. More details about the algorithm can be found at @cite Kroeger2016 . Includes three presets with preselected parameters to provide reasonable trade-off between speed and quality. However, even the slowest preset is still relatively fast, use DeepFlow if you need better quality and don't care about speed. This implementation includes several additional features compared to the algorithm described in the paper, including spatial propagation of flow vectors (@ref getUseSpatialPropagation), as well as an option to utilize an initial flow approximation passed to @ref calc (which is, essentially, temporal propagation, if the previous frame's flow field is passed).
Subclass of PDL::OpenCV::DenseOpticalFlow
Creates an instance of DISOpticalFlow
$obj = PDL::OpenCV::DISOpticalFlow->new; # with defaults $obj = PDL::OpenCV::DISOpticalFlow->new($preset);
one of PRESET_ULTRAFAST, PRESET_FAST and PRESET_MEDIUM
Finest level of the Gaussian pyramid on which the flow is computed (zero level corresponds to the original image resolution). The final flow is obtained by bilinear upscaling.
$res = $obj->getFinestScale;
See also: setFinestScale
$obj->setFinestScale($val);
@copybrief getFinestScale See also: getFinestScale
Size of an image patch for matching (in pixels). Normally, default 8x8 patches work well enough in most cases.
$res = $obj->getPatchSize;
See also: setPatchSize
$obj->setPatchSize($val);
@copybrief getPatchSize See also: getPatchSize
Stride between neighbor patches. Must be less than patch size. Lower values correspond to higher flow quality.
$res = $obj->getPatchStride;
See also: setPatchStride
$obj->setPatchStride($val);
@copybrief getPatchStride See also: getPatchStride
Maximum number of gradient descent iterations in the patch inverse search stage. Higher values may improve quality in some cases.
$res = $obj->getGradientDescentIterations;
See also: setGradientDescentIterations
$obj->setGradientDescentIterations($val);
@copybrief getGradientDescentIterations See also: getGradientDescentIterations
Number of fixed point iterations of variational refinement per scale. Set to zero to disable variational refinement completely. Higher values will typically result in more smooth and high-quality flow.
$res = $obj->getVariationalRefinementIterations;
$obj->setVariationalRefinementIterations($val);
Weight of the smoothness term
$res = $obj->getVariationalRefinementAlpha;
See also: setVariationalRefinementAlpha
$obj->setVariationalRefinementAlpha($val);
@copybrief getVariationalRefinementAlpha See also: getVariationalRefinementAlpha
Weight of the color constancy term
$res = $obj->getVariationalRefinementDelta;
See also: setVariationalRefinementDelta
$obj->setVariationalRefinementDelta($val);
@copybrief getVariationalRefinementDelta See also: getVariationalRefinementDelta
Weight of the gradient constancy term
$res = $obj->getVariationalRefinementGamma;
See also: setVariationalRefinementGamma
$obj->setVariationalRefinementGamma($val);
@copybrief getVariationalRefinementGamma See also: getVariationalRefinementGamma
Whether to use mean-normalization of patches when computing patch distance. It is turned on by default as it typically provides a noticeable quality boost because of increased robustness to illumination variations. Turn it off if you are certain that your sequence doesn't contain any changes in illumination.
$res = $obj->getUseMeanNormalization;
See also: setUseMeanNormalization
$obj->setUseMeanNormalization($val);
@copybrief getUseMeanNormalization See also: getUseMeanNormalization
Whether to use spatial propagation of good optical flow vectors. This option is turned on by default, as it tends to work better on average and can sometimes help recover from major errors introduced by the coarse-to-fine scheme employed by the DIS optical flow algorithm. Turning this option off can make the output flow field a bit smoother, however.
$res = $obj->getUseSpatialPropagation;
See also: setUseSpatialPropagation
$obj->setUseSpatialPropagation($val);
@copybrief getUseSpatialPropagation See also: getUseSpatialPropagation
Base class for dense optical flow algorithms
Subclass of PDL::OpenCV::Algorithm
Signature: ([phys] I0(l2,c2,r2); [phys] I1(l3,c3,r3); [io,phys] flow(l4,c4,r4); DenseOpticalFlowWrapper * self)
Calculates an optical flow.
$obj->calc($I0,$I1,$flow);
DenseOpticalFlow_calc ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Releases all inner buffers.
$obj->collectGarbage;
Class computing a dense optical flow using the Gunnar Farneback's algorithm.
$obj = PDL::OpenCV::FarnebackOpticalFlow->new; # with defaults $obj = PDL::OpenCV::FarnebackOpticalFlow->new($numLevels,$pyrScale,$fastPyramids,$winSize,$numIters,$polyN,$polySigma,$flags);
$res = $obj->getNumLevels;
$obj->setNumLevels($numLevels);
$res = $obj->getPyrScale;
$obj->setPyrScale($pyrScale);
$res = $obj->getFastPyramids;
$obj->setFastPyramids($fastPyramids);
$res = $obj->getWinSize;
$obj->setWinSize($winSize);
$res = $obj->getNumIters;
$obj->setNumIters($numIters);
$res = $obj->getPolyN;
$obj->setPolyN($polyN);
$res = $obj->getPolySigma;
$obj->setPolySigma($polySigma);
$res = $obj->getFlags;
$obj->setFlags($flags);
Kalman filter class.
The class implements a standard Kalman filter <http://en.wikipedia.org/wiki/Kalman_filter>, @cite Welch95 . However, you can modify transitionMatrix, controlMatrix, and measurementMatrix to get an extended Kalman filter functionality. @note In C API when CvKalman* kalmanFilter structure is not needed anymore, it should be released with cvReleaseKalman(&kalmanFilter)
$obj = PDL::OpenCV::KalmanFilter->new;
$obj = PDL::OpenCV::KalmanFilter->new2($dynamParams,$measureParams); # with defaults $obj = PDL::OpenCV::KalmanFilter->new2($dynamParams,$measureParams,$controlParams,$type);
Dimensionality of the state.
Dimensionality of the measurement.
Dimensionality of the control vector.
Type of the created matrices that should be CV_32F or CV_64F.
Signature: ([phys] control(l2,c2,r2); [o,phys] res(l3,c3,r3); KalmanFilterWrapper * self)
Computes a predicted state. NO BROADCASTING.
$res = $obj->predict; # with defaults $res = $obj->predict($control);
The optional input control
KalmanFilter_predict ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Signature: ([phys] measurement(l2,c2,r2); [o,phys] res(l3,c3,r3); KalmanFilterWrapper * self)
Updates the predicted state from the measurement. NO BROADCASTING.
$res = $obj->correct($measurement);
The measured system parameters
KalmanFilter_correct ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Base interface for sparse optical flow algorithms.
Signature: ([phys] prevImg(l2,c2,r2); [phys] nextImg(l3,c3,r3); [phys] prevPts(l4,c4,r4); [io,phys] nextPts(l5,c5,r5); [o,phys] status(l6,c6,r6); [o,phys] err(l7,c7,r7); SparseOpticalFlowWrapper * self)
Calculates a sparse optical flow. NO BROADCASTING.
($status,$err) = $obj->calc($prevImg,$nextImg,$prevPts,$nextPts);
First input image.
Second input image of the same size and the same type as prevImg.
Vector of 2D points for which the flow needs to be found.
Output vector of 2D points containing the calculated new positions of input features in the second image.
Output status vector. Each element of the vector is set to 1 if the flow for the corresponding features has been found. Otherwise, it is set to 0.
Optional output vector that contains error response for each point (inverse confidence).
SparseOpticalFlow_calc ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Class used for calculating a sparse optical flow.
The class can calculate an optical flow for a sparse feature set using the iterative Lucas-Kanade method with pyramids. See also: calcOpticalFlowPyrLK
Subclass of PDL::OpenCV::SparseOpticalFlow
Signature: (indx [phys] winSize(n2=2); int [phys] maxLevel(); int [phys] flags(); double [phys] minEigThreshold(); char * klass; TermCriteriaWrapper * crit; [o] SparsePyrLKOpticalFlowWrapper * res)
$obj = PDL::OpenCV::SparsePyrLKOpticalFlow->new; # with defaults $obj = PDL::OpenCV::SparsePyrLKOpticalFlow->new($winSize,$maxLevel,$crit,$flags,$minEigThreshold);
SparsePyrLKOpticalFlow_new ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Signature: (indx [o,phys] res(n2=2); SparsePyrLKOpticalFlowWrapper * self)
SparsePyrLKOpticalFlow_getWinSize ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Signature: (indx [phys] winSize(n2=2); SparsePyrLKOpticalFlowWrapper * self)
SparsePyrLKOpticalFlow_setWinSize ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
$res = $obj->getMaxLevel;
$obj->setMaxLevel($maxLevel);
$res = $obj->getTermCriteria;
$obj->setTermCriteria($crit);
$res = $obj->getMinEigThreshold;
$obj->setMinEigThreshold($minEigThreshold);
Base abstract class for the long-term tracker
Signature: ([phys] image(l2,c2,r2); indx [phys] boundingBox(n3=4); TrackerWrapper * self)
Initialize the tracker with a known bounding box that surrounded the target
$obj->init($image,$boundingBox);
The initial frame
The initial bounding box
Tracker_init ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Signature: ([phys] image(l2,c2,r2); indx [o,phys] boundingBox(n3=4); byte [o,phys] res(); TrackerWrapper * self)
Update the tracker, find the new most likely bounding box for the target
($boundingBox,$res) = $obj->update($image);
The current frame
The bounding box that represent the new target location, if true was returned, not modified otherwise
Returns: True means that target was located and false means that tracker cannot locate target in current frame. Note, that latter *does not* imply that tracker has failed, maybe target is indeed missing from the frame (say, out of sight)
Tracker_update ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
the CSRT tracker
The implementation is based on @cite Lukezic_IJCV2018 Discriminative Correlation Filter with Channel and Spatial Reliability
Subclass of PDL::OpenCV::Tracker
Create CSRT tracker instance
$obj = PDL::OpenCV::TrackerCSRT->new;
CSRT parameters TrackerCSRT::Params
Signature: ([phys] mask(l2,c2,r2); TrackerCSRTWrapper * self)
$obj->setInitialMask($mask);
TrackerCSRT_setInitialMask ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Constructor
$obj = PDL::OpenCV::TrackerDaSiamRPN->new;
DaSiamRPN parameters TrackerDaSiamRPN::Params
Return tracking score
$res = $obj->getTrackingScore;
the GOTURN (Generic Object Tracking Using Regression Networks) tracker * * GOTURN (
@cite GOTURN) is kind of trackers based on Convolutional Neural Networks (CNN). While taking all advantages of CNN trackers, * GOTURN is much faster due to offline training without online fine-tuning nature. * GOTURN tracker addresses the problem of single target tracking: given a bounding box label of an object in the first frame of the video, * we track that object through the rest of the video. NOTE: Current method of GOTURN does not handle occlusions; however, it is fairly * robust to viewpoint changes, lighting changes, and deformations. * Inputs of GOTURN are two RGB patches representing Target and Search patches resized to 227x227. * Outputs of GOTURN are predicted bounding box coordinates, relative to Search patch coordinate system, in format X1,Y1,X2,Y2. * Original paper is here: <http://davheld.github.io/GOTURN/GOTURN.pdf> * As long as original authors implementation: <https://github.com/davheld/GOTURN#train-the-tracker> * Implementation of training algorithm is placed in separately here due to 3d-party dependencies: * <https://github.com/Auron-X/GOTURN_Training_Toolkit> * GOTURN architecture goturn.prototxt and trained model goturn.caffemodel are accessible on opencv_extra GitHub repository.
$obj = PDL::OpenCV::TrackerGOTURN->new;
GOTURN parameters TrackerGOTURN::Params
the KCF (Kernelized Correlation Filter) tracker
* KCF is a novel tracking framework that utilizes properties of circulant matrix to enhance the processing speed. * This tracking method is an implementation of @cite KCF_ECCV which is extended to KCF with color-names features (@cite KCF_CN). * The original paper of KCF is available at <http://www.robots.ox.ac.uk/~joao/publications/henriques_tpami2015.pdf> * as well as the matlab implementation. For more information about KCF with color-names features, please refer to * <http://www.cvl.isy.liu.se/research/objrec/visualtracking/colvistrack/index.html>.
Create KCF tracker instance
$obj = PDL::OpenCV::TrackerKCF->new;
KCF parameters TrackerKCF::Params
The MIL algorithm trains a classifier in an online manner to separate the object from the background.
Multiple Instance Learning avoids the drift problem for a robust tracking. The implementation is based on @cite MIL . Original code can be found here <http://vision.ucsd.edu/~bbabenko/project_miltrack.shtml>
Create MIL tracker instance *
$obj = PDL::OpenCV::TrackerMIL->new;
MIL parameters TrackerMIL::Params
Variational optical flow refinement
This class implements variational refinement of the input flow field, i.e. it uses input flow to initialize the minimization of the following functional: E(U) = \int_{\Omega} \delta \Psi(E_I) + \gamma \Psi(E_G) + \alpha \Psi(E_S), where E_I,E_G,E_Sare color constancy, gradient constancy and smoothness terms respectively. \Psi(s^2)=\sqrt{s^2+\epsilon^2}is a robust penalizer to limit the influence of outliers. A complete formulation and a description of the minimization procedure can be found in @cite Brox2004
E(U) = \int_{\Omega} \delta \Psi(E_I) + \gamma \Psi(E_G) + \alpha \Psi(E_S)
E_I,E_G,E_S
\Psi(s^2)=\sqrt{s^2+\epsilon^2}
Creates an instance of VariationalRefinement
$obj = PDL::OpenCV::VariationalRefinement->new;
Signature: ([phys] I0(l2,c2,r2); [phys] I1(l3,c3,r3); [io,phys] flow_u(l4,c4,r4); [io,phys] flow_v(l5,c5,r5); VariationalRefinementWrapper * self)
$obj->calcUV($I0,$I1,$flow_u,$flow_v);
@ref calc function overload to handle separate horizontal (u) and vertical (v) flow components (to avoid extra splits/merges)
VariationalRefinement_calcUV ignores the bad-value flag of the input ndarrays. It will set the bad-value flag of all output ndarrays if the flag is set for any of the input ndarrays.
Number of outer (fixed-point) iterations in the minimization procedure.
$res = $obj->getFixedPointIterations;
See also: setFixedPointIterations
$obj->setFixedPointIterations($val);
@copybrief getFixedPointIterations See also: getFixedPointIterations
Number of inner successive over-relaxation (SOR) iterations in the minimization procedure to solve the respective linear system.
$res = $obj->getSorIterations;
See also: setSorIterations
$obj->setSorIterations($val);
@copybrief getSorIterations See also: getSorIterations
Relaxation factor in SOR
$res = $obj->getOmega;
See also: setOmega
$obj->setOmega($val);
@copybrief getOmega See also: getOmega
$res = $obj->getAlpha;
See also: setAlpha
$obj->setAlpha($val);
@copybrief getAlpha See also: getAlpha
$res = $obj->getDelta;
See also: setDelta
$obj->setDelta($val);
@copybrief getDelta See also: getDelta
$res = $obj->getGamma;
See also: setGamma
$obj->setGamma($val);
@copybrief getGamma See also: getGamma
To install PDL::OpenCV, copy and paste the appropriate command in to your terminal.
cpanm
cpanm PDL::OpenCV
CPAN shell
perl -MCPAN -e shell install PDL::OpenCV
For more information on module installation, please visit the detailed CPAN module installation guide.