class cv::DualTVL1OpticalFlow

Overview

“Dual TV L1” Optical Flow Algorithm. More…

#include <tracking.hpp>

class DualTVL1OpticalFlow: public cv::DenseOpticalFlow
{
public:
    // methods

    virtual
    double
    getEpsilon() const = 0;

    virtual
    double
    getGamma() const = 0;

    virtual
    int
    getInnerIterations() const = 0;

    virtual
    double
    getLambda() const = 0;

    virtual
    int
    getMedianFiltering() const = 0;

    virtual
    int
    getOuterIterations() const = 0;

    virtual
    int
    getScalesNumber() const = 0;

    virtual
    double
    getScaleStep() const = 0;

    virtual
    double
    getTau() const = 0;

    virtual
    double
    getTheta() const = 0;

    virtual
    bool
    getUseInitialFlow() const = 0;

    virtual
    int
    getWarpingsNumber() const = 0;

    virtual
    void
    setEpsilon(double val) = 0;

    virtual
    void
    setGamma(double val) = 0;

    virtual
    void
    setInnerIterations(int val) = 0;

    virtual
    void
    setLambda(double val) = 0;

    virtual
    void
    setMedianFiltering(int val) = 0;

    virtual
    void
    setOuterIterations(int val) = 0;

    virtual
    void
    setScalesNumber(int val) = 0;

    virtual
    void
    setScaleStep(double val) = 0;

    virtual
    void
    setTau(double val) = 0;

    virtual
    void
    setTheta(double val) = 0;

    virtual
    void
    setUseInitialFlow(bool val) = 0;

    virtual
    void
    setWarpingsNumber(int val) = 0;

    static
    Ptr<DualTVL1OpticalFlow>
    create(
        double tau = 0.25,
        double lambda = 0.15,
        double theta = 0.3,
        int nscales = 5,
        int warps = 5,
        double epsilon = 0.01,
        int innnerIterations = 30,
        int outerIterations = 10,
        double scaleStep = 0.8,
        double gamma = 0.0,
        int medianFiltering = 5,
        bool useInitialFlow = false
        );
};

Inherited Members

public:
    // methods

    virtual
    void
    clear();

    virtual
    bool
    empty() const;

    virtual
    String
    getDefaultName() const;

    virtual
    void
    read(const FileNode& fn);

    virtual
    void
    save(const String& filename) const;

    virtual
    void
    write(FileStorage& fs) const;

    template <typename _Tp>
    static
    Ptr<_Tp>
    load(
        const String& filename,
        const String& objname = String()
        );

    template <typename _Tp>
    static
    Ptr<_Tp>
    loadFromString(
        const String& strModel,
        const String& objname = String()
        );

    template <typename _Tp>
    static
    Ptr<_Tp>
    read(const FileNode& fn);

    virtual
    void
    calc(
        InputArray I0,
        InputArray I1,
        InputOutputArray flow
        ) = 0;

    virtual
    void
    collectGarbage() = 0;

protected:
    // methods

    void
    writeFormat(FileStorage& fs) const;

Detailed Documentation

“Dual TV L1” Optical Flow Algorithm.

The class implements the “Dual TV L1” optical flow algorithm described in [95] and [75]. Here are important members of the class that control the algorithm, which you can set after constructing the class instance:

  • member double tau Time step of the numerical scheme.
  • member double lambda Weight parameter for the data term, attachment parameter. This is the most relevant parameter, which determines the smoothness of the output. The smaller this parameter is, the smoother the solutions we obtain. It depends on the range of motions of the images, so its value should be adapted to each image sequence.
  • member double theta Weight parameter for (u - v)^2, tightness parameter. It serves as a link between the attachment and the regularization terms. In theory, it should have a small value in order to maintain both parts in correspondence. The method is stable for a large range of values of this parameter.
  • member int nscales Number of scales used to create the pyramid of images.
  • member int warps Number of warpings per scale. Represents the number of times that I1(x+u0) and grad( I1(x+u0) ) are computed per scale. This is a parameter that assures the stability of the method. It also affects the running time, so it is a compromise between speed and accuracy.
  • member double epsilon Stopping criterion threshold used in the numerical scheme, which is a trade-off between precision and running time. A small value will yield more accurate solutions at the expense of a slower convergence.
  • member int iterations Stopping criterion iterations number used in the numerical scheme.
  1. Zach, T. Pock and H. Bischof, “A Duality Based Approach for Realtime TV-L1 Optical Flow”. Javier Sanchez, Enric Meinhardt-Llopis and Gabriele Facciolo. “TV-L1 Optical Flow Estimation”.

Methods

virtual
double
getEpsilon() const = 0

Stopping criterion threshold used in the numerical scheme, which is a trade-off between precision and running time.

See also:

setEpsilon

virtual
double
getGamma() const = 0

coefficient for additional illumination variation term

See also:

setGamma

virtual
int
getInnerIterations() const = 0

Inner iterations (between outlier filtering) used in the numerical scheme.

See also:

setInnerIterations

virtual
double
getLambda() const = 0

Weight parameter for the data term, attachment parameter.

See also:

setLambda

virtual
int
getMedianFiltering() const = 0

Median filter kernel size (1 = no filter) (3 or 5)

See also:

setMedianFiltering

virtual
int
getOuterIterations() const = 0

Outer iterations (number of inner loops) used in the numerical scheme.

See also:

setOuterIterations

virtual
int
getScalesNumber() const = 0

Number of scales used to create the pyramid of images.

See also:

setScalesNumber

virtual
double
getScaleStep() const = 0

Step between scales (<1)

See also:

setScaleStep

virtual
double
getTau() const = 0

Time step of the numerical scheme.

See also:

setTau

virtual
double
getTheta() const = 0

Weight parameter for (u - v)^2, tightness parameter.

See also:

setTheta

virtual
bool
getUseInitialFlow() const = 0

Use initial flow.

See also:

setUseInitialFlow

virtual
int
getWarpingsNumber() const = 0

Number of warpings per scale.

See also:

setWarpingsNumber

virtual
void
setEpsilon(double val) = 0

Stopping criterion threshold used in the numerical scheme, which is a trade-off between precision and running time.

See also:

getEpsilon

virtual
void
setGamma(double val) = 0

coefficient for additional illumination variation term

See also:

getGamma

virtual
void
setInnerIterations(int val) = 0

Inner iterations (between outlier filtering) used in the numerical scheme.

See also:

getInnerIterations

virtual
void
setLambda(double val) = 0

Weight parameter for the data term, attachment parameter.

See also:

getLambda

virtual
void
setMedianFiltering(int val) = 0

Median filter kernel size (1 = no filter) (3 or 5)

See also:

getMedianFiltering

virtual
void
setOuterIterations(int val) = 0

Outer iterations (number of inner loops) used in the numerical scheme.

See also:

getOuterIterations

virtual
void
setScalesNumber(int val) = 0

Number of scales used to create the pyramid of images.

See also:

getScalesNumber

virtual
void
setScaleStep(double val) = 0

Step between scales (<1)

See also:

getScaleStep

virtual
void
setTau(double val) = 0

Time step of the numerical scheme.

See also:

getTau

virtual
void
setTheta(double val) = 0

Weight parameter for (u - v)^2, tightness parameter.

See also:

getTheta

virtual
void
setUseInitialFlow(bool val) = 0

Use initial flow.

See also:

getUseInitialFlow

virtual
void
setWarpingsNumber(int val) = 0

Number of warpings per scale.

See also:

getWarpingsNumber

static
Ptr<DualTVL1OpticalFlow>
create(
    double tau = 0.25,
    double lambda = 0.15,
    double theta = 0.3,
    int nscales = 5,
    int warps = 5,
    double epsilon = 0.01,
    int innnerIterations = 30,
    int outerIterations = 10,
    double scaleStep = 0.8,
    double gamma = 0.0,
    int medianFiltering = 5,
    bool useInitialFlow = false
    )

Creates instance of cv::DualTVL1OpticalFlow.