class cv::cuda::HOG

Overview

The class implements Histogram of Oriented Gradients ([17]) object detector. More…

#include <cudaobjdetect.hpp>

class HOG: public cv::Algorithm
{
public:
    // enums

    enum
    {
        DESCR_FORMAT_ROW_BY_ROW,
        DESCR_FORMAT_COL_BY_COL,
    };

    // methods

    static
    Ptr<HOG>
    create(
        Size win_size = Size(64, 128),
        Size block_size = Size(16, 16),
        Size block_stride = Size(8, 8),
        Size cell_size = Size(8, 8),
        int nbins = 9
        );

    virtual
    void
    compute(
        InputArray img,
        OutputArray descriptors,
        Stream& stream = Stream::Null()
        ) = 0;

    virtual
    void
    detect(
        InputArray img,
        std::vector<Point>& found_locations,
        std::vector<double>* confidences = NULL
        ) = 0;

    virtual
    void
    detectMultiScale(
        InputArray img,
        std::vector<Rect>& found_locations,
        std::vector<double>* confidences = NULL
        ) = 0;

    virtual
    size_t
    getBlockHistogramSize() const = 0;

    virtual
    Mat
    getDefaultPeopleDetector() const = 0;

    virtual
    int
    getDescriptorFormat() const = 0;

    virtual
    size_t
    getDescriptorSize() const = 0;

    virtual
    bool
    getGammaCorrection() const = 0;

    virtual
    int
    getGroupThreshold() const = 0;

    virtual
    double
    getHitThreshold() const = 0;

    virtual
    double
    getL2HysThreshold() const = 0;

    virtual
    int
    getNumLevels() const = 0;

    virtual
    double
    getScaleFactor() const = 0;

    virtual
    double
    getWinSigma() const = 0;

    virtual
    Size
    getWinStride() const = 0;

    virtual
    void
    setDescriptorFormat(int descr_format) = 0;

    virtual
    void
    setGammaCorrection(bool gamma_correction) = 0;

    virtual
    void
    setGroupThreshold(int group_threshold) = 0;

    virtual
    void
    setHitThreshold(double hit_threshold) = 0;

    virtual
    void
    setL2HysThreshold(double threshold_L2hys) = 0;

    virtual
    void
    setNumLevels(int nlevels) = 0;

    virtual
    void
    setScaleFactor(double scale0) = 0;

    virtual
    void
    setSVMDetector(InputArray detector) = 0;

    virtual
    void
    setWinSigma(double win_sigma) = 0;

    virtual
    void
    setWinStride(Size win_stride) = 0;
};

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);

protected:
    // methods

    void
    writeFormat(FileStorage& fs) const;

Detailed Documentation

The class implements Histogram of Oriented Gradients ([17]) object detector.

An example applying the HOG descriptor for people detection can be found at opencv_source_code/samples/cpp/peopledetect.cpp
A CUDA example applying the HOG descriptor for people detection can be found at opencv_source_code/samples/gpu/hog.cpp
(Python) An example applying the HOG descriptor for people detection can be found at opencv_source_code/samples/python/peopledetect.py

Methods

static
Ptr<HOG>
create(
    Size win_size = Size(64, 128),
    Size block_size = Size(16, 16),
    Size block_stride = Size(8, 8),
    Size cell_size = Size(8, 8),
    int nbins = 9
    )

Creates the HOG descriptor and detector.

Parameters:

win_size	Detection window size. Align to block size and block stride.
block_size	Block size in pixels. Align to cell size. Only (16,16) is supported for now.
block_stride	Block stride. It must be a multiple of cell size.
cell_size	Cell size. Only (8, 8) is supported for now.
nbins	Number of bins. Only 9 bins per cell are supported for now.

virtual
void
compute(
    InputArray img,
    OutputArray descriptors,
    Stream& stream = Stream::Null()
    ) = 0

Returns block descriptors computed for the whole image.

Parameters:

img	Source image. See cuda::HOGDescriptor::detect for type limitations.
descriptors	2D array of descriptors.
stream	CUDA stream.

virtual
void
detect(
    InputArray img,
    std::vector<Point>& found_locations,
    std::vector<double>* confidences = NULL
    ) = 0

Performs object detection without a multi-scale window.

Parameters:

img	Source image. CV_8UC1 and CV_8UC4 types are supported for now.
found_locations	Left-top corner points of detected objects boundaries.
confidences	Optional output array for confidences.

virtual
void
detectMultiScale(
    InputArray img,
    std::vector<Rect>& found_locations,
    std::vector<double>* confidences = NULL
    ) = 0

Performs object detection with a multi-scale window.

Parameters:

img	Source image. See cuda::HOGDescriptor::detect for type limitations.
found_locations	Detected objects boundaries.
confidences	Optional output array for confidences.

virtual
size_t
getBlockHistogramSize() const = 0

Returns the block histogram size.

virtual
Mat
getDefaultPeopleDetector() const = 0

Returns coefficients of the classifier trained for people detection.

virtual
size_t
getDescriptorSize() const = 0

Returns the number of coefficients required for the classification.

virtual
void
setDescriptorFormat(int descr_format) = 0

Descriptor storage format:

DESCR_FORMAT_ROW_BY_ROW - Row-major order.
DESCR_FORMAT_COL_BY_COL - Column-major order.

virtual
void
setGammaCorrection(bool gamma_correction) = 0

Flag to specify whether the gamma correction preprocessing is required or not.

virtual
void
setGroupThreshold(int group_threshold) = 0

Coefficient to regulate the similarity threshold. When detected, some objects can be covered by many rectangles. 0 means not to perform grouping. See groupRectangles.

virtual
void
setHitThreshold(double hit_threshold) = 0

Threshold for the distance between features and SVM classifying plane. Usually it is 0 and should be specfied in the detector coefficients (as the last free coefficient). But if the free coefficient is omitted (which is allowed), you can specify it manually here.

virtual
void
setL2HysThreshold(double threshold_L2hys) = 0

L2-Hys normalization method shrinkage.

virtual
void
setNumLevels(int nlevels) = 0

Maximum number of detection window increases.

virtual
void
setScaleFactor(double scale0) = 0

Coefficient of the detection window increase.

virtual
void
setSVMDetector(InputArray detector) = 0

Sets coefficients for the linear SVM classifier.

virtual
void
setWinSigma(double win_sigma) = 0

Gaussian smoothing window parameter.

virtual
void
setWinStride(Size win_stride) = 0

Window stride. It must be a multiple of block stride.

Table of Contents

Previous topic

Next topic

class cv::cuda::HOG

Overview

Inherited Members

Detailed Documentation

Methods