OpenCV中的 HOGDescriptor 类


其定义在  object.hpp中找到的：

struct CV_EXPORTS_W HOGDescriptor
{
public:
    enum { L2Hys=0 };
    enum { DEFAULT_NLEVELS=64 };

    CV_WRAP HOGDescriptor() : winSize(64,128), blockSize(16,16), blockStride(8,8),
        cellSize(8,8), nbins(9), derivAperture(1), winSigma(-1),
        histogramNormType(HOGDescriptor::L2Hys), L2HysThreshold(0.2), gammaCorrection(true),
        nlevels(HOGDescriptor::DEFAULT_NLEVELS)
    {}

    CV_WRAP HOGDescriptor(Size _winSize, Size _blockSize, Size _blockStride,
                  Size _cellSize, int _nbins, int _derivAperture=1, double _winSigma=-1,
                  int _histogramNormType=HOGDescriptor::L2Hys,
                  double _L2HysThreshold=0.2, bool _gammaCorrection=false,
                  int _nlevels=HOGDescriptor::DEFAULT_NLEVELS)
    : winSize(_winSize), blockSize(_blockSize), blockStride(_blockStride), cellSize(_cellSize),
    nbins(_nbins), derivAperture(_derivAperture), winSigma(_winSigma),
    histogramNormType(_histogramNormType), L2HysThreshold(_L2HysThreshold),
    gammaCorrection(_gammaCorrection), nlevels(_nlevels)
    {}

    CV_WRAP HOGDescriptor(const String& filename)
    {
        load(filename);
    }

    HOGDescriptor(const HOGDescriptor& d)
    {
        d.copyTo(*this);
    }

    virtual ~HOGDescriptor() {}

    CV_WRAP size_t getDescriptorSize() const;
    CV_WRAP bool checkDetectorSize() const;
    CV_WRAP double getWinSigma() const;

    CV_WRAP virtual void setSVMDetector(InputArray _svmdetector);

    virtual bool read(FileNode& fn);
    virtual void write(FileStorage& fs, const String& objname) const;

    CV_WRAP virtual bool load(const String& filename, const String& objname=String());
    CV_WRAP virtual void save(const String& filename, const String& objname=String()) const;
    virtual void copyTo(HOGDescriptor& c) const;

    CV_WRAP virtual void compute(const Mat& img,
                        CV_OUT vector<float>& descriptors,
                        Size winStride=Size(), Size padding=Size(),
                        const vector<Point>& locations=vector<Point>()) const;
    //with found weights output
    CV_WRAP virtual void detect(const Mat& img, CV_OUT vector<Point>& foundLocations,
                        CV_OUT vector<double>& weights,
                        double hitThreshold=0, Size winStride=Size(),
                        Size padding=Size(),
                        const vector<Point>& searchLocations=vector<Point>()) const;
    //without found weights output
    virtual void detect(const Mat& img, CV_OUT vector<Point>& foundLocations,
                        double hitThreshold=0, Size winStride=Size(),
                        Size padding=Size(),
                        const vector<Point>& searchLocations=vector<Point>()) const;
    //with result weights output
    CV_WRAP virtual void detectMultiScale(const Mat& img, CV_OUT vector<Rect>& foundLocations,
                                  CV_OUT vector<double>& foundWeights, double hitThreshold=0,
                                  Size winStride=Size(), Size padding=Size(), double scale=1.05,
                                  double finalThreshold=2.0,bool useMeanshiftGrouping = false) const;
    //without found weights output
    virtual void detectMultiScale(const Mat& img, CV_OUT vector<Rect>& foundLocations,
                                  double hitThreshold=0, Size winStride=Size(),
                                  Size padding=Size(), double scale=1.05,
                                  double finalThreshold=2.0, bool useMeanshiftGrouping = false) const;

    CV_WRAP virtual void computeGradient(const Mat& img, CV_OUT Mat& grad, CV_OUT Mat& angleOfs,
                                Size paddingTL=Size(), Size paddingBR=Size()) const;

    CV_WRAP static vector<float> getDefaultPeopleDetector();
    CV_WRAP static vector<float> getDaimlerPeopleDetector();

    CV_PROP Size winSize;
    CV_PROP Size blockSize;
    CV_PROP Size blockStride;
    CV_PROP Size cellSize;
    CV_PROP int nbins;
    CV_PROP int derivAperture;
    CV_PROP double winSigma;
    CV_PROP int histogramNormType;
    CV_PROP double L2HysThreshold;
    CV_PROP bool gammaCorrection;
    CV_PROP vector<float> svmDetector;
    CV_PROP int nlevels;


  // evaluate specified ROI and return confidence value for each location
  void detectROI(const cv::Mat& img, const vector<cv::Point> &locations,
                                  CV_OUT std::vector<cv::Point>& foundLocations, CV_OUT std::vector<double>& confidences,
                                  double hitThreshold = 0, cv::Size winStride = Size(),
                                  cv::Size padding = Size()) const;

  // evaluate specified ROI and return confidence value for each location in multiple scales
  void detectMultiScaleROI(const cv::Mat& img,
                                                      CV_OUT std::vector<cv::Rect>& foundLocations,
                                                      std::vector<DetectionROI>& locations,
                                                      double hitThreshold = 0,
                                                      int groupThreshold = 0) const;

  // read/parse Dalal's alt model file
  void readALTModel(std::string modelfile);
};

默认构造函数的几个参数：

winSize(64,128), blockSize(16,16), blockStride(8,8),
        cellSize(8,8), nbins(9), derivAperture(1), winSigma(-1),
        histogramNormType(HOGDescriptor::L2Hys), L2HysThreshold(0.2), gammaCorrection(true),
        nlevels(HOGDescriptor::DEFAULT_NLEVELS)

winSize ： 窗口的大小
blockSize ：块的大小
cellSize： 胞元的大小
nbins： 方向bin的个数 nBins表示在一个胞元（cell）中统计梯度的方向数目，例如nBins=9时，在一个胞元内统计9个方向的梯度直方图，每个方向为360/9=40度。

更详细请移步：

更多详情见请继续阅读下一页的精彩内容：

--------------------------------------分割线 --------------------------------------

Ubuntu Linux下安装OpenCV2.4.1所需包

Ubuntu 12.04 安装 OpenCV2.4.2

CentOS下OpenCV无法读取视频文件

Ubuntu 12.04下安装OpenCV 2.4.5总结

Ubuntu 10.04中安装OpenCv2.1九步曲

基于QT和OpenCV的人脸识别系统

--------------------------------------分割线 --------------------------------------

  • 1
  • 2
  • 下一页

相关内容