OpenCV2马拉松第12圈——直方图比較

收入囊中

• 使用4种不同的方法进行直方图比較

葵花宝典

要比較两个直方图,
首先必需要选择一个衡量直方图相似度的对照标准。也就是先说明要在哪个方面做对照。

我们能够想出非常多办法，OpenCV採用了下面4种

公式也都不难，我们自己就能实现。

d越小，表示差异越低，两幅图像越接近，越相似

初识API

C++: double compareHist(InputArray H1,
InputArray H2, int method)

C++: double compareHist(const
SparseMat& H1, const SparseMat& H2, int method)

	H1 – First compared histogram. H2 – Second compared histogram of the same size as H1 . method – Comparison method that could be one of the following: CV_COMP_CORREL Correlation CV_COMP_CHISQR Chi-Square CV_COMP_INTERSECT Intersection CV_COMP_BHATTACHARYYA Bhattacharyya distance CV_COMP_HELLINGER Synonym for CV_COMP_BHATTACHARYYA

荷枪实弹

#include "opencv2/highgui/highgui.hpp"
#include "opencv2/imgproc/imgproc.hpp"
#include <iostream>
#include <stdio.h>

using namespace std;
using namespace cv;

int main( int, char** argv )
{
	Mat src1,src2,gray1,gray2;
	src1 = imread(argv[1]);
	src2 = imread(argv[2]);
	cvtColor(src1, gray1, CV_BGR2GRAY);
	cvtColor(src2, gray2, CV_BGR2GRAY);

	int histSize = 256;
  	float range[] = { 0, 256 } ;
  	const float* histRange = { range };
 	int channels[] = {0};

 	Mat hist1,hist2;
	calcHist( &gray1, 1, channels, Mat(), hist1, 1, &histSize, &histRange);
	calcHist( &gray2, 1, channels, Mat(), hist2, 1, &histSize, &histRange);

	//相关：CV_COMP_CORREL
	//卡方：CV_COMP_CHISQR
	//直方图相交：CV_COMP_INTERSECT
	//Bhattacharyya距离：CV_COMP_BHATTACHARYYA
	double diff = compareHist(hist1,hist2,CV_COMP_BHATTACHARYYA);
	cout << diff << endl;
}

举一反三

这篇直方图比較很easy，没什么要说的了

计算机视觉讨论群：162501053

转载请注明：http://blog.csdn.net/abcd1992719g