利用OpenCV画HSV的3D直方图

此代码是根据opencv附带例子修改而成。

#include <iostream> #include <time.h> #include <cv.h> #include <highgui.h> #include <math.h> #include "CommandParser.h" using namespace std; int main(int argc, char** argv) { void WrongUsage(); CommandParser cp(argc,argv); // object to parse command line double total_time =0; double score= 0.8; //Load Template image char *param; char * filename; filename=param = cp.GetParameter("-i"); if(param==NULL) { cout<<"ERROR: Template image argument missing"; WrongUsage(); return -1; } IplImage * src= cvLoadImage(param); IplImage* hsv = cvCreateImage( cvGetSize(src), 8, 3 ); IplImage* h_plane = cvCreateImage( cvGetSize(src), 8, 1 ); IplImage* s_plane = cvCreateImage( cvGetSize(src), 8, 1 ); IplImage* v_plane = cvCreateImage( cvGetSize(src), 8, 1 ); IplImage* planes[] = { h_plane, s_plane ,v_plane}; /** H 分量划分为16个等级，S分量划分为8个等级 */ int h_bins = 10, s_bins = 6,v_bins = 6; param = cp.GetParameter("-h"); if(param) { h_bins=atoi(param); } param = cp.GetParameter("-s"); if(param) { s_bins=atoi(param); } param = cp.GetParameter("-v"); if(param) { v_bins=atoi(param); } int hist_size[] = {h_bins, s_bins,v_bins}; /** H 分量的变化范围 */ float h_ranges[] = { 0, 180 }; /** S 分量的变化范围*/ float s_ranges[] = { 0, 255 }; float v_ranges[] = { 0, 255 }; float* ranges[] = { h_ranges, s_ranges,v_ranges }; /** 输入图像转换到HSV颜色空间 */ cvCvtColor( src, hsv, CV_BGR2HSV ); cvCvtPixToPlane( hsv, h_plane, s_plane, v_plane, 0 ); /** 创建直方图，二维, 每个维度上均分 */ CvHistogram * hist = cvCreateHist(3, hist_size, CV_HIST_ARRAY, ranges, 1 ); /** 根据H,S两个平面数据统计直方图 */ cvCalcHist( planes, hist, 0, 0 ); /** 获取直方图统计的最大值，用于动态显示直方图 */ float max_value; int maxid[4]={0}; cvGetMinMaxHistValue( hist, 0, &max_value, 0, maxid ); /** 设置直方图显示图像 */ int height = 520; int ivh=height/v_bins; int width = (h_bins*s_bins*6);//每个bin6像素 IplImage* hist_img = cvCreateImage( cvSize(width,height), 8, 3 ); cvZero( hist_img ); CvFont font = cvFont(1,1); /** 用来进行HSV到RGB颜色转换的临时单位图像 */ IplImage * hsv_color = cvCreateImage(cvSize(1,1),8,3); IplImage * rgb_color = cvCreateImage(cvSize(1,1),8,3); int bin_w = width / (h_bins * s_bins); char dspStr1[80]; sprintf(dspStr1,"max=%f",max_value); cvPutText(hist_img,dspStr1,cvPoint(20,20),&font,cvScalar(0,0xff)); cout<< dspStr1 <<endl; float tatal_val=0; bool bh=false; float htotal=0.0; for(int v=0;v<v_bins;v++) { for(int h = 0; h < h_bins; h++) { bh=false; float hto=0.0; int i=0; for(int s = 0; s < s_bins; s++) { i = h*s_bins + s; /** 获得直方图中的统计次数，计算显示在图像中的高度 */ float bin_val = cvQueryHistValue_3D( hist, h, s,v ); int intensity = cvRound(bin_val*ivh/max_value); tatal_val+=bin_val; hto+=bin_val; if(bin_val==max_value) { bh=true; sprintf(dspStr1,"h=%d,s=%d,v=%d",h,s,v); cvPutText(hist_img,dspStr1,cvPoint(20,50),&font,cvScalar(0xff,0xaf)); cout<< dspStr1 <<endl; } /** 获得当前直方图代表的颜色，转换成RGB用于绘制 */ float vhr=(v+0.5)*255.f/v_bins;//避免显示黑色 cvSet2D(hsv_color,0,0,cvScalar(h*180.f / h_bins,(s+0.5)*255.f/s_bins,vhr,0)); cvCvtColor(hsv_color,rgb_color,CV_HSV2BGR); CvScalar color = cvGet2D(rgb_color,0,0); cvRectangle( hist_img, cvPoint(i*bin_w,(v+1)*ivh), cvPoint((i+1)*bin_w,(v+1)*ivh - intensity), color, -1, 8, 0 ); } if(bh) { htotal=hto; } cvLine(hist_img,cvPoint((i+1)*bin_w,v*ivh),cvPoint((i+1)*bin_w,(v+1)*ivh),cvScalar(0xA,0xff,0x98)); } cvLine(hist_img,cvPoint(0,(v+1)*ivh),cvPoint(width,(v+1)*ivh),cvScalar(0xA,0xff,0x98)); } sprintf(dspStr1,"pixels=%f",tatal_val); cvPutText(hist_img,dspStr1,cvPoint(20,80),&font,cvScalar(0,0xff)); cout<< dspStr1 <<endl; sprintf(dspStr1,"rate=%f",htotal/tatal_val); cvPutText(hist_img,dspStr1,cvPoint(20,110),&font,cvScalar(0,0xff,0xff)); cout<< dspStr1 <<endl; sprintf(dspStr1,"maxrt=%f",max_value/tatal_val); cvPutText(hist_img,dspStr1,cvPoint(20,140),&font,cvScalar(0xff,0xaf)); cout<< dspStr1 <<endl; cvNamedWindow(filename, 1 ); cvShowImage( filename, src ); cvNamedWindow( "H-S Histogram", 1 ); cvShowImage( "H-S Histogram", hist_img ); cvWaitKey(0); }  

代码效果如下：

原图

3D 直方图

图上表明最大像素数的HSV的值，以及该值所占比率。