提取輪廓在OpenCV里有一個函數?cvFindContours：

int cvFindContours( CvArr* image, CvMemStorage* storage, CvSeq** first_contour,int header_size=sizeof(CvContour),int mode=CV_RETR_LIST,int method=CV_CHAIN_APPROX_SIMPLE, CvPoint offset=cvPoint(0,0) );這個函數用起來很方便，但是隨著你使用的深入，你會發現有一些迷惑在這里。比如當你提取輪廓時只需要最外圍的一個輪廓，但是你會發現當輪廓畫出來時是好幾個；當你需要找一個最大輪廓時卻發現找出來的卻根本就不是你想要的那個。帶著這樣問題我們再來仔細看看cvFindContours這個函數。下邊的是一位仁兄寫的測試程序和測試圖片，說明提取輪廓的兩種方法及繪制輪廓中最大等級分析的問題，非常感謝他的分享，原文戳這里：/************************************************************************/??? /* 提取輪廓兩種方法對比及繪制輪廓'最大等級'分析???????????????????????? */??? /************************************************************************/??? #include "stdafx.h"??? #include "cv.h"??? #include "highgui.h"??? ??? int main()??? {??? ??? IplImage* img = cvLoadImage("lena.jpg", CV_LOAD_IMAGE_GRAYSCALE);??? ??? IplImage* img_temp = cvCreateImage(cvGetSize(img), 8, 1);??? ??? cvThreshold(img, img, 128, 255, CV_THRESH_BINARY);??? ??? CvMemStorage* mem_storage = cvCreateMemStorage(0);??? ??? CvSeq *first_contour = NULL, *c = NULL;??? ??? //??? ??? // 1、??? ??? cvNamedWindow("contour1");??? ??? cvCopyImage(img, img_temp);??? ??? double t = (double)cvGetTickCount();? ??? cvFindContours(img_temp, mem_storage, &first_contour);??? ??? cvZero(img_temp);??? ??? cvDrawContours(??? ??????? img_temp,???? ??????? first_contour,??? ??????? cvScalar(100),??? ??????? cvScalar(100),??? ??????? 1??? ??????? );??? ??? t = (double)cvGetTickCount() - t;?? ??? cvShowImage("contour1", img_temp);??? ??? printf("run1 = %gms\n", t/(cvGetTickFrequency()*1000.));??? ??? cvClearMemStorage(mem_storage);??? ??? //??? ??? // 2、??? ??? cvNamedWindow("contour2");??? ??? cvCopyImage(img, img_temp);??? ??? t = (double)cvGetTickCount();? ??? CvContourScanner scanner = cvStartFindContours(img_temp, mem_storage);??? ??? while (cvFindNextContour(scanner));??? ??? first_contour = cvEndFindContours(&scanner);??? ??????? ??? cvZero(img_temp);??? ??? cvDrawContours(??? ??????? img_temp,???? ??????? first_contour,??? ??????? cvScalar(100),??? ??????? cvScalar(100),??? ??????? 1??? ??????? );??? ??? t = (double)cvGetTickCount() - t;?? ??? cvShowImage("contour2", img_temp);??? ????? ??? printf("run2 = %gms\n", t/(cvGetTickFrequency()*1000.));??? ??????? ??? cvClearMemStorage(mem_storage);??? ??? cvReleaseImage(&img);??? ??? cvReleaseImage(&img_temp);??? ??? ??? cvWaitKey();??? ? ??? /************************************************************************/??? ??? /* 經測試 run1 = 16.1431ms run2 = 15.8677ms (參考) ?????? 不過可以肯定這兩中算法時間復雜度是相同的???????????????????????????????????? */??? ??? /************************************************************************/??? ??????? ??? //??? ??? // 上述兩種方法完成了對輪廓的提取,如想繪制輪廓都得配合cvDrawContours來使用??? ??? // 而cvDrawContours 函數第5個參數為 max_level 經查ICVL含義如下:??? ??? //??? ??? // 繪制輪廓的最大等級。如果等級為0，繪制單獨的輪廓。如果為1，繪制輪廓及在其后的相同的級別下輪廓。??? ??? // 如果值為2，所有的輪廓。如果等級為2，繪制所有同級輪廓及所有低一級輪廓，諸此種種。如果值為負數，??? ??? // 函數不繪制同級輪廓，但會升序繪制直到級別為abs(max_level)-1的子輪廓。??? ??? //??? ??? // 相信好多讀者初次都無法理解等級的含義,而且測試時候輸入>=1 的整數效果幾乎一樣??? ??? // 只有提取輪廓時候的提取模式設為 CV_RETR_CCOMP CV_RETR_TREE 時這個參數才有意義??? ??? //??? ??? // 經查FindContours 函數里面這樣介紹提取模式(mode)的這兩個參數:??? ??? // CV_RETR_CCOMP - 提取所有輪廓，并且將其組織為兩層的 hierarchy: 頂層為連通域的外圍邊界，次層為洞的內層邊界。???? ??? // CV_RETR_TREE - 提取所有輪廓，并且重構嵌套輪廓的全部 hierarchy???? ??? //???? ??? // 下面用第一種方法進行測試??? ??? ??? cvNamedWindow("contour_test");??? ??? cvNamedWindow("contour_raw");??? ??? img = cvLoadImage("contour.jpg", CV_LOAD_IMAGE_GRAYSCALE);??? ??? cvShowImage("contour_raw", img);??? ??? cvThreshold(img, img, 128, 255, CV_THRESH_BINARY);??? ??? img_temp = cvCloneImage(img);??? ??? cvFindContours(??? ??????? img_temp,???? ??????? mem_storage,???? ??????? &first_contour,??? ??????? sizeof(CvContour),??? ??????? CV_RETR_CCOMP?????????? //#1 需更改區域??? ??????? );??? ??? ??? cvZero(img_temp);??? ??? cvDrawContours(??? ??????? img_temp,???? ??????? first_contour,??? ??????? cvScalar(100),??? ??????? cvScalar(100),??? ??????? 1?????????????????????? //#2 需更改區域??? ??????? );??? ??? cvShowImage("contour_test", img_temp);??? ??? /************************************************************************/??? ??? /* (1, 2) = (CV_RETR_CCOMP, 1)? 如圖1? ?????? (1, 2) = (CV_RETR_CCOMP, 2)? 如圖2? ?????? (1, 2) = (CV_RETR_TREE, 1)?? 如圖3? ?????? (1, 2) = (CV_RETR_TREE, 2)?? 如圖4? ?????? (1, 2) = (CV_RETR_TREE, 6)?? 如圖5? ?????? 經分析CV_RETR_CCOMP 只把圖像分為兩個層次,頂層和次層,一等級輪廓只匹配與其最接近? ?????? 的內側輪廓即2等級? ?????? CV_RETR_TREE 則從輪廓外到內按等級1 - n 全部分配????????? ?????? CV_RETR_LIST 全部輪廓均為1級??????????????????????? */??? ??? /************************************************************************/??? ??? ??? cvWaitKey();??? ??? cvReleaseImage(&img);??? ??? cvReleaseImage(&img_temp);??? ??? cvReleaseMemStorage(&mem_storage);??? ??? cvDestroyAllWindows();??? ??? return 0;??? }

原圖

圖一

圖二

圖三

圖四

圖五

這是OpenCV的經典一個例子：

#include "cv.h"#include "cxcore.h"#include "highgui.h"#include <math.h>#endif?#pragma?? comment(lib,"cv.lib")? #pragma?? comment(lib,"highgui.lib")? #pragma?? comment(lib,"cxcore.lib")#define w 500int levels = 3;CvSeq* contours = 0;?void on_trackbar(int pos){??? IplImage* cnt_img = cvCreateImage( cvSize(w,w), 8, 3 );??? CvSeq* _contours = contours;??? int _levels = levels - 3;??? if( _levels <= 0 ) // get to the nearest face to make it look more funny??????? _contours = _contours->h_next->h_next->h_next->h_next->h_next->h_next->h_next->v_next->h_next->h_next;//_contours = _contours->v_next;??? cvZero( cnt_img );??? cvDrawContours( cnt_img, _contours, CV_RGB(255,0,0), CV_RGB(0,255,0), _levels);//, 3, CV_AA, cvPoint(0,0) );??? /*_levels:3，所有外輪廓及包含的內輪廓及里面的內輪廓2：所有外輪廓及包含的內輪廓1:所有外輪廓0，第一個外輪廓-1：第一個外輪廓及包含的內輪廓-2：第一個外輪廓及包含的內輪廓及里面的內輪廓?? _contours->h_next:同級的下一個輪廓_contours->v_next父級下的下層區域；*/cvShowImage( "contours", cnt_img );??? cvReleaseImage( &cnt_img );}?int main( int argc, char** argv ){??? int i, j;??? CvMemStorage* storage = cvCreateMemStorage(0);??? IplImage* img = cvCreateImage( cvSize(w,w), 8, 1 );???? cvZero( img );???? for( i=0; i < 6; i++ )??? {??????? int dx = (i%2)*250 - 30;//0%2=0;??????? int dy = (i/2)*150;??????? CvScalar white = cvRealScalar(255);??????? CvScalar black = cvRealScalar(0);???????? if( i == 0 )??????? {??????????? for( j = 0; j <= 10; j++ )??????????? {??????????????? double angle = (j+5)*CV_PI/21;??????????????? cvLine(img, cvPoint(cvRound(dx+100+j*10-80*cos(angle)),??????????????????? cvRound(dy+100-90*sin(angle))),??????????????????? cvPoint(cvRound(dx+100+j*10-30*cos(angle)),??????????????????? cvRound(dy+100-30*sin(angle))), white, 1, 8, 0);??????????? }??????? }???????? cvEllipse( img, cvPoint(dx+150, dy+100), cvSize(100,70), 0, 0, 360, white, -1, 8, 0 );??????? cvEllipse( img, cvPoint(dx+115, dy+70), cvSize(30,20), 0, 0, 360, black, -1, 8, 0 );??????? cvEllipse( img, cvPoint(dx+185, dy+70), cvSize(30,20), 0, 0, 360, black, -1, 8, 0 );??????? cvEllipse( img, cvPoint(dx+115, dy+70), cvSize(15,15), 0, 0, 360, white, -1, 8, 0 );??????? cvEllipse( img, cvPoint(dx+185, dy+70), cvSize(15,15), 0, 0, 360, white, -1, 8, 0 );??????? cvEllipse( img, cvPoint(dx+115, dy+70), cvSize(5,5), 0, 0, 360, black, -1, 8, 0 );??????? cvEllipse( img, cvPoint(dx+185, dy+70), cvSize(5,5), 0, 0, 360, black, -1, 8, 0 );??????? cvEllipse( img, cvPoint(dx+150, dy+100), cvSize(10,5), 0, 0, 360, black, -1, 8, 0 );??????? cvEllipse( img, cvPoint(dx+150, dy+150), cvSize(40,10), 0, 0, 360, black, -1, 8, 0 );??????? cvEllipse( img, cvPoint(dx+27, dy+100), cvSize(20,35), 0, 0, 360, white, -1, 8, 0 );??????? cvEllipse( img, cvPoint(dx+273, dy+100), cvSize(20,35), 0, 0, 360, white, -1, 8, 0 );??? }???? cvNamedWindow( "image", 1 );??? cvShowImage( "image", img );???? cvFindContours( img, storage, &contours, sizeof(CvContour),??????????????????? 2, CV_CHAIN_APPROX_SIMPLE, cvPoint(0,0) );???? // comment this out if you do not want approximation??? contours = cvApproxPoly( contours, sizeof(CvContour), storage, CV_POLY_APPROX_DP, 3, 1 );?//cvApproxPoly:???????????????????????????????????????????????? 逼近方法???? 精度 逼近曲線是否封閉??? cvNamedWindow( "contours", 1 );??? cvCreateTrackbar( "levels+3", "contours", &levels, 7, on_trackbar );???? on_trackbar(0);??? cvWaitKey(0);??? cvReleaseMemStorage( &storage );??? cvReleaseImage( &img );???? return 0;}

主要還是理解下int mode=CV_RETR_LIST,int method=CV_CHAIN_APPROX_SIMPLE,CvPoint offset=cvPoint(0,0));

當mode?為CV_RETR_CCOMP?只把圖像分為兩個層次,頂層和次層,一等級輪廓只匹配與其最接近??；

cvDrawContours?函數第5個參數為?max_level=0時，笑臉圖像會顯示第一個找到的輪廓，左邊的白色耳朵一只；

max_level=1時，所有白色區域的輪廓都會被顯示出來，因為他們都屬于等級1；

max_level=2時；每個白色區域里面的黑色區域會被顯示出來，可能一個白色區域下面有多個黑色區域，但他們都是同級的；

這里你要注意的的是每個白色區域下的黑色區域，如臉下面有4個黑色區域，白色眼珠下有一個黑色區域，這個黑色區域與臉下的那三個區域時同級的，也就是說他不屬于臉的內區域，他是白色眼珠的內區域；

當mode為???????CV_RETR_LIST?全部輪廓均為1級

???????????

再分享一下我老師大神的人工智能教程吧。零基礎！通俗易懂！風趣幽默！還帶黃段子！希望你也加入到我們人工智能的隊伍中來！https://blog.csdn.net/jiangjunshow

總結

以上是生活随笔為你收集整理的OpenCV函数cvFindContours的全部內容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網站內容還不錯，歡迎將生活随笔推薦給好友。