OpenCV图像处理车牌检测与定位应用

最新推荐文章于 2024-07-26 09:47:27 发布
最新推荐文章于 2024-07-26 09:47:27 发布
阅读量5.3k 收藏 10
点赞数
分类专栏： OpenCV

这两天在做关于车牌识别的实验，用了几种方式：

1.车牌颜色分布（HSV空间，YCrCb空间的没有颜色分布图谱，无法实验）；利用HSV的H通道，效果一般，受环境影响大。

#include "highgui.h"
#include "cv.h"
#include <stdio.h>   
#include <math.h>  
#include <string>
#include<iostream>
using namespace std;

CvPoint Point;
IplImage* img=0;
// skin region location using rgb limitation
void SkinRGB(IplImage* rgb,IplImage* _dst)
{
	assert(rgb->nChannels==3&& _dst->nChannels==3);

	static const int R=2;
	static const int G=1;
	static const int B=0;

	IplImage* dst=cvCreateImage(cvGetSize(_dst),8,3);
	cvZero(dst);

	for (int h=0;h<rgb->height;h++) {
		unsigned char* prgb=(unsigned char*)rgb->imageData+h*rgb->widthStep;
		unsigned char* pdst=(unsigned char*)dst->imageData+h*dst->widthStep;
		for (int w=0;w<rgb->width;w++) {
			if ((prgb[R]>95 && prgb[G]>40 && prgb[B]>20 &&
				prgb[R]-prgb[B]>15 && prgb[R]-prgb[G]>15/*&&
				!(prgb[R]>170&&prgb[G]>170&&prgb[B]>170)*/)||//uniform illumination 
				(prgb[R]>200 && prgb[G]>210 && prgb[B]>170 &&
				abs(prgb[R]-prgb[B])<=15 && prgb[R]>prgb[B]&& prgb[G]>prgb[B])//lateral illumination
				) {
					memcpy(pdst,prgb,3);
			}			
			prgb+=3;
			pdst+=3;
		}
	}
	cvCopyImage(dst,_dst);
	cvReleaseImage(&dst);
}
// skin detection in rg space
void cvSkinRG(IplImage* rgb,IplImage* gray)
{
	assert(rgb->nChannels==3&&gray->nChannels==1);
	
	const int R=2;
	const int G=1;
	const int B=0;

	double Aup=-1.8423;
	double Bup=1.5294;
	double Cup=0.0422;
	double Adown=-0.7279;
	double Bdown=0.6066;
	double Cdown=0.1766;
	for (int h=0;h<rgb->height;h++) {
		unsigned char* pGray=(unsigned char*)gray->imageData+h*gray->widthStep;
		unsigned char* pRGB=(unsigned char* )rgb->imageData+h*rgb->widthStep;
		for (int w=0;w<rgb->width;w++) 
		{
			int s=pRGB[R]+pRGB[G]+pRGB[B];
			double r=(double)pRGB[R]/s;
			double g=(double)pRGB[G]/s;
			double Gup=Aup*r*r+Bup*r+Cup;
			double Gdown=Adown*r*r+Bdown*r+Cdown;
			double Wr=(r-0.33)*(r-0.33)+(g-0.33)*(g-0.33);
			if (g<Gup && g>Gdown && Wr>0.004)
			{
				*pGray=255;
			}
			else
			{ 
				*pGray=0;
			}
			pGray++;
			pRGB+=3;
		}
	}

}
// implementation of otsu algorithm
// author: onezeros#yahoo.cn
// reference: Rafael C. Gonzalez. Digital Image Processing Using MATLAB
void cvThresholdOtsu(IplImage* src, IplImage* dst)
{
	int height=src->height;
	int width=src->width;

	//histogram
	float histogram[256]={0};
	for(int i=0;i<height;i++) {
		unsigned char* p=(unsigned char*)src->imageData+src->widthStep*i;
		for(int j=0;j<width;j++) {
			histogram[*p++]++;
		}
	}
	//normalize histogram
	int size=height*width;
	for(int i=0;i<256;i++) {
		histogram[i]=histogram[i]/size;
	}

	//average pixel value
	float avgValue=0;
	for(int i=0;i<256;i++) {
		avgValue+=i*histogram[i];
	}

	int threshold;	
	float maxVariance=0;
	float w=0,u=0;
	for(int i=0;i<256;i++) {
		w+=histogram[i];
		u+=i*histogram[i];

		float t=avgValue*w-u;
		float variance=t*t/(w*(1-w));
		if(variance>maxVariance) {
			maxVariance=variance;
			threshold=i;
		}
	}

	cvThreshold(src,dst,threshold,255,CV_THRESH_BINARY);
}

void cvSkinOtsu(IplImage* src, IplImage* dst)
{
	assert(dst->nChannels==1&& src->nChannels==3);

	IplImage* ycrcb=cvCreateImage(cvGetSize(src),8,3);
	IplImage* cr=cvCreateImage(cvGetSize(src),8,1);
	cvCvtColor(src,ycrcb,CV_BGR2YCrCb);
	cvSplit(ycrcb,0,cr,0,0);

	cvThresholdOtsu(cr,cr);
	cvCopyImage(cr,dst);
	cvReleaseImage(&cr);
	cvReleaseImage(&ycrcb);
}

void cvSkinYUV(IplImage* src,IplImage* dst)
{
	IplImage* ycrcb=cvCreateImage(cvGetSize(src),8,3);
	//IplImage* cr=cvCreateImage(cvGetSize(src),8,1);
	//IplImage* cb=cvCreateImage(cvGetSize(src),8,1);
	cvCvtColor(src,ycrcb,CV_BGR2YCrCb);
	//cvSplit(ycrcb,0,cr,cb,0);

	static const int Cb=2;
	static const int Cr=1;
	static const int Y=0;

	//IplImage* dst=cvCreateImage(cvGetSize(_dst),8,3);
	cvZero(dst);

	for (int h=0;h<src->height;h++) {
		unsigned char* pycrcb=(unsigned char*)ycrcb->imageData+h*ycrcb->widthStep;
		unsigned char* psrc=(unsigned char*)src->imageData+h*src->widthStep;
		unsigned char* pdst=(unsigned char*)dst->imageData+h*dst->widthStep;
		for (int w=0;w<src->width;w++) {
			if ((pycrcb[Cr]<=126||pycrcb[Cr]>=130)&&(pycrcb[Cb]<=126||pycrcb[Cb]>=130))
			{
					memcpy(pdst,psrc,3);
			}
			pycrcb+=3;
			psrc+=3;
			pdst+=3;
		}
	}
	//cvCopyImage(dst,_dst);
	//cvReleaseImage(&dst);
}

void cvSkinHSV(IplImage* src,IplImage* dst)
{
	IplImage* hsv=cvCreateImage(cvGetSize(src),8,3);
	//IplImage* cr=cvCreateImage(cvGetSize(src),8,1);
	//IplImage* cb=cvCreateImage(cvGetSize(src),8,1);
	cvCvtColor(src,hsv,CV_BGR2HSV);
	//cvSplit(ycrcb,0,cr,cb,0);

	static const int V=2;
	static const int S=1;
	static const int H=0;

	//IplImage* dst=cvCreateImage(cvGetSize(_dst),8,3);
	cvZero(dst);

	for (int h=0;h<src->height;h++) {
		unsigned char* phsv=(unsigned char*)hsv->imageData+h*hsv->widthStep;
		unsigned char* psrc=(unsigned char*)src->imageData+h*src->widthStep;
		unsigned char* pdst=(unsigned char*)dst->imageData+h*dst->widthStep;
		for (int w=0;w<src->width;w++) {
			if (phsv[H]>=90&&phsv[H]<=135)
			{
					memcpy(pdst,psrc,3);
			}
			phsv+=3;
			psrc+=3;
			pdst+=3;
		}
	}
	//cvCopyImage(dst,_dst);
	//cvReleaseImage(&dst);
}

void on_mouse(int event,int x,int y,int flags,void* param )  
{  
	 switch(event)  
     {  
	 case CV_EVENT_LBUTTONUP:  
         {  
             Point=cvPoint(x,y);  
         }  
			cvCircle(img,Point,1,CV_RGB(255,0,0),1);
			CvScalar HSV=cvGet2D(img,x,y);
			cout<<"H:"<<HSV.val[0]<<"\t S:"<<HSV.val[1]<<"\t V:"<<HSV.val[2]<<endl; 
         break;  
     }  
  
//printf("( %d, %d) ",x,y);  
//printf("The Event is : %d ",event);  
//printf("The flags is : %d ",flags);  
//printf("The param is : %d\n",param);  
}
int main()
{   
	
    IplImage* img0= cvLoadImage("D:/image/car/00.jpg"); //随便放一张jpg图片在D盘或另行设置目录
	img=cvCreateImage(cvSize(400,300),8,3);
	cvResize(img0,img);
	IplImage* dstRGB=cvCreateImage(cvGetSize(img),8,3);
	IplImage* dstRG=cvCreateImage(cvGetSize(img),8,1);
	IplImage* dst_crotsu=cvCreateImage(cvGetSize(img),8,1);
	IplImage* dst_YUV=cvCreateImage(cvGetSize(img),8,3);
	IplImage* dst_HSV=cvCreateImage(cvGetSize(img),8,3);


    cvNamedWindow("inputimage", CV_WINDOW_AUTOSIZE);
    cvShowImage("inputimage", img);
    cvWaitKey(0);
	/*
	SkinRGB(img,dstRGB);
	cvNamedWindow("outputimage1", CV_WINDOW_AUTOSIZE);
    cvShowImage("outputimage1", dstRGB);
    cvWaitKey(0);
	cvSkinRG(img,dstRG);
	cvNamedWindow("outputimage2", CV_WINDOW_AUTOSIZE);
    cvShowImage("outputimage2", dstRG);
	cvWaitKey(0);
	cvSkinOtsu(img,dst_crotsu);
	cvNamedWindow("outputimage3", CV_WINDOW_AUTOSIZE);
    cvShowImage("outputimage3", dst_crotsu);
	cvWaitKey(0);
	
	cvSkinYUV(img,dst_YUV);
	cvNamedWindow("outputimage4", CV_WINDOW_AUTOSIZE);
    cvShowImage("outputimage4", dst_YUV);
	//cvSaveImage("D:/skin04.jpg",dst_YUV);
	cvWaitKey(0);
	*/
	cvSkinHSV(img,dst_HSV);
	cvNamedWindow("outputimage5", CV_WINDOW_AUTOSIZE);
    cvShowImage("outputimage5", dst_HSV);
	cvSaveImage("D:/image/car/car00.jpg",dst_HSV);
	cvWaitKey(0);
	
    return 0;
}

2.Canny+Hough；效果也不好，但学习了hough变换的有关内容。 

#include <cv.h>
#include <highgui.h>
#include <math.h>
 
int main(int argc, char** argv)
{
    const char* filename = argc >= 2 ? argv[1] : "D:/image/car/car04.jpg";
    IplImage* src = cvLoadImage( filename, 0 );
	cvDilate(src,src);
    IplImage* dst;
    IplImage* color_dst;
    CvMemStorage* storage = cvCreateMemStorage(0);
    CvSeq* lines = 0;
    int i;
 
    if( !src )
        return -1;
 
    dst = cvCreateImage( cvGetSize(src), 8, 1 );
    color_dst = cvCreateImage( cvGetSize(src), 8, 3 );
 
    cvCanny( src, dst, 50, 150, 3 );
    cvCvtColor( dst, color_dst, CV_GRAY2BGR );
	
#if 0
    lines = cvHoughLines2( dst, storage, CV_HOUGH_STANDARD, 1, CV_PI/180, 100, 0, 0 );
 
    for( i = 0; i < MIN(lines->total,100); i++ )
    {
        float* line = (float*)cvGetSeqElem(lines,i);
        float rho = line[0];
        float theta = line[1];
        CvPoint pt1, pt2;
        double a = cos(theta), b = sin(theta);
        double x0 = a*rho, y0 = b*rho;
        pt1.x = cvRound(x0 + 1000*(-b));
        pt1.y = cvRound(y0 + 1000*(a));
        pt2.x = cvRound(x0 - 1000*(-b));
        pt2.y = cvRound(y0 - 1000*(a));
        cvLine( color_dst, pt1, pt2, CV_RGB(255,0,0), 3, CV_AA, 0 );
    }
#else
	
    lines = cvHoughLines2( dst, storage, CV_HOUGH_PROBABILISTIC, 1, CV_PI/180, 50, 5, 3 );
    for( i = 0; i < lines->total; i++ )
    {
        CvPoint* line = (CvPoint*)cvGetSeqElem(lines,i);
        cvLine( color_dst, line[0], line[1], CV_RGB(255,0,0), 3, CV_AA, 0 );
    }
//#endif
    cvNamedWindow( "Source", 1 );
    cvShowImage( "Source", src );
 
    cvNamedWindow( "Hough", 1 );
    cvShowImage( "Hough", color_dst );
 
    cvWaitKey(0);
 
    return 0;
}

3.Coutour检测；效果勉强。 

#include "cv.h"
#include "highgui.h"
#include <cxcore.h>
#include <stdio.h>
 
int BinarizeImageByOTSU (IplImage * src)
{ 
	assert(src != NULL);
 
	//get the ROI
	CvRect rect = cvGetImageROI(src);
 
	//information of the source image
	int x = rect.x;
	int y = rect.y;
	int width = rect.width; 
	int height = rect.height;
	int ws = src->widthStep;
 
	int thresholdValue=1;//阈值
	int ihist [256] ; // 图像直方图, 256个点
	int i, j, k,n, n1, n2, Color=0;
	double m1, m2, sum, csum, fmax, sb;
	memset (ihist, 0, sizeof (ihist)) ; // 对直方图置 零...
 
	for (i=y;i< y+height;i++) // 生成直方图
	{ 
		int mul =  i*ws;
		for (j=x;j<x+width;j++)
		{ 
			//Color=Point (i,j) ;
			Color = (int)(unsigned char)*(src->imageData + mul+ j);
			ihist [Color] +=1;
		}
	}
	sum=csum=0.0;
	n=0;
	for (k = 0; k <= 255; k++)
	{ 
		sum+= (double) k* (double) ihist [k] ; // x*f (x) 质量矩
		n +=ihist [k]; //f (x) 质量
	}
	// do the otsu global thresholding method
	fmax = - 1.0;
	n1 = 0;
	for (k=0;k<255;k++) 
	{
		n1+=ihist [k] ;
		if (! n1)
		{ 
			continue; 
		}
		n2=n- n1;
		if (n2==0) 
		{
			break;
		}
		csum+= (double) k*ihist [k] ;
		m1=csum/ n1;
		m2= (sum- csum) /n2;
		sb = ( double) n1* ( double) n2* ( m1 - m2) * (m1- m2) ;
 
		if (sb>fmax) 
		{
			fmax=sb;
			thresholdValue=k;
		}
	}
 
	//binarize the image 
	cvThreshold( src, src ,thresholdValue, 255, CV_THRESH_BINARY ); 
	return 0;
} 
 
int main( int argc, char* argv[])
{
	IplImage* src;
	if((src=cvLoadImage("D:/image/car/05sobel.jpg", 0)))//载入图像
	{
		//为轮廓显示图像申请空间,3通道图像，以便用彩色显示
		IplImage* dst = cvCreateImage( cvGetSize(src), 8, 3);
		//创建内存块，将该块设置成默认值,当前默认大小为64k
		CvMemStorage* storage = cvCreateMemStorage(0);
		//可动态增长元素序列
		CvSeq* contour = 0;
		//对图像进行自适二值化
		BinarizeImageByOTSU(src);
		//图像膨胀
		cvDilate(src,src);
		//图像腐蚀
		cvErode(src,src);
		//显示源图像的二值图
		cvNamedWindow( "Source", 1 );
		cvShowImage( "Source", src );
		//在二值图像中寻找轮廓
		cvFindContours( src, storage, &contour, sizeof(CvContour), CV_RETR_CCOMP, CV_CHAIN_APPROX_SIMPLE );
		cvZero( dst );//清空数组
		cvCvtColor(src,dst,CV_GRAY2BGR);
		//目标轮廓最小下限
		int mix_area = 2500;
		//目标轮廓最大上限
		int max_area = 3500;
		//可存放在1-，2-，3-，4-TUPLE类型的捆绑数据的容器
		CvScalar color = CV_RGB( 255, 0, 0);
		//在图像中绘制外部和内部的轮廓
		for( ; contour != 0; contour = contour->h_next)
		{
			//取得轮廓的最小矩形
			CvRect aRect = cvBoundingRect( contour, 1 );
			//取得矩形的面积
			int tmparea=aRect.height*aRect.height;
			//用车牌的形态做判断
			if (((double)aRect.width/(double)aRect.height>3)
				&& ((double)aRect.width/(double)aRect.height<6))
			{
				cvRectangle(dst,cvPoint(aRect.x,aRect.y),cvPoint(aRect.x+aRect.width ,aRect.y+aRect.height),color,2);
				//cvDrawContours( dst, contour, color, color, -1, 1, 8 );
			}
		}
 
		cvNamedWindow( "Components", 1 );
		cvShowImage( "Components", dst );
		cvWaitKey(0);	
		cvDestroyWindow("Components");
		cvReleaseImage(&dst);
		cvDestroyWindow("Source");
		cvReleaseImage(&src);
 
		return 0;
	}	
	return 1;
}

4.Squares方式：Canny||Threshold+cvFindContours+cvApproxPoly；效果一般 

#ifdef _CH_
#pragma package <opencv>
#endif

#ifndef _EiC
#include "cv.h"
#include "highgui.h"
#include <stdio.h>
#include <math.h>
#include <string.h>
#endif

int thresh = 50;
IplImage* img = 0;
IplImage* img0 = 0;
CvMemStorage* storage =  cvCreateMemStorage(0);
CvPoint pt[4];
const char* wndname = "Square Detection Demo";

// helper function:
// finds a cosine of angle between vectors
// from pt0->pt1 and from pt0->pt2 
double angle( CvPoint* pt1, CvPoint* pt2, CvPoint* pt0 )
{
    double dx1 = pt1->x - pt0->x;
    double dy1 = pt1->y - pt0->y;
    double dx2 = pt2->x - pt0->x;
    double dy2 = pt2->y - pt0->y;
    return (dx1*dx2 + dy1*dy2)/sqrt((dx1*dx1 + dy1*dy1)*(dx2*dx2 + dy2*dy2) + 1e-10);
}

// returns sequence of squares detected on the image.
//返回图像中的四边形序列
// the sequence is stored in the specified memory storage
//序列存储在特定的storage中
CvSeq* findSquares4( IplImage* img, CvMemStorage* storage )
{
	
    CvSeq* contours;
    int i, c, l, N = 11;
    CvSize sz = cvSize( img->width & -2, img->height & -2 );
    IplImage* timg = cvCloneImage( img ); // make a copy of input image复制输入图像
    IplImage* gray = cvCreateImage( sz, 8, 1 ); 
    IplImage* pyr = cvCreateImage( cvSize(sz.width/2, sz.height/2), 8, 3 );//尺度减小为1/2
    IplImage* tgray;
    CvSeq* result;
    double s, t;
    // create empty sequence that will contain points -
    // 4 points per square (the square's vertices)
	//建立一个空序列存储每个四边形的四个顶点
    CvSeq* squares = cvCreateSeq( 0, sizeof(CvSeq), sizeof(CvPoint), storage );
    
    // select the maximum ROI in the image
    // with the width and height divisible by 2
	//设定timg的ROI为最大值（）
    cvSetImageROI( timg, cvRect( 0, 0, sz.width, sz.height ));
    
    // down-scale and upscale the image to filter out the noise
	//金字塔方式升和降来滤波去除噪声
    //cvPyrDown( timg, pyr, 7 );
    //cvPyrUp( pyr, timg, 7 );
    tgray = cvCreateImage( sz, 8, 1 );
    
    // find squares in every color plane of the image
	//寻找每个通道的四边形
    for( c = 0; c < 3; c++ )
    {
        // extract the c-th color plane
		//提取第c个通道
        cvSetImageCOI( timg, c+1 );
        cvCopy( timg, tgray, 0 );
        
        // try several threshold levels
		//尝试每个阈值等级
        for( l = 0; l < N; l++ )
        {
            // hack: use Canny instead of zero threshold level.
            // Canny helps to catch squares with gradient shading   
            //Canny代替零阈值，Canny通过梯度变化程度大来寻找四边形

			if( l == 0 )
            {
                // apply Canny. Take the upper threshold from slider
                // and set the lower to 0 (which forces edges merging)
				// l=0使用Canny
                cvCanny( tgray, gray,60, 180, 3 );
				// 
                // dilate canny output to remove potential
                // holes between edge segments 
                cvDilate( gray, gray, 0, 1 );
            }
            else
            {
                // apply threshold if l!=0:
                // tgray(x,y) = gray(x,y) < (l+1)*255/N ? 255 : 0
                //cvThreshold( tgray, gray, (l+1)*255/N, 255, CV_THRESH_BINARY );
				cvThreshold( tgray, gray, 50, 255, CV_THRESH_BINARY );
            }
            
            // find contours and store them all as a list
            cvFindContours( gray, storage, &contours, sizeof(CvContour),
                CV_RETR_LIST, CV_CHAIN_APPROX_SIMPLE, cvPoint(0,0) );
            
            // test each contour
            while( contours )
            {
                // approximate contour with accuracy proportional
                // to the contour perimeter
				//用指定精度逼近多边形曲线 
                result = cvApproxPoly( contours, sizeof(CvContour), storage,
                    CV_POLY_APPROX_DP, cvContourPerimeter(contours)*0.02, 0 );
                // square contours should have 4 vertices after approximation
                // relatively large area (to filter out noisy contours)
                // and be convex.
                // Note: absolute value of an area is used because
                // area may be positive or negative - in accordance with the
                // contour orientation
                if( result->total == 4 &&
                    fabs(cvContourArea(result,CV_WHOLE_SEQ)) > 1000 &&	//cvContourArea计算整个轮廓或部分轮廓的面积 
                    cvCheckContourConvexity(result) )					//CheckContourConvexity
                {
                    s = 0;
                    
                    for( i = 0; i < 5; i++ )
                    {
                        // find minimum angle between joint
                        // edges (maximum of cosine)
                        if( i >= 2 )
                        {
                            t = fabs(angle(
                            (CvPoint*)cvGetSeqElem( result, i ),
                            (CvPoint*)cvGetSeqElem( result, i-2 ),
                            (CvPoint*)cvGetSeqElem( result, i-1 )));
                            s = s > t ? s : t;
                        }
                    }
                    
                    // if cosines of all angles are small
                    // (all angles are ~90 degree) then write quandrange
                    // vertices to resultant sequence 
                    if( s < 0.3 )
                        for( i = 0; i < 4; i++ )
                            cvSeqPush( squares,
                                (CvPoint*)cvGetSeqElem( result, i ));
                }
                
                // take the next contour
                contours = contours->h_next;
            }
        }
    }
    
    // release all the temporary images
    cvReleaseImage( &gray );
    cvReleaseImage( &pyr );
    cvReleaseImage( &tgray );
    cvReleaseImage( &timg );
    
    return squares;
}


// the function draws all the squares in the image
void drawSquares( IplImage* img, CvSeq* squares )
{
    CvSeqReader reader;
    IplImage* cpy = cvCloneImage( img );
    int i;
    
    // initialize reader of the sequence
    cvStartReadSeq( squares, &reader, 0 );
    
    // read 4 sequence elements at a time (all vertices of a square)
    for( i = 0; i < squares->total; i += 4 )
    {
        CvPoint* rect = pt;
        int count = 4;
        
        // read 4 vertices
        memcpy( pt, reader.ptr, squares->elem_size );
        CV_NEXT_SEQ_ELEM( squares->elem_size, reader );
        memcpy( pt + 1, reader.ptr, squares->elem_size );
        CV_NEXT_SEQ_ELEM( squares->elem_size, reader );
        memcpy( pt + 2, reader.ptr, squares->elem_size );
        CV_NEXT_SEQ_ELEM( squares->elem_size, reader );
        memcpy( pt + 3, reader.ptr, squares->elem_size );
        CV_NEXT_SEQ_ELEM( squares->elem_size, reader );
        
        // draw the square as a closed polyline 
        cvPolyLine( cpy, &rect, &count, 1, 1, CV_RGB(0,255,0), 3, CV_AA, 0 );
    }
    
    // show the resultant image
    cvShowImage( wndname, cpy );
    cvReleaseImage( &cpy );
}


void on_trackbar( int a )
{
    if( img )
        drawSquares( img, findSquares4( img, storage ) );
}
//char* names[] = { "D:/image/car/00.jpg", "D:/image/car/01.jpg", "D:/image/car/02.jpg",
//                  "D:/image/car/03.jpg", "D:/image/car/04.jpg", "D:/image/car/05.jpg", 0 };
//char* names[] = { "D:/image/car/car00.jpg", "D:/image/car/car01.jpg", "D:/image/car/car02.jpg",
//                  "D:/image/car/car03.jpg", "D:/image/car/car04.jpg", "D:/image/car/car05.jpg", 0 };
//char* names[] = { "D:/image/car/00sobel.jpg", "D:/image/car/01sobel.jpg", "D:/image/car/02sobel.jpg",
//                "D:/image/car/03sobel.jpg", "D:/image/car/04sobel.jpg", "D:/image/car/05sobel.jpg", 0 };
char* names[] = { "D:/image/car/06sobel_normal.jpg", 
				  "D:/image/car/0sobel_normal.jpg",
				  "D:/image/car/08sobel_normal.jpg",
                  "D:/image/car/09sobel_normal.jpg", 
				  "D:/image/car/10sobel_normal.jpg",
				  "D:/image/car/11sobel_normal.jpg",
				  "D:/image/car/12sobel_normal.jpg", 
				  "D:/image/car/13sobel_normal.jpg",
				  "D:/image/car/14sobel_normal.jpg",
                  "D:/image/car/15sobel_normal.jpg", 
				  "D:/image/car/16sobel_normal.jpg",
				  "D:/image/car/17sobel_normal.jpg",
				  "D:/image/car/18sobel_normal.jpg", 
				  "D:/image/car/19sobel_normal.jpg",
				  "D:/image/car/20sobel_normal.jpg",
                  "D:/image/car/21sobel_normal.jpg", 
				  "D:/image/car/22sobel_normal.jpg",
				  "D:/image/car/23sobel_normal.jpg",
				  "D:/image/car/00sobel_normal.jpg", 
				  "D:/image/car/01sobel_normal.jpg",
				  "D:/image/car/02sobel_normal.jpg",
                  "D:/image/car/03sobel_normal.jpg", 
				  "D:/image/car/04sobel_normal.jpg",
				  "D:/image/car/05sobel_normal.jpg",
				  0 };
int main(int argc, char** argv)
{
    int i, c;
    // create memory storage that will contain all the dynamic data
    storage = cvCreateMemStorage(0);

    for( i = 0; names[i] != 0; i++ )
    {
        // load i-th image
        img0 = cvLoadImage( names[i], 1 );
        if( !img0 )
        {
            printf("Couldn't load %s/n", names[i] );
            continue;
        }
        img = cvCloneImage( img0 );
        
        // create window and a trackbar (slider) with parent "image" and set callback
        // (the slider regulates upper threshold, passed to Canny edge detector) 
        cvNamedWindow( wndname,0 );
        cvCreateTrackbar( "canny thresh", wndname, &thresh, 1000, on_trackbar );
        
        // force the image processing
        on_trackbar(0);
        // wait for key.
        // Also the function cvWaitKey takes care of event processing
        c = cvWaitKey(0);
        // release both images
        cvReleaseImage( &img );
        cvReleaseImage( &img0 );
        // clear memory storage - reset free space position
        cvClearMemStorage( storage );
        if( c == 27 )
            break;
    }
    
    cvDestroyWindow( wndname );
    
    return 0;
}

#ifdef _EiC
main(1,"squares.c");
#endif

5.Sobel（横向求导，保留纵向纹理）+（颜色反向）+cvMorphologyEx（Close操作，IplConvKernel*（3x1）横向闭运算）+FindContours+cvBoundingRect+cvRectangle(满足一定条件)正确率65% 主要由于没有加入仿射变换或变形

#include "cv.h"
#include "highgui.h"
#include "cxcore.h"
#include <stdio.h>
#include <math.h>
#include <string.h>
#include <string>
using namespace std;


CvPoint pt[4];
IplImage* img = 0;
IplImage* img0 = 0;
const char* wndname = "Demo";

char* names[] = { "D:/image/car/06.jpg", 
				  "D:/image/car/07.jpg",
				  "D:/image/car/08.jpg",
                  "D:/image/car/09.jpg", 
				  "D:/image/car/10.jpg",
				  "D:/image/car/11.jpg",
				  "D:/image/car/12.jpg", 
				  "D:/image/car/13.jpg",
				  "D:/image/car/14.jpg",
                  "D:/image/car/15.jpg", 
				  "D:/image/car/16.jpg",
				  "D:/image/car/17.jpg",
				  "D:/image/car/18.jpg", 
				  "D:/image/car/19.jpg",
				  "D:/image/car/20.jpg",
                  "D:/image/car/21.jpg", 
				  "D:/image/car/22.jpg",
				  "D:/image/car/23.jpg",
				  "D:/image/car/00.jpg", 
				  "D:/image/car/01.jpg",
				  "D:/image/car/02.jpg",
                  "D:/image/car/03.jpg", 
				  "D:/image/car/04.jpg",
				  "D:/image/car/05.jpg",
				  0 };

void FindContours(IplImage* src);

int main(int argc, char** argv)
{
    int i;

    for( i = 0; names[i] != 0; i++ )
    {
        // load i-th image
        img0 = cvLoadImage( names[i], 0 );
        if( !img0 )
        {
            printf("Couldn't load %s/n", names[i] );
            continue;
        }
		img=cvCreateImage(cvSize(400,300),8,1);
		IplImage* pyr=cvCreateImage(cvSize(img->width/2,img->height/2),IPL_DEPTH_8U,1);
		cvResize(img0,img);
		cvNamedWindow("input",1);
		cvShowImage("input",img);
		cvSmooth(img,img,CV_MEDIAN);
		//cvPyrDown( img, pyr, 7 );
		//cvPyrUp( pyr, img, 7 );


        //img = cvCloneImage( img0 );
		IplImage* imgS=cvCreateImage(cvGetSize(img),IPL_DEPTH_16S,1);
		IplImage* imgTh=cvCreateImage(cvGetSize(img),IPL_DEPTH_8U,1);
		IplImage* temp=cvCreateImage(cvGetSize(img),IPL_DEPTH_8U,1);
		

        cvSobel(img,imgS,2,0,3);
        cvNormalize(imgS,imgTh,255,0,CV_MINMAX);

        cvNamedWindow( wndname,1);

		cvNamedWindow("Sobel",1);
		cvShowImage("Sobel",imgTh);


		//cvAdaptiveThreshold(imgTh,imgTh,255,0,0,5,5);
		cvThreshold( imgTh, imgTh, 100, 255, CV_THRESH_BINARY );
		
		for (int k=0; k<img->height; k++)

			for(int j=0; j<img->width; j++)

			{

				imgTh->imageData[k*img->widthStep+j] = 255 - imgTh->imageData[k*img->widthStep+j];

			}
			
		cvNamedWindow("Th",1);
		cvShowImage("Th",imgTh);
		IplConvKernel* K=cvCreateStructuringElementEx(3,1,0,0,CV_SHAPE_RECT);
		IplConvKernel* K1=cvCreateStructuringElementEx(3,3,0,0,CV_SHAPE_RECT);
		
		cvMorphologyEx(imgTh,imgTh,temp,K,CV_MOP_CLOSE,10);
		cvMorphologyEx(imgTh,imgTh,temp,K1,CV_MOP_OPEN,1);
		//cvDilate(imgTh,imgTh,K,15);
		//cvErode(imgTh,imgTh,K,15);
		cvShowImage(wndname,imgTh);
		string a=names[i];
		a.insert(15,"sobel_normal");
		//cvSaveImage(a.c_str(),imgTh);
		//cvWaitKey(0);
		FindContours(imgTh);

		//cvShowImage(wndname,imgTh);
		
		

		
	}
}

void FindContours(IplImage* src)
{
	CvMemStorage* storage = cvCreateMemStorage(0);
	IplImage* dst = cvCreateImage( cvGetSize(src), 8, 3);
	cvCvtColor(src,dst,CV_GRAY2BGR);
	CvScalar color = CV_RGB( 255, 0, 0);
	CvSeq* contours=0;
	
	 //建立一个空序列存储每个四边形的四个顶点
   // CvSeq* squares = cvCreateSeq( 0, sizeof(CvSeq), sizeof(CvPoint), storage );

	//cvFindContours( src, storage, &contours, sizeof(CvContour),CV_RETR_CCOMP, CV_CHAIN_APPROX_SIMPLE );//外界边界h_next 和 孔用v_next连接
	cvFindContours( src, storage, &contours, sizeof(CvContour), CV_RETR_CCOMP, CV_CHAIN_APPROX_SIMPLE );
	 for( ; contours != 0; contours = contours->h_next)
            {
				//使用边界框的方式
				CvRect aRect = cvBoundingRect( contours, 1 );
				int tmparea=aRect.height*aRect.height;
				if (((double)aRect.width/(double)aRect.height>3)
				&& ((double)aRect.width/(double)aRect.height<6)&& tmparea>=200&&tmparea<=2500)
			{
				cvRectangle(dst,cvPoint(aRect.x,aRect.y),cvPoint(aRect.x+aRect.width ,aRect.y+aRect.height),color,2);
				//cvDrawContours( dst, contours, color, color, -1, 1, 8 );
			}
		}
	 cvNamedWindow("contour",1);
	 cvShowImage("contour",dst);
	 cvWaitKey(0);
				//多边形曲线逼近方法
				/*
				//用指定精度逼近多边形曲线 
                result = cvApproxPoly( contours, sizeof(CvContour), storage,
                    CV_POLY_APPROX_DP, cvContourPerimeter(contours)*0.02, 0 );

				if( result->total == 4 &&
                    fabs(cvContourArea(result,CV_WHOLE_SEQ)) > 1000 &&	//cvContourArea计算整个轮廓或部分轮廓的面积 
                    cvCheckContourConvexity(result) )					//CheckContourConvexity
                {
				*/
	 
}


Augusdi
关注
专栏目录
Python 基于 OpenCV 视觉图像处理实战 之 OpenCV 简单人脸检测/识别实战案例 之五 简单进行车牌检测和识别
仙魁XAN
04-23 1182
Python是一种跨平台的计算机程序设计语言。是一种面向对象的动态类型语言，最初被设计用于编写自动化脚本(shell)，随着版本的不断更新和语言新功能的添加，越多被用于独立的、大型项目的开发。Python是一种解释型脚本语言，可以应用于以下领域： Web 和 Internet开发、科学计算和统计、人工智能、教育、桌面界面开发、软件开发、后端开发、网络爬虫。这里使用 Python 基于 OpenCV 进行视觉图像处理，......
使用opencv3.1.0识别图形进行精准定位
06-23
使用opencv对图像中的圆形抓取填充以及计算出圆形的周长面积，包括圆心的坐标非常实用用于工业摄像头定位方面
OpenCV之图像匹配与定位
learning attitude
12-25 1698
两张图片文件的像素分辨率和图像位置都发生了变化，如果要通过portrait.png去匹配定位它在QQ.png所在的坐标位置，自动化工具PyAutoGUI肯定是无法实现的。（2）根据两张图片的特征描述符（即变量descriptor）进行匹配，匹配算法有暴力匹配和FLANN匹配，不同的匹配算法所产生的匹配结果存在一定的差异。（4）在匹配结果里抽取中位数，利用中位数来反向查询它在目标图片所对应像素点的坐标位置，这个坐标位置也是自动化开发中使用的图片定位坐标。
苦学Opencv的第十二天：图形检测
最新发布
qq_71745595的博客
07-26 1208
轮廓，轮廓拟合，凸包，canny，霍夫变换
OpenCV（三）：一步步实现图像定位（Python版）
热门推荐
强强的博客
01-22 2万+
一、预期目标 如下图，要识别图中的国旗，然后框选出来，并且返回国旗的中心位置，效果如下: 彩色图像大小: (400,264) 目标中心位置: (225, 218) 二、准备工作 将 ...
计算机视觉与Python：利用OpenCV进行视觉定位和目标识别
程序员光剑
01-18 4561
1.背景介绍 计算机视觉是一种利用计算机处理和分析图像和视频的技术。它广泛应用于各个领域，如人脸识别、自动驾驶、物体检测等。OpenCV(Open Source Computer Vision Library)是一个开源的计算机视觉库，提供了大量的功能和工具，帮助开发者快速实现计算机视觉任务。本文将介绍如何使用OpenCV进行视觉定位和目标识别。 2.核心概念与联系 2.1 视觉定位 视觉...
opencv实现图像叠加+坐标定位
05-10
opencv实现图像叠加+坐标定位，初步实现旋转
车牌识别_边缘检测_opencv_opencv车牌识别_opencv识别车牌_车牌识别
06-21
总结来说，本项目展示了如何结合OpenCV的Canny边缘检测和颜色识别功能，实现车牌的自动检测与识别。通过理解并实践这个过程，可以提升在计算机视觉领域的技能，特别是对于图像处理和模式识别的应用。
MFC_OPENCV-chepai.rar_opencv 车牌_opencv 检测车牌_opencv 车牌识别_车牌 open
07-14
源代码中的MFC_OPENCV车牌检测部分，可能是具体实现车牌检测的函数或类。通过阅读和理解这段代码，我们可以学习到如何在实际项目中运用OpenCV进行车牌识别。 总结，OpenCV作为一个强大的图像处理库，提供了丰富的...
opencv车牌识别、车牌识别是⼀种图像处理技术
03-08
1. **车牌检测**：这是识别流程的第一步，目的是在图像中定位车牌的位置。OpenCV中的轮廓检测可以用来寻找潜在的车牌轮廓。如果对车牌的尺寸、颜色和大致位置有先验知识，可以通过调整算法参数提高检测精度。此外，...
python+opencv实现车牌定位功能(实例代码)
12-23
在计算机视觉领域，车牌定位是一项常见的任务，它涉及到图像处理和模式识别技术。本实例展示了如何利用Python和OpenCV库来完成这一任务，特别适合初学者理解基本的图像处理流程。下面将详细解释代码中的关键步骤： ...
车牌检测和识别opencv代码
03-12
车牌识别代码，采用C++和OPENCV2实现，详细实现过程可以参见我的博客车牌识别文章。训练所用的数据科查看我的另外一个资源。
OpenCV基础知识（8）— 图形检测（图像轮廓、轮廓拟合、Canny边缘检测和霍夫变换）
突然好想你
08-26 3420
图形检测是计算机视觉的一项重要功能，本节课就介绍如何检测图像的形状、图像所占的区域，以及如何查找图像中出现的几何图形等！
OpenCV单目视觉定位实现方法
weixin_43956732的博客
08-05 4880
OpenCV单目视觉定位（测量）系统 The System of Vision Location with Signal Camera Abstract:This passage mainly describes how to locate with signalcamera,which bases on OpenCV library. Key words: OpenCV; Locate;Signalcamera 摘  要：本文主要描述的是利用开源计算机视觉库OpenCV实现单目视觉定位系统。
学习OpenCV——车牌检测(定位)
龙在天涯
11-07 6497
这两天在做关于车牌识别的实验，用了几种方式： 1.车牌颜色分布（HSV空间，YCrCb空间的没有颜色分布图谱，无法实验）；利用HSV的H通道，效果一般，受环境影响大。 转载自：http://blog.csdn.net/sangni007/article/details/7444470 [cpp] view plaincopyprint? #include "high
08- OpenCV进行图像的特征检测 (OpenCV系列) (机器视觉)
March_A的博客
01-08 1685
推荐Shi-Tomasi角点检测Harris角点检测计算的稳定性和K有关, 不稳定, shi-tomasi是优化版)。三种特征检测算法对比:SIFT 最慢, 准确率最高SURF 速度比SIFT快些, 准确率差些ORB速度最快, 可以实时检测, 准确率最差.两种特征匹配方式:BFMatcherFLANN(处理大数据集时效果较好)暴力特征匹配单应性矩阵的应用, 查找对应图形:(图像查找)直接读取为灰度图Harris角点检测shi-tomasi角点检测坐标画点cv2.circle。
【计算机视觉】反向投影直方图检测特定的图像内容
ZhangPY的专栏
11-14 1990
【计算机视觉】反向投影直方图检测特定的图像内容标签（空格分隔）： 【图像处理】原理：获取特定的图像内容直方图，并进行归一化，然后将输入图像的某一位置上像素值（多维或灰度）对应在直方图的一个bin上的值来代替该像素值，所以得到的反向投影图是单通的。举个小例 例如灰度图像如下 Image=0 1 2 3 4 5 6 7 8 9 10 11
OpenCV项目实践——物体检测与跟踪
欢迎来到“Python算法探索”，我是GT算法工程师，一名热爱计算机算法的博主。在这里，我们将一起深入探索Python世界中的各种算法，从基础到高级，从理论到实践。我参加过很多算法比赛，积累了丰富的经验，希望通过这个播客与你分享。让我们一起用Python探
01-26 978
本文将带领你完成一个基于OpenCV的物体检测与跟踪项目。我们将使用OpenCV的强大功能，从基础物体检测到动态跟踪，逐步深入探索计算机视觉的奥秘。通过这个项目，你将了解物体检测与跟踪的基本原理、方法和技术，并亲自动手实现一个实用的物体检测与跟踪系统。
基于C++的数字图像处理系统：车牌定位实践与OpenCV应用
"本文主要探讨了利用Qt集成开发环境，结合C++编程语言和OpenCV库，开发一个数字图像处理系统的过程，特别是车牌定位的应用。文章详细阐述了系统设计、需求分析、算法实现以及结果分析，旨在为图像处理研究者提供一个...
评论 2
添加红包

请填写红包祝福语或标题

个

红包个数最小为10个

元

红包金额最低5元

当前余额3.43元 前往充值 >
需支付：10.00元
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付元
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明：

1.余额是钱包充值的虚拟货币，按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载，可以购买VIP、付费专栏及课程。

余额充值