OpenCV依据颜色的车牌定位

使用颜色属性：

Mat srcImage=imread("image/t10.jpg");
	Mat srcShowImage;
	srcImage.copyTo(srcShowImage);
	//imshow("a",srcImage);
	int i,j;
	int cPointB,cPointG,cPointR;
	for(i=1;i<srcImage.rows;i++)
		for(j=1;j<srcImage.cols;j++)
		{
			cPointB=srcImage.at<Vec3b>(i,j)[0];
			cPointG=srcImage.at<Vec3b>(i,j)[1];
			cPointR=srcImage.at<Vec3b>(i,j)[2];
			if(cPointB>80&cPointR<80&cPointG<80)    //提取蓝色，将该区域设置为黑色
			{
				srcImage.at<Vec3b>(i,j)[0]=0;
				srcImage.at<Vec3b>(i,j)[1]=0;
				srcImage.at<Vec3b>(i,j)[2]=0;
			}

			else if(cPointB>200&cPointR>200&cPointG>200)  //提取白色，将其设置为黑色
			{
				srcImage.at<Vec3b>(i,j)[0]=0;
				srcImage.at<Vec3b>(i,j)[1]=0;
				srcImage.at<Vec3b>(i,j)[2]=0;
			}

			else
			{
				srcImage.at<Vec3b>(i,j)[0]=255;
				srcImage.at<Vec3b>(i,j)[1]=255;
				srcImage.at<Vec3b>(i,j)[2]=255;
			}

		}
		cvtColor(srcImage,srcImage, CV_BGR2GRAY);  
		threshold(srcImage,srcImage,127, 255,CV_THRESH_BINARY);   
		//使用差分法，去掉不相关的区域。
		for(i=1;i<srcImage.rows;i++)
			for(j=1;j<srcImage.cols-1;j++)
			{
				srcImage.at<uchar>(i,j)=srcImage.at<uchar>(i,j+1)-srcImage.at<uchar>(i,j);

			}

			threshold(srcImage,srcImage,127, 255,CV_THRESH_BINARY_INV);//通过二值化的方式来取反。
			//erode(srcImage,srcImage,Mat(5,5,CV_8U),Point(-1,-1),2);  //腐蚀
			//	dilate(src,src,Mat(5,5,CV_8U),Point(-1,-1),2); //膨胀
			//	morphologyEx(src,src,MORPH_OPEN,Mat(3,3,CV_8U),Point(-1,-1),1);   //开运算
			//	 morphologyEx(src,src,MORPH_CLOSE,Mat(3,3,CV_8U),Point(-1,-1),1);  //闭运算
			erode(srcImage,srcImage,Mat(3,3,CV_8U),Point(-1,-1),5);
			threshold(srcImage,srcImage,127,255,CV_THRESH_BINARY_INV);
			imshow("a",srcImage);
			vector<vector<Point> > contours;
			vector<Vec4i> hierarchy;
			findContours(srcImage, contours, hierarchy, CV_RETR_EXTERNAL, CV_CHAIN_APPROX_SIMPLE, Point(0, 0) );
			Scalar color = Scalar( rng.uniform(0, 255), rng.uniform(0,255), rng.uniform(0,255) );
			for( int i = 0; i < contours.size(); i++ )
			{  
				//使用边界框的方式  
				CvRect aRect =  boundingRect(contours[i]);
				int tmparea=aRect.height*aRect.height;  
				if (((double)aRect.width/(double)aRect.height>2)&& ((double)aRect.width/(double)aRect.height<6)&& tmparea>=2000&&tmparea<=25000)  
				{  
					rectangle(srcShowImage,cvPoint(aRect.x,aRect.y),cvPoint(aRect.x+aRect.width ,aRect.y+aRect.height),color,2);  
					//cvDrawContours( dst, contours, color, color, -1, 1, 8 );  
				}  
			} 


			imshow("da",srcShowImage);

颜色可以考虑更细致，或者考虑在其他颜色空间内实现。

OpenCV依据颜色的车牌定位,布布扣,bubuko.com

OpenCV依据颜色的车牌定位

原文：http://blog.csdn.net/superdont/article/details/24936341