OpenCV实现拼接图像的简单方法
本文实例为大家分享了OpenCV实现拼接图像的具体方法,供大家参考,具体内容如下
用iphone拍摄的两幅图像:
拼接后的图像:
相关代码如下:
//读取图像 Mat leftImg=imread("left.jpg"); Mat rightImg=imread("right.jpg"); if(leftImg.data==NULL||rightImg.data==NULL) return; //转化成灰度图 Mat leftGray; Mat rightGray; cvtColor(leftImg,leftGray,CV_BGR2GRAY); cvtColor(rightImg,rightGray,CV_BGR2GRAY); //获取两幅图像的共同特征点 int minHessian=400; SurfFeatureDetector detector(minHessian); vector<KeyPoint> leftKeyPoints,rightKeyPoints; detector.detect(leftGray,leftKeyPoints); detector.detect(rightGray,rightKeyPoints); SurfDescriptorExtractor extractor; Mat leftDescriptor,rightDescriptor; extractor.compute(leftGray,leftKeyPoints,leftDescriptor); extractor.compute(rightGray,rightKeyPoints,rightDescriptor); FlannBasedMatcher matcher; vector<DMatch> matches; matcher.match(leftDescriptor,rightDescriptor,matches); int matchCount=leftDescriptor.rows; if(matchCount>15) { matchCount=15; sort(matches.begin(),matches.begin()+leftDescriptor.rows,DistanceLessThan); } vector<Point2f> leftPoints; vector<Point2f> rightPoints; for(int i=0; i<matchCount; i++) { leftPoints.push_back(leftKeyPoints[matches[i].queryIdx].pt); rightPoints.push_back(rightKeyPoints[matches[i].trainIdx].pt); } //获取左边图像到右边图像的投影映射关系 Mat homo=findHomography(leftPoints,rightPoints); Mat shftMat=(Mat_<double>(3,3)<<1.0,0,leftImg.cols, 0,1.0,0, 0,0,1.0); //拼接图像 Mat tiledImg; warpPerspective(leftImg,tiledImg,shftMat*homo,Size(leftImg.cols+rightImg.cols,rightImg.rows)); rightImg.copyTo(Mat(tiledImg,Rect(leftImg.cols,0,rightImg.cols,rightImg.rows))); //保存图像 imwrite("tiled.jpg",tiledImg); //显示拼接的图像 imshow("tiled image",tiledImg); waitKey(0);
以上就是本文的全部内容,希望对大家的学习有所帮助,也希望大家多多支持我们。
赞 (0)