拼接两幅图像
Step #1: Detect keypoints (DoG, Harris, etc.) and extract local invariant descriptors (SIFT, SURF, etc.) from the two input images.
Step #2: Match the descriptors between the two images.
Step #3: Use the RANSAC algorithm to estimate a homography matrix using our matched feature vectors.
Step #4: Apply a warping transformation using the homography matrix obtained from Step #3.
参考资料
https://www.pyimagesearch.com/2016/01/11/opencv-panorama-stitching/
近期评论