提出了一种融合SURF算法和SC—RANSAC算法的图像配准方法。首先利用SURF算法提取待匹配图像的特征,然后用最近邻方法找出匹配点,最后运用SC—RANSAC算法剔除错误的匹配点,实现图像的正确配准。实验结果表明,该方法在保持较高的特征点正确匹配率的前提下,配准速度高于SURF和RANSAC相结合的方法和SIFT和RANSAC相结合的方法。
This paper utilized a hybrid algorithm combined SURF and SC-RANSAC together. This algorithm could be utilized to do image registration. It utilized SURF algorithm to extract image features, and then utilized a nearest-neighbor-search method to find the matching points of two images. At last, it utilized SC-RANSAC algorithm to get rid of the error matches then the results were used to do image registration. Experiments show that the proposed algorithm has a higher matching feature points, and runs faster than algorithm combining SURF or SIFT with RANSAC.