一种适应地形起伏的误匹配粗差剔除方法

Method of Eliminating the Gross Error Adapted to the Fluctuant Topography

为了实现更高精度的图像匹配,综合利用SIFT算法和APERS算法各自的优势,提出了一种可以剔除图像匹配粗差的改进的APERS算法。该算法先采用SIFT算法提取特征点并进行粗匹配,再采用改进的APERS算法剔除误匹配点对,解决了SIFT算法中存在的误匹配等问题,通过与RANSAC算法的对比实验验证了该算法的有效性。

In order to achieve the higher accuracy of image matching, the paper puts forward the improved APERS algorithm combined with the advantage of SIFT and APERS. The APERS algorithm uses SIFT algorithm to extract feature points and complete the coarse matching, then it removes the error matching points and solves the problems like error matching. The effectiveness of the improved algorithm is verified by the contrast experiment with the RANSAC algorithm.