Research on Feature Matching Algorithm Based on Improved AKAZE
Firstly,aiming at the poor matching performance of AKAZE algorithm under image blur changes and scale chang-es,the improved BRISK descriptor is used to replace the original AKAZE algorithm by reducing the sampling points of BRISK de-scriptor and optimizing the sampling points selection mechanism.Secondly,in order to improve the problem of AKAZE algorithm with many false matching points and poor matching speed,the selection mechanism of high quality points by RANSAC algorithm is added,and the improved RANSAC algorithm is combined with two stages to remove false matching points.Experimental results show that the proposed algorithm has better matching effect under scale and fuzzy changes,and the accuracy and speed of feature point matching are better than AKAZE algorithm.
万方数据
维普
