针对不同传感器、不同时相、不同分辨率的异源遥感影像匹配困难的问题,引入尺度不变特征(Scale Invariant Feature Transform,SIFT)算法;针对传统SIFT匹配算法的不足,利用SIFT特征向量匹配对的唯一性约束改进传统SIFT算法的匹配策略,采用双向匹配以达到在匹配过程中准确寻找匹配点对的目的,提高影像匹配的正确率,实验证明,该方法适用于异源影像匹配。

According to the dii culty matching the multi-source remote sensing images since dif erent sensors, dif erent temporal and dif erent resolution, a new matching algorithm based on SIFT was proposed. The match constraint was proposed to make up for the insui cient of SIFT based enunidirectional match algorithm. And a bidirectional matching strategy on its optimization was used to improve the matching reliability. The experimental results demonstrate that this approach is ef ective and can be used for multi-source remote sensing images matching.