Multi-Source Remote Sensing Imageries Matching Based on SIFT Feature with Match-Support Measurement

In this paper, an automatic image match algorithm based on SIFT features with match-support measure is presented for mutlsource remote sensing images. In order to adjust SIFT algorithm applied in the matching processing for different-source remote sensing images, we introduce the match support measure for similarity measure. Firstly, it builds SIFT feature descriptor and selects the points satisfied the minimum Euclidean distance for candidate match result between reference image and match image. Afterward, it calculates the match-support measure among the candidates separately. Finally, it employs the relaxation method to discard the false matching pairs. We used the two groups of different source remote sensing images for image match experiment, which have shown the improvement in image matching processing with our algorithm.