Automatic Matching Method of Control Points for AVHRR Remote Sensing Image
Ying Jie Xia, Linjun Zhu, Xiaobo Luo, Hae-Young Bae · Advanced Science and Technology Letters · 2014
Remote sensing image automatic registration technology is one of the image processing technologies that have been developed rapidly in recent years, and automatic matching of control points is the core work of the automatic registration. Considering the influence of the cloud points and the phenomenon of uneven distribution of control points in AVHRR images, An automatic matching method of control points is proposed in this paper. In this method, the cloud points are detected first, and then feature points are extracted in the overlapping region of the two images. Finally, matching is performed based on image blocks method and the results of cloud point detection. Experimental result shows that this method not only can obtain even distributed control point pairs and avoid the interference of cloud points, but also can improve the accuracy of registration of AVHRR remote sensing images