Automatic Registration of Multi-Source Imagery Based on Global Image Matching.

Image fusion is important in the synthetic application of multisource remotelv sensed images, and image registration is the basis of image fusion. The traditional method of manual registration is a laborious, tedious, and complex task. For improving the efficiency of image fusion, automatic methods of image registration must be used. In this paper, automatic registration based on the theory and methods of image matching is presented. The main contents include extraction of single point matching, reliability strategy based on a hierarchical pyramid image structure, and global image matching. The automatic registration of multi-source remote sensing images is then discussed, the presentation being based on global image matching. Finally, the successful results of the automated registration and fusion of TM and SPOT images are presented. [ABSTRACT FROM AUTHOR]