A tiny facet primitive remote sensing image registration method based on SIFT key points

Abstract

Multisensory remote sensing images were achieved with different sensors in different environments; it can result in the ground information difference of the same area. So the high accurate image registration has become a pivotal hotpot problem of remote sensing image fusion research. In order to solve the remote sensing image registration problem with high accurate image registration method, a tiny facet primitive image registration method based on scale invariant feature transform (SIFT) key points is presented in this paper. By introducing the conventional tiny facet primitive method, the SIFT method can rationally solve the detection problem of the registration control points (RCPs) with the key points. Empirical results show that the detection process is with higher performance. The SIFT-based tiny facet primitive method can effectively solve the detection of RCPs and it proved to be an effective method for remote sensing image fusion and high accurate image registration.