Remote sensing image registration with spatial restraint based on moment invariants and fast generalized fuzzy clustering

摘要

Image registration is a key component in remote sensing image processing. In this paper, we present a remote sensing image registration method by incorporating spatial restraint based on moment invariants and fast generalized fuzzy clustering. Seven moment invariants are extracted as features of objects obtained by the fast generalized fuzzy c-means (FGFCM) algorithm. The objects are matched through these features. Then, we detect the keypoints in corresponding matching regions. Through the spatial restraint, the outliers are removed and the correct matches are increased. The proposed algorithm is evaluated on multi-spectral images, multi-temporal images, and multi-sensor images. Extensive experimental studies prove that the proposed algorithm is promising.