Synthetic aperture radar image matching based on improved scale-invariant feature transform

Abstract

In this paper, an improved scale-invariant feature transform (SIFT) algorithm for synthetic aperture radar (SAR) image matching is proposed. Initially, feature descriptors based on gradient ratio (GR) are constructed by utilizing traditional SIFT method. In order to measure the matching degree between images, the similarities of the descriptors are then calculated via the symmetry Kullback-Leibler divergence (SKLD) criterion. Experimental results show that the proposed method has a good performance on image matching for SAR images.