SAR image registration based on KECA-SAR-SIFT operator

Synthetic aperture radar (SAR) image registration is a key technology in SAR image processing. The accuracy and efficiency of registration directly affect the quality of subsequent image processing. In order to further improve the accuracy and computational efficiency of the traditional SAR-SIFT (SAR-scale invariant feature transform, SAR-SIFT) image registration algorithm, an improved SAR-SIFT algorithm based on Kernel entropy component analysis (KECA) is proposed. Firstly, the SAR-Harris scale space is established, the extreme points in the space are selected and the main directions of the key points are calculated; then, the SAR-SIFT descriptor is generated, and the KECA algorithm is used for further feature extraction, and the KECA-SAR-SIFT descriptors is obtained after dimensionality reduction; finally, feature points of two or more pictures are matched by comparing the similarity of the extracted descriptors. The experimental results showed that the proposed method effectively improved the matching accuracy, shortened the matching time, and had certain robustness.