Research on registration algorithm based on feature points for human eye retina images
Yuran Liu, Huizhen Yang, Liyun Su, Yudong Zhang, Xuejun Rao · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2009
As a rule, according to the characteristics of living retinal images, such as gray and resolution unconsistency, some differences in detail, etc, it is a fairly difficult problem to carry out images stitching. In this paper, a robust image registration method based on feature points is proposed. To begin with, use RANSAC method to make rough match for the corner. Then, Mahalanobis distance is applied in the accurate purification. Finally, execute the image stitching with the purified corner. Practical experiments and application show that the proposed approach is feasible and effective.