TPS-SURF-SAC matching approach of feature point applied to deformation measurement of nonrigid tissues from MR images

Due to the nonlinear deformation of the nonrigid and nonuniform biological tissues, it is difficult whereas important to correctly match a number of feature points distributed somewhat uniform in the tissues from MR images for deformation measurement. In this paper, the authors present TPS-SURF-SAC matching method and mismatching elimination method based on TPS clustering. Firstly the matching region is identified by a TPS for every query point. Then the SURF descriptors and the proposed Spatial Association Correspondence (SAC) method are combined to match the feature points. Finally, using clustering the coordinate differences between the matching points obtained using TPS-SURF-SAC method and the matching points matched by TPS model, most of wrong match points are eliminated. After every iterative processing of matching and mismatching elimination, the updated TPS model becomes more accurate and more correctly matched points can be identified than that of the previous iteration. The experimental results showed that the proposed method outperformed the single SURF and SIFT methods.