DOI: 10.21276/sjpms.2018.5.4.3

Pages: 231-235

Aiming at the problem that the RANSAC algorithm has high computational complexity and low efficiency in image matching. A fast coarse-to-fine strategy is introduced in this paper: the RANSAC algorithm has been used in the inliers sub-set to estimate the point transformation matrix between two images accurately by using least square method. It filters the noisy or wrong input data iteratively, and then estimates the model parameters through pure data, so it has strong error-tolerant capacity for the image noise and inaccuracy of feature extraction.