Accurate Feature Point matching Based on Affine Iterative Model

A highly accurate feature point matching method for stereo was proposed. It was proved that the homography between corresponding feature windows can be approximated by an affine model. Then the projective distortion of corresponding feature windows was estimated and rectified by a fast iterative scheme, at the same time, the location error of corresponding feature points arising at the detection stage was computed and compensated. The matching results achieve subpixel precision which leads to a highly accurate epipolar geometry. Experimental results of real images and the comparisons with other methods strongly demonstrate the validity and accuracy of our method