This paper gives a new method for image rectification. The method is based on an estimation of the epipolar constraints and homography matrix H, which de-scribes the relationship of the corresponding epipolar lines.The approach makes the resampling images extremely sim-ple by using Bresenham Algorithm to extract pixels alongthe corresponding epipolar line. For a large set of cameramotions, remapping to a plane has the drawback of cre-ating rectified images that are potentially infinitely largeand presents a loss of pixel information a long the epipolar lines. In contrast, our method guarantees that the recti-fied images are bounded for all possible camera motions and minimizes the loss of pixel information along epipo-lar lines. Excellent experimental results obtained with a binocular stereovision images are presented and detailed analysis is provided.
展开▼
