用于立体图像矫正:Harries角点、NCC匹配、RANSAC计算基本矩阵完成立体图像对的极线校正，自己书写的opencv函数-For three-dimensional image correction: Harries corner NCC matching, RANSAC calculation of the completion of the fundamental matrix the epipolar rectification of the stereo image pairs o

图像自动分割算法，可以从简单背景中分割出长方形的兴趣图片，并提供方向校正和仿射较正。可以用于名片分割，试卷分割，所有长方形目标的分割。-And automatic image segmentation:, which can segment the rectangular object the simple background image, and provide direction correction and affine rectification. This algorithm c

3篇双目视觉的英文文献，利用8点法求出立体视觉系统的基本矩阵，进而求出8对点对应的极点和极限- A New Rectification Scheme for Uncalibrated Stereo Image Pairs and Its Application to Intermediate View Reconstruction  Planar Affine Rectification Change of Scale  A Study