坐标网格立体视觉测量中的图像匹配技术

利用立体视觉技术，实现对板料成形分析中应用的大批量坐标网格的变形进行无接触自动测量与分析，以减少工作量并提高测量的速度和精度．主要针对圆形坐标网格变形的立体视觉测量中的图像匹配技术进行了研究．在对网格图像进行特征提取与编码的基础上，实现网格单元的子域匹配，并通过对摄像位置进行预约束，建立起左右对应像素点纵向图像坐标间的约束关系，从而实现左右图像特征点的快速准确匹配方法．该方法程序实现简单，匹配结果可靠，无二义性，可推广应用于其他具有相似结构特征的三维图像的立体匹配中．

Image Matching Technology in Coordinate Grids Measurement Based on Stereo Vision

Based on stereo vision technology, the non contact and automatic measurement of large quantity of deformed coordinate grids which is used widely in sheet metal forming analysis was implemented. A new approach was proposed to realize the rapid and accurate image matching in stereo measurement especially for circular grids. The sub region matching is based on the feature abstraction and coding and the feature point matching is implemented by applying constrains to camera modeling to construct the coordinate value relation between two corresponding pixels. The results show that the method is reliable and can be further used for stereo matching of other structured images.