视觉传感集成体系中不同图像间映射匹配关系

Mapping and Matching Among the Different Images in a Vision Sensing Integrated System

提出一种由三角法面形传感器和 2个 CCD摄像机组成的机器视觉传感集成系统 ,利用前者所获得的三维面形与后者所摄得的二维图像在空间上特定的几何对应关系 ,建立 3幅图像间的映射。可实现在二维 CCD图像上附加距离数据 ,及以三维面形数据为网格点粘贴亮度、颜色和纹理等信息 ;并进一步可实现 2幅二维图像基于映射小区域间的匹配 ,用于双 CCD摄像机交汇测量。保证了三维面形与二维图像间的信息融合和互补 ,以满足机器视觉需要。

A machine vision integrated sensor, which includes one 3-D profilometer based on a synchronized laser scanner and two CCD cameras, has been presented. The mapping relationship among the 3-D surface profile and two 2-D CCD images was developed by means of their special geometric correspondence in space. Therefore range information from the 3-D surface data could be added to the corresponding small area on two 2-D CCD images, and meanwhile the intensity, color and vein information from the CCD images could also be affixed to the grid of 3-D surface profile, moreover the matching relation between two CCD images could be achieved through their small area mapping to the common pixel on the 3-D image so that the intersection measurement based on the two CCD images could be realized. Data fusion and information complementation among 3-D surface profile and 2-D CCD images can be ensured in order to meet the need of machine vision requirement.