利用编码和极线约束相结合的方法实现工业摄影测量中的点匹配  

Point Matching in Industrial Photogrammetry with Coded Point and Epipolar Constr aint

作　　者：董明利[1] 齐晓娟[2] 吕乃光[2] 王永强[2] 潘志康[2] 祝连庆[2] 

机构地区：[1]北京理工大学 [2]北京机械工业学院

出　　处：《工具技术》2006年第4期73-75,共3页Tool Engineering

基　　金：国家自然基金资助项目(项目编号:50475176);北京市自然科学基金资助项目(项目编号:KZ200511232019);北京市优秀人才培养专项经费资助项目(项目编号:20041D0500601)

摘　　要：提出一种利用编码和极线约束相结合实现点匹配的方法,该方法利用圆环式编码点作为识别码标识区域。此类编码点具有旋转、缩放、变形的无关性,测量点是与编码点中心圆斑大小一样的圆斑。针对大尺寸测量对象中点的匹配问题,本文提出了分区的思想。首先进行区域匹配,即粗匹配;然后利用极线约束匹配小区域测量点,即细匹配,这样就完成了整个匹配。试验表明该方法能够有效地降低数据处理时间,提高匹配率和自动化程度。A point matching method with coded points and epipolar constrain is described. The circular ceded points are known as identifiers that are used to identify distracts. Identifying the ceded points is not affected when the coded points come forth rotation, zooming and deformation. The radius of the measured point is the same as the radius of the ceded points circle. To solve the point matching in huge shape measurement, the idea of the plotting area is presented. First, the subareas are matched, namely, crassitude matching, then the measured points are matched with epipolar constraint in the subareas of left and right images. The resuh shows that this method is valid and reliable in lowering data processing time, improving image matching veracity and au- tomatic degree.

关 键 词：编码点 极线约束 图像匹配 摄影测量 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TP391.41[自动化与计算机技术—计算机科学与技术]