摘要
针对数字近景摄影测量中的人工标志点快速自动匹配问题,提出了一种基于双片空间前方交会的匹配算法。首先,通过计算投影线间的最短距离确定一组初始匹配点;然后通过双片空间前方交会计算相应物方点坐标;最后反求该物方点坐标在其他像片上的像点坐标,通过比较该像点与初始匹配点的坐标差确定同名像点。两组实验均证明,该算法计算速度快,具有高匹配率和低误匹配率。
A new matching method based on two-image space intersection was developed, aiming to the fast automatic matching of artificial target points in digital close range photogrammetry. Firstly, an initial matched points group was found by calibrating the shortest distance between two projecting beams. Secondly, the coordinate of corresponding object point was calculated through two-image space intersection. Finally, we got image point coordinates of the object point in the other images, and fixed on the homologous points by comparing the coordinate differences between the calculated image point and initial points. Both experiences proved the advantage of the matching method, which was at high speed, with high matching quotiety and low miss matching quotiety.
出处
《测绘科学技术学报》
北大核心
2008年第1期32-34,共3页
Journal of Geomatics Science and Technology
关键词
数字近景摄影测量
匹配
双片空间前方交会
digital close range photogrammetry
matching
two-image space intersection
