多种匹配算法的结合在影像全自动内定向中的应用

The Application of the Combination of Several Matching Algorithms in the Automatic Interior Orientation of Images

本文首先将两种匹配算法即金字塔分层匹配和基于遗传算法的影像匹配运用于全自动内定向中框标点的预匹配 ,再利用最小二乘匹配进行框标点的精匹配。为减小搜索范围 ,提高定向速度 ,本文还提出了一种新的框标点概略定位方法。实验证明该算法自动化程度高、快速。

This paper applies two matching algorithms in the initial matching of the frame point of automatic interior orientation and LSM is used for the precise matching of frame points. In order to diminish the range of searching and improve the speed of orientation, this paper provides a new method of orienting frame points roughly. Experiments evince this method is very automatic, fast and effective.