Matching of terrestrial laser intensity image and optical image

摘要

The matching of terrestrial laser intensity image and optical image is one of the active research areas in terrestrial laser scanning technique. In this paper, we introduce an algorithm based on gray scale to match intensity image and optical image. The similarity measure and the size of matching template are two crucial issues of this algorithm. Firstly, the correlation coefficient based on gray values and the mutual information based on statistical distribution are combined. Thus an integrated criterion is produced to extract corresponding points in an intensity image and an optical image. Secondly, an experiment was done to perceive the relations between the correlation coefficient and the mutual information of corresponding points and the size of a matching template. This image matching algorithm has two steps, coarse image matching and fine image matching. The coarse image matching step is carried out to search three precise pairs of corresponding points to calculate a mapping polynomial, by which a primary search region can be obtained for reducing the possibility of false correspondence and increasing the speed of image matching. While the fine image matching step aims at obtaining more precise corresponding points in the region determined by the mapping polynomial.