Feature extraction and stereo matching algorithm for lunar surface

Abstract

According to the characteristics of the lunar surface and the application of the matching results, a stability and robust feature extraction and stereo matching algorithm for lunar surface was presented. In order to improve the efficiency of the subsequent stereo matching, firstly, a compact algorithm based on the camera parameters for rectification of stereo images was used. Then the CLAHE algorithm was studied to enhance each region's contrast, which make it more conducive to feature extraction and stereo matching. Based on the above results, the invariant SIFT extraction algorithm for lunar surface environment was proposed and improved according to the characteristics of images. At last, a matching algorithm was executed on a pair of SIFT feature corner extracted stereo image, and some commonly constraints in stereo matching were studied to eliminate pseudo matching of features. Experimental results indicate that the algorithm performs well when used in the images of simulated lunar surface terrain environment.