Video Sequence Image Matching Algorithm Based on Adaptive Gamma Correction

Abstract

Aiming at the low matching accuracy and poor robustness in video sequence image matching, a video sequence image matching algorithm based on Adaptive Gamma Correction is proposed in the paper. Firstly, Adaptive Gamma Correction is used to enhance the contrast of video sequence images. Then feature extraction and description are realized by the ORB algorithm. The RANSAC algorithm is optimized in matching by improving the interior point ratio after Brute Force Match. The experimental results show that the algorithm proposed improves the correct matching rate by 19.51%, 12.53%, and 4.87% respectively compared with the SIFT, ORB, and ORB+RANSAC algorithms without increasing the running time. Meanwhile, the algorithm in this paper has good robustness in video sequence image matching.