Research on Image Matching Method Based on ORB and Grid Motion Statistics

The ORB algorithm has a small amount of calculation and can quickly abstract numerous feature points. The grid motion statistics algorithm can filter the initial match, reduce false matches, and retain more inliers. Combining ORB and grid motion statistics algorithm to complete the indoor matching of two adjacent frames of pictures. Proved by experiment: The algorithm used in this paper can eliminate mismatched point pairs and retain more inliers. Compared with the ORB+RANSAC algorithm, the inliers ratio is increased by 7.62% on average. The algorithm in this paper can be applied to the indoor positioning of vision-based mobile robots and has strong robustness in the feature point matching of image rotation and scaling.