Research on Feature Extraction and Matching Algorithm in Visual SLAM

Based on the application of visual SLAM, an improved ORB feature extraction and matching algorithm is proposed to overcome the shortcomings of the original ORB algorithm. In order to improve the stability of the scale, a multi-scale spatial pyramid is established to detect and extract the FAST feature points of each grid in different scales, so as to improve the stability of the scale; the quadtree method is used to divide the feature points to make the distribution more uniform; after the coarse matching, the PRASAC method is used to eliminate the wrong matching points, which improves the matching accuracy. Experimental results show that the algorithm can effectively improve the matching accuracy and efficiency, and meet the requirements of robustness and real-time in visual SLAM.