Image Matching Algorithm Based on Improved FAST and RANSAC

Abstract

Aiming at the problems of large amount of calculation, low matching accuracy and long matching time in image matching in visual positioning system, this paper proposes an image matching algorithm based on improved FAST (MFAST) and RANSAC (P-RANSAC). First, the multi-level FAST algorithm is used to extract the corner points, and the SURF algorithm is used to determine the main direction to generate the feature descriptor; then the fast approximate nearest neighbor algorithm is used to complete the rough matching of the feature points. Use the pre-sampling algorithm to select a new sample set for sampling and test the calculated model and discard the incorrect model parameters. Experimental results show that the proposed algorithm can effectively improve the accuracy and real-time performance of image matching compared with traditional algorithms.