In the process of visual Simultaneous Localization and Mapping (SLAM), the image feature matching algorithm ofRandom Sampling Consensus (RANSAC) is usually used to estimate the image model randomly, which is easy to cause theuncertainty of algorithm time complexity, and then lead to excessive image matching time consumption. It is difficult to meet thereal-time requirements of visual SLAM. In order to improve the problem, the algorithm of Progressive Sampling Consensus(PROSAC) is used to screen image features and reject mismatched feature points, which effectively improves the efficiency androbustness of image feature matching, and further enhances the stability and real-time performance of visual SLAM. Experimentalverification shows that compared with the current visual SLAM feature matching algorithm, the computational efficiency issignificantly improved.