Abstract: In view of complexity of underground environment of coal mine and limitations of image stitching algorithms, an automatic and quick stitching algorithm of mine monitoring image was proposed. The algorithm combines advantages of Harris algorithm and SIFT algorithm, and uses improved RANSAC algorithm for purification and matching of extracted feature points and model parameter estimation, so that the anti-scaling performance and noise immunity are greatly improved. It uses locality-sensitive hashing algorithm to improve success rate and real-time performance of image stitching. The experimental results show that the algorithm is robust and has fast stitching capability, and can be applied to image automatic stitching of mine monitoring image.