A Robust Image Mosaic Method Based on Graph-Model Optimization

In this paper, we introduce an automatic image mosaic method based on graph-model, which is built with matching mean square error, and graph theory method is used to minimize the global error. We apply optimized Dijkstra algorithm to speed up the path-finding algorithm. As a result, the quality of the output panorama is quite good and the proposed algorithm gets the lowest error compared with other methods.