In this paper, we proposed an image segmentation algorithm based on the shortest path. The algorithm is superior to the traditional region growing algorithm (In this paper we called RGA), it devoid of certain drawbacks experienced by traditional region growing based on competing seeds. This method was enlightened by the water free movement in surface of terrain. In this method, every pixel node will be assigned to an optimal path, in order to ensure integrity and continuity of segmentation objects, we added leaking detection into the algorithm. The results of running both the proposed and the traditional algorithm on medical tongue images and other images prove the superiority of the proposed algorithm based on the shortest path. The proposed algorithm can portrays the contours of the object area accurately, especially when it is used to segment the local object of the image, the segmentation results is useful to sequence image analysis and patter recognition.
展开▼
