A Shortest Path Algorithm of Image Segmentation Based on Fuzzy-Rough Grid

This paper studies a new fuzzy-rough sets theory, and proposes an approximation method which estimates fuzzy-roughly things in a certain range. According to this theory, the paper put forward a method parting fuzzy-roughly soft boundary grid. Based on the graph theory combined with the thoughts of the shortest path algorithm of watershed transformation, the paper puts forward a shortest path segmentation algorithm based on rough fuzzy grid, to each fuzzy-rough grid of the digital image is assigned to a shortest path. A shortest path algorithm to fully consider the relation between the positions of the grids, At the same time, this algorithm was used in the relationship between the size of mesh grid of path cost function, this would ensure the grid pixel values of similar to the same area of grid partition and the accuracy of the segmentation. It was applied in TCM tongue image segmentation experiment, the experimental results show that this method is fast, from the noise and the singularity of interference, image segmentation effect is good.