Research on Interactive Segmentation Algorithm Based on Search Path Optimization

For many images with complicated foreground and background, it is difficult to locate interest object area by automatic segmentation algorithm. Interactive segmentation algorithms can provide effective control on these occasions. Interactive live-wire segmentation algorithm, which can provide accurate, reproducible boundary extraction, is widely applied. However, the algorithm, which employs Dijkstra algorithm to search the optimal path between two nodes, is time consuming, especially in the case of real-time segmenting sequence of slice images including plenty of pixels. The paper presents two methods of search path optimization to improve the real-time capability of the algorithm. The constraint that the search ends as soon as finds the goal node, is set up to greatly reduce searched nodes. Meanwhile, heap sort is used for searching the globally optimal path from the start node to the goal node. Algorithm analysis and experiments indicate that the presented search strategy can evidently improve the efficiency of live-wire algorithm.