Fast merging algorithm for image segmentation based on date structure

Abstract

For the complexity of the algorithm for the merger of the regional boundary. A new algorithm — the Full λ Edge List Segmentation Algorithm(FELSA) is presented. Firstly, A new list — edge list in date structure of segmentation region is introduced to improve the boundaries of multi-partition approach, and the updates value of merger list in FLSA algorithm is re-defined. Thus, segmentation algorithm may select the best boundary for the merger according to the edge list. Furthermore, the adjacent part of the region is defined by edge list, so that a new edge can be retrieved more efficiently. The results show that this improved algorithm become more efficient and effective to the merger of image segmentation.