Edge detection plays an important role in many domains, such ascomputer vision, pattern recognition and content-based image retrievalsystems. In this paper, we propose a method for prompt contourdetection to extract the boundaries of objects from an image. Comparedwith conventional methods, the proposed method significantly reducesthe amount of calculation. Furthermore, it overcomes the multipleobjects detection problem in the active contour method. Theexperimental results confirmed the efficiency and robustness of thismethod.
Proceedings of the Seventh International Conference on Distributed Multimedia Systems (DMS 2001)，頁439-443