• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2006, Vol. 28 ›› Issue (1): 86-88.

• 论文 • 上一篇    下一篇

基于A^*搜索方法的主动轮廓提取算法

陈欣[1,2] 熊岳山[1] 谭珂[3] 郭光友[3]   

  • 出版日期:2006-01-01 发布日期:2010-05-20

  • Online:2006-01-01 Published:2010-05-20

摘要:

在主动轮廓模型Snake的研究与应用中,如何降低Snake对初始轮廓的敏感性以厦如何保证能量极小化过程收敛到全局最小值,是两个极其重要的问题。为提取两点间的目标轮廓线,本文提出了基于A^*搜索过程的Snake算法,将Snake能量极小化问题转换成势能地图的最短路径搜索问题。实验结果表明,该算法简化了Snake的初始化过程,而且Snaake能量收敛至全局最小值,具有良好的实用性与鲁棒性。

关键词: 主动轮廓模型 几何Snake模型 势能地图 图搜索 最短路径 A^*算法

Abstract:

How to alleviate the sensitivity of active contour models to their initial curves and bow to keep convergent to the global minimum during the iterative procedure of energy minimization are difficult problems for solving the active eontour models. In order to extract open contour between two given poin ts, a Snake approach based on the A^* graph search procedure is proposed in this paper. The classical energy minimization is turned into searching the minimal weighted paths in a potential map. Our approach for active contour extraction simplifies the initialization of the Snake model, just needing two boundary points for an open contour, and the experimental results show that this method is sure to find the global minimum of the Snake energy. It is   able to give satisfying solutions for contour extraction.

Key words: active contour model, geometric Snake model, potential map, graph search, minimal path , A^* method