计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (29): 188-190.DOI: 10.3778/j.issn.1002-8331.2008.29.053

• 图形、图像、模式识别 • 上一篇    下一篇

基于分散微粒群算法的二维模糊最大熵图像分割

田 杰1,曾建潮1,侯明冬2   

  1. 1.太原科技大学 系统仿真与计算机应用研究所,太原 030024
    2.山东劳动职业技术学院 电气自动化系,济南 250000
  • 收稿日期:2007-11-20 修回日期:2008-01-31 出版日期:2008-10-11 发布日期:2008-10-11
  • 通讯作者: 田 杰

2-D fuzzy maximum entropy method in image segmentation based on DPSO

TIAN Jie1,ZENG Jian-chao1,HOU Ming-dong2   

  • Received:2007-11-20 Revised:2008-01-31 Online:2008-10-11 Published:2008-10-11
  • Contact: TIAN Jie

摘要: 该文研究了基于二维模糊信息熵的图像分割方法,针对二维模糊信息熵图像分割方法求取阈值时存在的计算复杂、时间长、实用性差等问题,提出了基于优化微粒群算法的二维最大熵图像分割方法。DPSO算法对图像的二维阈值空间进行全局搜索,并将搜索得到的二维熵最大值所对应的点灰度-区域灰度均值作为阈值进行图像分割。同时,为了避免该算法收敛到局部最优解的问题,在算法中引入了变异策略。通过实验显示了该算法在收敛性和计算效率上较QPSO在内其它优化算法具有更好的优越性。

关键词: 图像分割, 二维模糊最大熵, 分散粒子群优化算法, 粒子群优化算法, 算子行为的微粒群优化算法

Abstract: The 2-D fuzzy maximum entropy image segmentation method is studied in this paper,for the problems that the method is complex,time-consuming and lack of practicability during evaluating threshold,a 2-D fuzzy maximum entropy image segmentation method based on DPSO is presented.The proposed method searches the 2-D space of threshold using DPSO,and takes the gray scale value of pixel and the gray scale mean value of region corresponding to the 2-D maximum entropy value in the search space as the threshold for image segmentation.Furthermore,a mutation strategy is designed to avoid premature convergence.Simulation results show the good efficiency of DPSO to image segmentation.

Key words: image segmentation, 2-D fuzzy maximum entropy, Discrete Particle Swarm Optimization(DPSO), Particle Swarm Optimization(PSO), Quantum-behaved Particle Swarm Optimization(QPSO)