J4 ›› 2011, Vol. 38 ›› Issue (6): 152-158+178.doi: 10.3969/j.issn.1001-2400.2011.06.025

• 研究论文 • 上一篇    下一篇

基于细菌觅食算法的SAR图像阈值分割

马苗1;梁建慧1,2;郭敏1
  

  1. (1. 陕西师范大学 计算机科学学院,陕西 西安  710062;
    2. 海南大学 应用科技学院,海南 儋州  571737)
  • 收稿日期:2010-10-08 出版日期:2011-12-20 发布日期:2011-11-29
  • 通讯作者: 马苗
  • 作者简介:马苗(1977-),女,副教授,博士,E-mail: mmthp@snnu.edu.cn
  • 基金资助:

    国家自然科学基金资助项目(60803088;10974130);陕西省青年科技新星资助项目(2011kjxx17);中央高校基本科研业务费专项资金重点资助项目(GK200901006);陕西师范大学研究生培养创新基金资助项目(2010CXS010)

SAR image thresholding segmentation based on the bacteria foraging algorithm

MA Miao1;LIANG Jianhui1,2;GUO Min1
  

  1. (1. College of Computer Sci., Shaanxi Normal Univ., Xi'an  710062, China;
    2. College of Applied Sci. adn Tech. of Hainan Univ., Danzhou  571737, China)
  • Received:2010-10-08 Online:2011-12-20 Published:2011-11-29
  • Contact: MA Miao

摘要:

提出了一种SAR图像阈值分割新方法.该方法在深入分析基本细菌觅食算法的基础上,缩小菌群的觅食空间以进一步提高分割阈值的搜索速度,然后采用改进的二维灰熵模型作为细菌觅食算法的适应度函数,通过菌群的趋化、复制和驱散3种行为模式并行搜索最佳阈值.实验结果初步显示,该方法在收敛速度、稳定性和分割效果3个方面,均优于基于遗传算法、人工鱼群算法等群体智能优化算法的分割方法.

关键词: 图像分割, 细菌觅食算法, 全局优化, SAR图像

Abstract:

In order to increase the speed and the accuracy of Synthetic Aperture Radar(SAR) image segmentation, a new thresholding segmentation method is proposed, which integrates the Bacterial Foraging Algorithm (BFA) and two-dimensional grey entropy. After the basic BFA is analyzed deeply, the method improves the search speed of the best threshold via shrinking the foraging space of the bacterial swarm. And then, an improved two-dimensional grey entropy model is regarded as the fitness function of our optimized BFA. Finally, the best threshold is located gradually and quickly by three behaviors of the bacterial swarm, i.e., chemotaxis, and reproduction, elimination and dispersal. Experimental results show that the proposed method is superior to some segmentation methods based on the Genetic Algorithm (GA) and Artificial Fish Swarm Algorithm (AFSA) in convergence, stability and segmentation effects.

Key words: image segmentation, bacterial foraging algorithm, global optimization, SAR image

中图分类号: 

  • TP391.41,N941.5