Computer Engineering and Applications ›› 2007, Vol. 43 ›› Issue (35): 102-104.

• 学术探讨 • Previous Articles     Next Articles

2-D maximum entropy method of image segmentation based on PSO algorithm

LIU Jian-chen,SHEN Hong-yuan   

  1. School of Information,Hunan University of Science and Technology,Xiangtan,Hunan 411201,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-12-11 Published:2007-12-11
  • Contact: LIU Jian-chen

基于微粒群算法的二维最大熵图像分割算法

刘健辰,沈洪远   

  1. 湖南科技大学 信息学院,湖南 湘潭 411201
  • 通讯作者: 刘健辰

Abstract: The PSO algorithm is used to optimize the 2-D maximum entropy for the image segmentation based on thresholds.The 2-D maximum entropy criteria function for image segmentation is built.Then,the PSO algorithm is used to maximize the criterion function.Finally,the sample images with noise are segmented effectively by the proposed algorithm.Segmentation examples show that the method proposed has greater resistance capability to noise,and is faster than the common 2-D maximum entropy approach and the 2-D maximum entropy approach based on genetic algorithm.

Key words: image segmentation, PSO algorithm, entropy, 2-D histogram

摘要: 将微粒群算法运用于二维最大熵图像阈值分割法。首先构建图像分割的二维最大熵准则函数,然后采用适用于整数规划的微粒群算法最大化该准则函数,最终实现含噪声干扰下图像的有效分割。分割实验表明,该方法具有较强的抗噪声能力,且比普通和基于遗传算法的二维最大熵法运算速度更快。

关键词: 图像分割, 微粒群算法, 熵, 二维直方图