作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程

• 人工智能及识别技术 • 上一篇    下一篇

基于质心的Copula EDA及其在图像去噪中的应用

严莉娜,王丽芳   

  1. (太原科技大学复杂系统与计算智能实验室,太原 030024)
  • 收稿日期:2014-11-21 出版日期:2016-02-15 发布日期:2016-01-29
  • 作者简介:严莉娜(1988-),女,硕士研究生,主研方向为智能信息处理;王丽芳,副教授。
  • 基金资助:
    国家自然科学青年基金资助项目(61003053);山西省优秀研究生创新基金资助项目(20113121);太原科技大学研究生科技创新基金资助项目(20134025)。

Copula Estimation of Distribution Algorithm Based on Centroid and Its Application in Image Denoising

YAN Lina,WANG Lifang   

  1. (Laboratory of Complex System and Computational Intelligence,Taiyuan University of Science and Technology,Taiyuan 030024,China)
  • Received:2014-11-21 Online:2016-02-15 Published:2016-01-29

摘要: 针对传统Copula分布估计算法(EDA)局部搜索能力较差的缺点,提出基于质心的Copula EDA。在传统算法的基础上加入质心变异算子,使种群个体带着较优个体的经验信息向最优解方向进行搜索,并将其应用到图像去噪中,利用优化BP神经网络的初始权值和阈值区分出污染像素做去噪处理。实验结果表明,该算法可使BP网络分类更精确,通过分类后去噪的图像具有较高的峰值信噪比。

关键词: Copula函数, 分布估计算法, 质心, BP神经网络, 图像去噪

Abstract: For the disadvantage of poor local search ability of traditional Copula Estimation of Distribution Algorithm(EDA),a Copula EDA based on Centroid(CEDAC) is proposed.On the basis of traditional algorithm,it joins the centroid mutation operator to make the individual species with better information searching to the direction of the optimal solution.Through the optimination of initial weights and threshold of BP neural network,it effectively distinguishes the noise pollution pixels.Experimental results show that the optimized BP neural network classifies pixels more accurate and images present a higher Peak Signal to Noise Ratio(PSNR) after denoising through classification.

Key words: Copula function, Estimation of Distribution Algorithm(EDA), centroid, BP neural network, image denoising

中图分类号: