首页 | 官方网站   微博 | 高级检索  
     

基于遗传算法的彩色图像滤波器研究
引用本文:杨欧,郭宝平,郭轩,胡涛.基于遗传算法的彩色图像滤波器研究[J].计算机工程与应用,2007,43(24):7-11.
作者姓名:杨欧  郭宝平  郭轩  胡涛
作者单位:1. 华中科技大学,光电子科学与工程学院,武汉,430074
2. 深圳大学,光电子学研究所,广东,深圳,518060
基金项目:国家高技术研究发展计划(863计划)
摘    要:为实现彩色图像噪声的滤波,基于遗传算法技术构建了一种新型的彩色图像滤波器,在彩色图像的矢量中值滤波过程中,该滤波器利用遗传算法具有寻找全局最优解的能力,获得滤波器窗口的最优权值。从滤波效果看,有一定提高,并进一步提出了一种基于个体相似性的遗传算法:当种群的多样性较好时,采用标准的交叉策略;当种群的多样性较差时,根据个体的相似性选择个体配对,避免相同的个体配对交叉,减少高度相似个体进行配对交叉的概率,从而提高交叉操作的效率。

关 键 词:遗传算法  矢量方向滤波器  脉冲噪声  交叉操作  变异操作
文章编号:1002-8331(2007)24-0007-05
修稿时间:2007-05

Research on color image filter based on genetic algorithm
YANG Ou,GUO Bao-ping,GUO Xuan,HU Tao.Research on color image filter based on genetic algorithm[J].Computer Engineering and Applications,2007,43(24):7-11.
Authors:YANG Ou  GUO Bao-ping  GUO Xuan  HU Tao
Affiliation:1.Institute of Optoelectronics Science and Engineering,Huazhong University of Science and Technology,Wuhan 430074,China; 2.Institute of Optoelectronics,Shenzhen University,Shenzhen,Guangdong 518060,China
Abstract:In order to filter the impulse noise in the color image or picture,this article proposes a new weighted vector directional filter based on the GA(Genetic Algorithm).This filter obtains the ability to get globally optimal weight value using GA.There are much enhancement applications in color image filtering.This paper further proposes an improved genetic algorithm based on the similarity of chromosomes,standard crossover operation is used when the diversity of population is good;otherwise,chromosomes are matched based on their similarity,crossover between two same chromosomes are avoided,and the matching probability of two highly relevant chromosomes is decreased,the efficiency of crossover operation is improved.The simulation result shows that this improved algorithm is effective.
Keywords:Genetic Algorithm  vector directional filter  impulse noise  crossover  mutation
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《计算机工程与应用》浏览原始摘要信息
点击此处可从《计算机工程与应用》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号