华南理工大学学报(自然科学版) ›› 2015, Vol. 43 ›› Issue (1): 59-65.doi: 10.3969/j.issn.1000-565X.2015.01.010

• 电子、通信与自动控制 • 上一篇    下一篇

基于人工蜂群优化的 NSCT 域图像模糊集增强方法

吴一全1,2,3,4 殷骏1 戴一冕1   

  1. 1. 南京航空航天大学 电子信息工程学院, 江苏 南京 210016 ; 2. 农业部渔业装备与工程技术重点实验室, 上海 200092 ;3. 农业部淡水渔业和种质资源利用重点实验室, 江苏 无锡 214081 ;4. 南京财经大学 食品科学与工程学院 ∥ 江苏省粮油品质控制及深加工技术重点实验室, 江苏 南京 210023
  • 收稿日期:2014-03-25 修回日期:2014-09-03 出版日期:2015-01-25 发布日期:2014-12-01
  • 通信作者: 吴一全(1963-),男,博士,教授,博士生导师,主要从事图像处理与分析、目标检测与识别、智能信息处理研究. E-mail:nuaaimage@163.com
  • 作者简介:吴一全(1963-),男,博士,教授,博士生导师,主要从事图像处理与分析、目标检测与识别、智能信息处理研究.
  • 基金资助:

    国家自然科学基金资助项目( 60872065 );农业部渔业装备与工程技术重点实验室开放基金资助项目( 2013001 );农业部淡水渔业与种质资源利用重点实验室开放基金资助项目( KF201313 );江苏省粮油品质控制及深加工技术重点实验室开放基金资助项目( LYPK201304 );深圳市城市轨道交通重点实验室开放基金资助项目( SZCSGD201306 );江苏省高校优势学科建设工程资助项目; 2013 年研究生学位论文创新与创优基金资助项目( DZS201203 )

Image Enhancement in NSCT Domain Based on Fuzzy Sets and Artificial Bee Colony Optimization

Wu Yi - quan1,2,3,4 Yin Jun1 Dai Yi - mian1   

  1. 1. College of Electronic and Information Engineering , Nanjing University of Aeronautics and Astronautics , Nanjing 210016 ,Jiangsu , China ; 2. Key Laboratory of Fishery Equipment and Engineering , Ministry of Agriculture of the People ’ s Republic of China ,Shanghai 200092 , China ; 3. Key Laboratory of Freshwater Fisheries and Germplasm Resources Utilization , Ministry of Agriculture of the People ’ s Republic of China , Wuxi 214081 , Jiangsu , China ; 4. Jiangsu Key Laboratory of Quality Control and Further Processing of Cereals and Oils , Nanjing University of Finance and Economics , Nanjing 210023 , Jiangsu , China
  • Received:2014-03-25 Revised:2014-09-03 Online:2015-01-25 Published:2014-12-01
  • Contact: 吴一全(1963-),男,博士,教授,博士生导师,主要从事图像处理与分析、目标检测与识别、智能信息处理研究. E-mail:nuaaimage@163.com
  • About author:吴一全(1963-),男,博士,教授,博士生导师,主要从事图像处理与分析、目标检测与识别、智能信息处理研究.
  • Supported by:
    Supported by the National Natural Science Foundation of China ( 60872065 )

摘要: 针对实际应用中所采集的图像对比度低、边缘细节模糊的问题,提出了基于非下采样 Contourlet 变换 ( NSCT ) 、模糊集、人工蜂群 ( ABC ) 优化的自适应图像增强方法 . 首先对输入图像进行 NSCT 分解,得到一个低频子带和多个高频子带;然后依据贝叶斯萎缩阈值和非线性增益函数增强高频子带系数,采用模糊增强法增强低频子带系数,并利用 ABC 算法优化其中的模糊参数,以提高模糊增强法的自适应性;接着用低频子带图像的信息熵作为 ABC 算 法的适应度函数,同时引入较劣种群随机初始化策略改进 ABC 算法,以缩短增强方法的运行时间 . 文中采用该增强方法对淡水鱼、铁轨表面、储粮害虫 3 类图像进行了
增强实验,并依据主观视觉效果和对比度增益、清晰度增益、信息熵 3 个客观定量评价指标,对文中方法及其他 3 种同类增强方法进行了比较 . 结果表明,所提出的方法视觉效果最佳,能提高图像的对比度和清晰度,目标边缘光滑,且增加了图像的信息量,便于后续准确地进行图像检测与识别.

关键词: 图像增强, 非下采样 Contourlet 变换, 模糊集, 人工蜂群算法, 贝叶斯萎缩阈值, 非线性增益, 自适应增强

Abstract: Proposed in this paper is an adaptive image enhancement method on the basis of nonsubsampled Contourlet transform (NSCT) , fuzzy sets and artificial bee colony (ABC) optimization , which helps improve the low contrast and definition of the image acquired in practical applications. In this method , first , an input image is decomposed into a low-frequency sub-band and several high-frequency sub-bands through NSCT. Secondly ,the coefficients of high-frequency sub-bands are enhanced according to Bayesian shrinkage threshold and nonlinear gain function , while that of the low-frequency sub-band is enhanced by using the fuzzy enhancement method with its adaptability improved by fuzzy parameter optimization via ABC algorithm. Then , for the purpose of reducing running time , the entropy of low-frequency sub-band image is used as the fitness function of ABC algorithm and a random initializing strategy of inferior populations is introduced to improve ABC algorithm. The proposed enhancement method is finally employed to process three kinds of images of freshwater fish , rail surface and grain pest , and a comparison is made between the proposed method and three other similar enhancement methods in terms of subjective visual effect and such objective quantitative evaluation indices as contrast gain , definition gain and entropy. Experimental results show that the proposed method is of the most
excellent visual effect because it helps obtain images with improved contrast and definition , smooth edge and greater information amount , which benefits further accurate image detection and recognition.

Key words: image enhancement, nonsubsampled Contourlet transform, fuzzy sets, artificial bee colony algo-rithm, Bayesian shrinkage threshold, nonlinear gain, adaptive enhancement

中图分类号: