计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (21): 109-111.DOI: 10.3778/j.issn.1002-8331.2009.21.032

• 图形、图像、模式识别 • 上一篇    下一篇

一种最大模糊熵多尺度彩色图像增强方法

杜海顺1,汪凤泉2   

  1. 1.河南大学 先进控制与智能信息处理研究所,河南 开封 475004
    2.东南大学 仪器科学与工程学院,南京 210096
  • 收稿日期:2008-04-25 修回日期:2008-07-21 出版日期:2009-07-21 发布日期:2009-07-21
  • 通讯作者: 杜海顺

Multi-scale color image enhancement algorithm based on human visual system and maximal fuzzy entropy

DU Hai-shun1,WANG Feng-quan2   

  1. 1.Institute of Advanced Control and Intelligent Information Processing,Henan University,Kaifeng,Henan 475004,China
    2.College of Instrument Science & Engineering,Southeast University,Nanjing 210096,China
  • Received:2008-04-25 Revised:2008-07-21 Online:2009-07-21 Published:2009-07-21
  • Contact: DU Hai-shun

摘要: 在HSI颜色空间提出了一种基于视觉特性和最大模糊熵的多尺度彩色图像增强算法。对亮度分量I进行小波多尺度分解,对于不同分解尺度的高频小波子带,采用不同的非线性函数进行处理,以实现图像细节的增强,同时最大限度地抑制噪声。对包含亮度信息的低频小波子带,首先进行基于最大模糊熵的模糊化处理,其中模糊化隶属度函数采用遗传算法进行优化;然后,对经过模糊化的低频小波子带,采用基于人眼视觉特性的线性矫正函数(LRF)进行处理,以实现对图像亮度更好的动态调节。对饱和度分量S,采用指数拉伸方法进行增强处理,色调H保持不变。实验证明,增强后的彩色图像更加清晰,色彩更加鲜艳而不失真。

关键词: 彩色图像增强, 小波变换, 最大模糊熵, 遗传算法

Abstract: An algorithm based on human visual system and maximal fuzzy entropy is presented in the HSI color space.The luminance component I is multi-scale wavelet decomposed first.To enhance image detail and suppress noise,different no-linear function is used to process different high frequency wavelet sub-bands.First,based on maximal fuzzy entropy,the low frequency wavelet sub-band is transformed into fuzzy domain by fuzzy function that is optimized by genetic algorithm.Then,the fuzzy low frequency wavelet sub-band is processed by Linear Rectification Function(LRF) to adjust image brightness better.The saturation component S is processed by exponential method,and the hue component H is invariant.Experiment results confirm that improved color image enhancement is achieved.

Key words: color image enhancement, wavelet transform, maximal fuzzy entropy, genetic algorithm