基于域滤波的自适应Retinex图像增强

计算机工程与科学

基于域滤波的自适应Retinex图像增强

涂清华，戴声奎

（华侨大学信息科学与工程学院厦门市移动多媒体通信重点实验室，福建厦门 361021）

收稿日期:2015-05-26 修回日期:2015-09-23 出版日期:2016-09-25 发布日期:2016-09-25

Adaptive Retinex image enhancement based on domain transform filter

TU Qing-hua,DAI Sheng-kui

（Xiamen Key Laboratory of Mobile Multimedia Communication,
School of Information Science and Engineering,Huaqiao University,Xiamen 361021,China）

Received:2015-05-26 Revised:2015-09-23 Online:2016-09-25 Published:2016-09-25

摘要/Abstract

摘要：

为了提高低照度图像的亮度和对比度，提出了一种新的基于Retinex理论的彩色图像增强方法。首先，基于Retinex理论,提出对HSV空间V分量进行域滤波估计图像光照分量，然后将V分量与光照分量相除得到反射分量的方法。之后，采用自适应Gamma校正对光照分量进行亮度提升，然后采用CLAHE对其进行对比度增强。最后，将亮度校正光照分量与反射分量相乘得到增强后的V分量，并将增强后的图像转化为RGB空间图像，达到彩色图像增强的目的。本算法可以获得更自然的增强效果，能抑制亮度较大像素点的增强，很好地突出图像中的细节信息，克服了图像增强中增强图像对比度低、颜色失真、过增强及光照突变处出现光晕现象等缺点。本算法对多种图像有效，例如高动态(HDR)图像、非均匀光照图像及低曝光图像。通过验证，本算法得到的结果相比于传统方法视觉效果更佳。

关键词: Retinex理论, 彩色图像增强, 域滤波, 自适应Gamma校正, CLAHE

Abstract:

In order to improve the brightness and contrast of low illumination images, we propose a new color image enhancement method based on Retinex theory. Firstly, we evaluate the illumination image of the V component by domain transform filtering in the HSV space based on the Retinex theory. We divide the V component by illumination to get the reflection image. Then, we enhance the illumination image by the adaptive Gamma correction and increase the contrast of the illumination image by the CLAHE. Finally, we multiply the enhanced illumination image by the reflection image and then transform the enhanced image into RGB space. The proposed method can achieve more natural enhancement effect, suppress strong light enhancement and highlight image details. Furthermore, the proposed method overcomes the drawbacks of low contrast, color distortion, over enhancement and halo artifact, which exist in the former methods. It can also deal with many kinds of images, such as high dynamic range images (HDR), non-uniform illumination images and low exposure images. Experimental results demonstrate that the proposed method can achieve better visual quality than traditional methods.

Key words: Retinex theory, color image enhancement, domain transform filtering, adaptive Gamma correction, CLAHE

涂清华，戴声奎. 基于域滤波的自适应Retinex图像增强[J]. 计算机工程与科学.

TU Qing-hua,DAI Sheng-kui. Adaptive Retinex image enhancement based on domain transform filter [J]. Computer Engineering & Science.

参考文献

［1］Land E H. McCann: Lightness and retinex theory［J］. Journal of the Optical Society of America,1971,61(1):1-11.
［2］Jobson D J,Rahman Z,Ga W.Properties and performance of a center/surround retinex［J］.IEEE Transactions on Image Processing,1997,6(3):451-462.
［3］Jobson D,Rahmann Z,Woodell G.A multiscale retinex for bridging the gap between color images & the human observations of scenes［J］. IEEE Transanctions on Image Processing,1997,14(6):965-976.
［4］Rahman Z U,Jobson D J,Woodell G A.Retinex processing for automatic image enhancement ［J］.Journal of Electronic Imaging,2004,13(1):100-110.
［5］Li Xiao-xia,Li Cheng-guo,Zou Jian-hua,et al.New low illumination color image enhancement algorithm［J］.Computer Application Research,2011,28(9): 3554-3555.doi: 10.3969/j.issn.1001-3695.2011.09.100.(in Chinese)
［6］Chen Fang-jin,Du Xiao-jun,Ma Li,et al.Low-light image enhancement based on retinex［J］.Television Technology,2013,37(15):4-6.doi: doi: 10.3969 / j.issn.1002-8692.2013.15.002.(in Chinese)
［7］Kou Xing-yuan.The research of image enhancement algorithm based on retinex theory［D］.Shenyang: Shenyang Aerospace University,2015.
［8］Wang S,Zheng J,Hu H M,et al.Naturalness preserved enhancement algorithm for non-uniform illumination images.［J］.IEEE Transactions on Image Processing,2013,22(9):3538-3548.(in Chinese)
［9］Chen S,Beghdadi A.Natural rendering of color image based on retinex［C］∥Proc of International Conference on Image Processing,2009:1813-1816.
［10］Mahmood Z,Ali T,Khattak S,et al.A color image enhancement technique using multiscale retinex［C］∥Proc of International Conference on Frontiers of Information Technology.2013:119-124.
［11］Shin Y, Jeong S, Lee S. Adaptive brightness equalization based on retinex theory［C］∥Proc of International Technical Conference on Circuits/Systems,Computers and Communications(ITC-CSCC2013),2013:610-612.
［12］Shin Y,Jeong S,Lee S.Content awareness-based color image enhancement［C］∥Proc of the 18th IEEE International Symposium on Consumer Electronics (ISCE 2014),2014: 1-2.
［13］Xiao Quan,Ding Xing-hao,Wang Shou-jue,et al.The effective elimination of the halo phenomenon and color retention of color image enhancement algorithm［J］.Journal of Computer Aided Design and Graphics,2010,22(8):1246-1252.(in Chinese)
［14］Gastal E S L,Oliveira M M.Domain transform for edge-aware image and video processing［J］.ACM Transactions on Graphics (TOG),2011,30(4): 69.
［15］Zuiderveld K.Contrast limited adaptive histogram equalization［M］∥Graphic Gems.CA:Academic Press Professional, Inc,1994:474-485.
［16］Matkovic K,Neumann L,Neumann A,et al.Global contrast factor-a new approach to image contrast［C］∥Pro of the 1st Eurographics Conference on Computational Aesthetics in Graphic,Visualization and Imaging,2005:159-167.
附中文参考文献：
［5］李小霞,李铖果,邹建华,等.一种新的低照度彩色图像增强算法［J］.计算机应用研究,2011,28(9): 3554-3555.doi:10.3969/j.issn.1001-3695.2011.09.100.
［6］程芳瑾,杜晓骏,马丽,等.基于Retinex的低照度图像增强［J］.电视技术,2013,37(15):4-6.doi: 10.3969 / j.issn.1002-8692.2013.15.002.
［7］寇兴元.基于Retinex理论的图像增强算法研究［D］.沈阳:沈阳航空航天大学,2015.
［13］肖泉,丁兴号,王守觉,等.有效消除光晕现象和颜色保持的彩色图像增强算法［J］.计算机辅助设计与图形学学报,2010,22(8):1246-1252.

[1]	安静，张贵仓，刘燕妮. 基于多尺度top-hat变换的自适应彩色图像增强[J]. 计算机工程与科学, 2017, 39(07): 1317-1321.
[2]	钱增磊，梁久祯. 一种基于凸壳的压缩域运动对象快速分割算法[J]. J4, 2015, 37(06): 1196-1202.