摘要
针对水下环境存在的颜色衰减和散射效应导致水下图像颜色严重失真的问题,提出一种多通道均衡化的水下图像增强算法.首先,对原始图像在对数域上进行归一化处理后转换到HSI(色调-饱和度-亮度)颜色空间;然后,对亮度分量利用McCann Retinex算法在四个方向(纵横)进行比较、实现增强,并根据图像全局亮度信息进行照度增强;最后,将图像重新转换到RGB(红-绿-蓝)空间,计算各通道的累积分布函数,对密集部分进行拉伸处理,达到颜色均衡的效果.针对多幅水下彩色图像进行增强对比实验,结果表明:通过该方法得到的增强图像颜色失真程度减弱,图像对比度和清晰度显著提高,色彩更加鲜艳;该算法在改善水下图像照度信息的同时,保留了饱和度和色度信息,解决了水下图像增强的颜色失真问题,使水下图像具有较高的对比度和清晰度.
In view of the color attenuation and scattering effect of underwater environment, the color of underwater image is seriously distorted.In this paper,a multi-channel equalization algorithm for underwater image enhancement was proposed.First,the original image was converted to the HIS(hue-saturation-intensity)color space after the normalization of the original image in the logarithmic domain.Then,the intensity component was enhanced by using the McCann Retinex algorithm in four directions(the vertical and horizontal directions), and the illumination was enhanced according to the global brightness information of the image.Finally,it was re-converted to RGB(red-green-blue) space,and the cumulative distribution function of each channel was calculated, and the dense part was stretched to achieve the effect of color equalization. The contrast experiments of multiple underwater color images were carried out.The results show that the color distortion of the enhanced image obtained by the proposed method is effectively relieved, the contrast and clarity of the image are significantly improved and the color is brighter. The algorithm,while improving the information of underwater image,preserves the information of saturation and chromaticity,and solves the problem of color distortion of underwater image enhancement.In addition,it has high contrast and clarity.
引文
[1]JONASZ M,FOURNIER G R.Light scattering by particles in water[M].New York:Academic,2007.
[2]蔡珣,孟祥旭,向辉.光照色调颜色恒常性算法研究[J].中国图象图形学报,2004,9(8):922-926.
[3]CHAMBAH M,RIZZI A.Underwater color constancy:enhancement of automatic live fish recognition[J].Proc of SPIE:The International Society for Optical Engineering,2004,5293:157-168.
[4]WEN H,TIAN Y,HUANG T,et al.Single underwater image enhancement with a new optical model[C]//Proc of IEEE International Symposium on Circuits and Systems.Beijing:IEEE,2013:753-756.
[5]倪锦艳,李庆武,周亚琴,等.基于透射率优化和色温调节的水下图像复原[J].激光与光电子学进展,2017,54(1):90-97.
[6]杨淼,纪志成.基于模糊形态筛和四元数的水下彩色图像增强[J].仪器仪表学报,2012,33(7):1601-1605.
[7]孙斌,卜雄洙,王正成,等.基于多尺度Retinex的缺陷图像增强算法[J].无损检测,2017,39(6):24-27.
[8]翟广涛.基于人眼视觉系统特性的数字图像处理技术的研究[D].上海:上海交通大学图书馆,2009.
[9]FRANKLE J A,MCCANN J J.Method and apparatus for lightness imaging:US,US4384336[P].1983-05-17.
[10]LAND E H,MCCANN J J.Lightness and Retinex theory[J].Journal of the Optical Society of America,1971,61(1):1-11.
[11]杨爱萍,郑佳,王建,等.基于颜色失真去除与暗通道先验的水下图像复原[J].电子与信息学报,2015,37(11):2541-2547.
[12]GONZALEZ R C,WOODS R E,EDDINS S L.Digital image processing using MATLAB[M].Knoxville:Gatesmark Publishing,2009.
[13]闫钧宣,张科.HSI空间亮度信息的多尺度Retinex图像增强研究[J].计算机工程与应用,2010,46(23):31-33.
[14]WANG W,LI B,ZHENG J,et al.A fast multi-scale Retinex algorithm for color image enhancement[C]//Proc of International Conference on Wavelet Analysis and Pattern Recognition.Baoding:IEEE,2008:80-85.
[15]何春,宋国琴.基于MSRCR理论的算法实现[J].智能计算机与应用,2017,7(3):114-116.
[16]应凌楷,李子印,张聪聪.融合梯度信息与HVS滤波器的无参考清晰度评价[J].中国图象图形学报,2015,20(11):1446-1452.
[17]李福文,金伟其,陈伟力,等.基于Retinex模型的彩色图像全局增强算法[J].北京理工大学学报,2010,30(8):947-951.
[18]MITTAL A,MOORTHY A K,BOVIK A C.No-reference image quality assessment in the spatial domain[J].IEEETransactions on Image Processing,2012,21(12):4695-4708.
[19]MITTAL A,SOUNDARAJAN,BOVIK A C.Making a“completely blind”image quality analyzer[J].IEEESignal Processing Letters,2012,20(3):209-212.
[20]ZHANG L,ZHANG L,BOVIK A C.A feature-enriched completely blind image quality evaluator[J].IEEE Transactions on Image Processing,2015,24(8):2579-2591.