基于限制对比度颜色校正的水下图像增强  

Underwater Image Enhancement Based on Contrast Limited Color Correction

作　　者：余义德[1] 周曼丽[2] 王红萍[1] YU Yi-de ZHOU Man-li WANG Hong-ping(Unit 91550, PLA, Dalian Liaoning 116023, China College of Information Science ＆ Engineering, Ocean University of China, Qingdao Shandong 266100, China)

机构地区：[1]中国人民解放军第91550部队 [2]中国海洋大学信息科学与工程学院

出　　处：《无线电工程》2017年第9期16-20,共5页Radio Engineering

摘　　要：针对水下图像对比度低、颜色失真的问题,提出了限制对比度自适应的颜色校正模型算法,在灰度的基础上,先求出RGB各个通道的均值然后进行对比判断,根据均值的大小来决定是一端线性拉伸还是两端线性拉伸。该方法基于颜色平衡、RGB色彩模型和HSI颜色模型的对比度校正。基于增强水下图像质量的需求,该算法在CLAHE分块进行局部处理方式的基础上,对每小块区域进行限制对比度,然后运用颜色校正模型算法,并采用双线性插值来提高算法的效率,实验结果表明该算法更加优异。Because of the special absorption and scattering characteristics of water medium, underwater image can not show natural color image quality.This paper proposes a color correction model algorithm that limits the contrast adaptation.The color correction model algorithm is to obtain the mean values of the RGB channels and compare them based on the gray scale,and then determine to do the linear stretching of one end or two ends according to the size of the average. The method is a contrast correction based on color balance, RGB color model and HSI color model.The algorithm is based on CLAHE that divides the pictures into smM1 blocks to process, and then by limiting the contrast of each small block, applies the color correction model algorithm to process pictures based on underwater image enhancement requirement. Also, the bilinear interpolation is used to improve the efficiency of the algorithm. The practical operation shows that the algorithm has good treatment efficiency.

关 键 词：水下图像 图像增强 颜色校正模型 自适应算法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]