基于Retinex理论改进的低照度图像增强算法  

Research on image enhancement algorithm of low illumination based on Retinex theory

作　　者：闫保中[1] 韩旭东 何伟 YAN Baozhong;HAN Xudong;HE Wei(College of Automation,Harbin Engineering University,Harbin 150001,China)

机构地区：[1]哈尔滨工程大学自动化学院,黑龙江哈尔滨150001

出　　处：《应用科技》2020年第5期74-78,共5页Applied Science and Technology

摘　　要：针对光照不均匀、光线暗等环境导致图像采集单元采集到的图像视觉效果差、噪声大等问题,本文提出一种基于视网膜和皮层(Retinex)理论改进的低照度图像增强算法去恢复图像原有的视觉特征。将低照度图像从红、绿、蓝(RGB)空间转换到色调、饱和度、亮度(HSV)空间,在HSV空间的V通道去对低照度图像进行处理,这样能够避免图像三基色比例关系被破坏;采用改进的多尺度Retinex(MSR)算法估计光照分量,用非局部均值(NLM)滤波代替高斯滤波,利用滤波窗口与相邻窗口间的递归关系来简化计算,不仅能准确估计光照分量,还能够提高图像的处理速度;最后进行颜色空间逆变换,转换到人眼习惯的RGB颜色空间。实验结果表明该算法可以有效提高图像清晰度,保护图像的细节信息。In order to solve the problems of poor visual effect and high noise caused by uneven illumination and dark environment,an improved low illuminance image enhancement algorithm is proposed in this paper based on retinacortex(Retinex)theory,so as to restore the original visual features of the image.Firstly,the low illuminance image is transformed from red,green,blue(RGB)space to hue,saturation,value(HSV)space,and the low illuminance image is processed in the V channel of HSV space,thus the destruction of the three primary color proportion relationship of the image can be avoided.Secondly,the improved multi-scale Retinex(MSR)algorithm is used to estimate the illumination component,and non local means(NL-means)filtering is used to replace Gaussian filtering.The recursive relationship between filtering windows and adjacent windows can be used to simplify the calculation.This not only can accurately estimate the light component,but also can improve the processing speed of the image.Finally,the inverse transformation of color space is carried out,and it is converted to RGB space used by human eyes.Experimental results show that the algorithm can effectively improve image clarity and protect image details.

关 键 词：RETINEX理论 低照度图像 RGB空间 图像增强 MSR算法 HSV空间 非局部均值算法 光照分量 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]