结合灰度世界和Retinex的低照度图像增强算法  

Low Illumination Image Enhancement Algorithm Based on Gray World and Retinex

作　　者：安芷楠 魏赟[1] AN Zhinan;WEI Yun(School of Optical-Electrical and Computer Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China)

机构地区：[1]上海理工大学光电信息与计算机工程学院,上海200093

出　　处：《小型微型计算机系统》2024年第2期477-482,共6页Journal of Chinese Computer Systems

基　　金：国家重点研发计划项目(2018YFB1700902)资助。

摘　　要：针对图像增强处理中亮度偏低、细节模糊、色彩失真等问题,提出了一种基于Gabor滤波的多尺度Retinex与自适应补偿的灰度世界理论结合的图像增强算法.通过双伽马函数预处理图像暗区对比度并抑制亮区过增强;考虑Gabor滤波良好的方向选择性和纹理细节敏感度,将其作为多尺度Retinex算法的中心环绕函数在双曲正切域上进行卷积来突出细节,并添加增益偏移算子动态地调整反射分量的亮度;最后使用自适应补偿的灰度世界算法进行色彩保持和校正,得到最终的输出图像.实验结果表明该算法在结构相似性、标准差和平均梯度上比对比算法的平均值分别高出47%、38%和50.3%,处理后的图像在提高亮度、突出细节的同时保持了较高的色彩度.Aiming at the problems of low brightness,fuzzy details and color distortion in image enhancement processing,an image enhancement algorithm based on Gabor filter multi-scale Retinex and adaptive compensation gray world theory is proposed.The double gamma function is used to preprocess the contrast of dark areas and suppress the over enhancement of bright areas;Considering the good directional selectivity and texture detail sensitivity of Gabor filter,it is convoluted on the hyperbolic tangent domain as the central surround function of multi-scale Retinex algorithm to highlight the details,and a gain offset operator is added to dynamically adjust the brightness of the reflection component;Finally,the adaptive compensation gray world algorithm is used to maintain and correct the color,and the final output image is obtained.The experimental results show that the algorithm is 47%,38% and 50.3% higher than the average value of the comparison algorithm in structural similarity,standard deviation and average gradient respectively.The processed image maintains a high color degree while improving brightness and highlighting details.

关 键 词：图像增强 Lab颜色空间 RETINEX算法 GABOR滤波 颜色恢复 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]