基于改进的Retinex矿井低照度图像增强算法  

Low-light Mine Image Enhancement Algorithm Based on Improved Retinex

作　　者：董振良 王梦姣 刘晓佩[2] 田丰[2] DONG Zhenliang;WANG Mengjiao;LIU Xiaopei;TIAN Feng(China Coal Energy Research Institute Co.,Ltd.,Xi′an 710054,China;College of Communication and Information Engineering,Xi′an University of Science and Technology,Xi′an 710054,China)

机构地区：[1]中煤能源研究院有限责任公司,西安710054 [2]西安科技大学通信与信息工程学院,西安710054

出　　处：《煤炭技术》2025年第4期249-253,共5页Coal Technology

摘　　要：传统Retinex算法对低照度图像增强时存在局部光晕模糊、边缘细节保持不足和噪声严重等问题。为解决这些问题,提出了一种基于改进的Retinex矿井低照度图像增强算法。在HSV空间中,以改进的多尺度引导滤波替代Retinex中的高斯滤波,对光照分量运用韦伯-费希纳定律进行亮度均衡,将对比度受限的自适应直方图均衡算法与改进的引导滤波融合,对反射分量进行对比度增强和去噪。同时,对饱和分量进行自适应拉伸,最后转换到RGB空间得到增强图像。该算法有效提高了图像亮度和对比度,在避免光晕伪影的同时保留了边缘细节,降低了噪声影响,为低光照图像增强提供了理论参考。The traditional Retinex algorithm for low-light enhancement suffers from local halo blurring,insufficient retention of edge details,and severe noise.To solve these problems,an improved Retinex mine low-light image enhancement algorithm based on improved Retinex is proposed.The Gaussian filter in the Retinex algorithm is replaced by an improved multi-scale bootstrap filter in HSV space,the Weber-Fechner law is applied to the illumination component for illumination equalization,the contrastlimited adaptive histogram equalization algorithm is fused with the improved bootstrap filter,and contrast enhancement and denoising are performed on the reflection component.Meanwhile,adaptive stretching is performed on the saturation component,and finally converted to RGB space to obtain the enhanced image.This indicates that the algorithm effectively improves image brightness and contrast,preserves edge details while avoiding halo artifacts,reduces noise effects,and provides a theoretical reference for low-light image enhancement.

关 键 词：引导滤波 低照度 RETINEX 图像增强 HSV 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]