基于全局双约束的矿井尘雾图像增强方法  

Mine dust image enhancement method based on global double constraints

作　　者：冀常鹏[1] 贺丽娜 代巍 JI Changpeng;HE Lina;DAI Wei(School of Electronics and Information Engineering,Liaoning Technical University,Huludao 125105,China)

机构地区：[1]辽宁工程技术大学电子与信息工程学院,辽宁葫芦岛125105

出　　处：《辽宁工程技术大学学报（自然科学版）》2024年第2期225-231,共7页Journal of Liaoning Technical University (Natural Science)

基　　金：辽宁省教育厅基本科研项目(LJKMZ20220677)。

摘　　要：为提高煤矿尘雾图像的可观测性,提出一种基于全局双约束的Retinex算法的尘雾图像增强算法(GCFCDL-Retinex)。首先,将输入图像进行内外循环,训练聚类和稀疏双重约束下的过完备字典,对图像中的噪声分量进行抑制;然后,通过Retinex算法对照度分量和反射分量进行估计及提取,并对提取的照度分量进行自适应Gamma校正;最后输出增强后的图像。研究结果表明:在煤矿井下的复杂环境中,所提出的图像增强算法能够有效提高矿井下尘雾图像的对比度和清晰度,去除真实粉尘,同时抑制图像光晕、边缘模糊的现象,增强后的图像色彩自然,视觉效果明显提升。研究结论为矿井下视频监控清晰化的工程应用提供理论依据。In order to improve the observability of dust haze images in coal mines,a dust haze image enhancement algorithm based on global double constraint Retinex algorithm was proposed.Firstly,the input image is cycled inside and outside,and the overcomplete dictionary under the double constraints of clustering and sparse is trained to suppress the noise component in the image.Then,the Retinex algorithm was used to estimate and extract illuminance and reflection components,and adaptive Gamma correction was performed for the extracted illuminance components.The final output of the enhanced image.The research results show that the proposed image enhancement algorithm can effectively improve the contrast and clarity of dust fog image in the complex underground environment of coal mine,remove the real dust,and suppress the phenomenon of image halo and edge blur.The enhanced image color is natural,and the visual effect is significantly improved.The research conclusion provides a theoretical basis for the engineering application of mine video monitoring sharpening.

关 键 词：图像增强 图像去噪 稀疏约束 聚类约束 RETINEX算法 

分 类 号：TD67[矿业工程—矿山机电]