融合暗通道与MSRCR算法的图像去雾研究  

Research on Image Dehazing Using Dark Channel and MSRCR Algorithm

作　　者：李玉鑫 于会山[1] 梁天全 Li Yuxin;Yu Huishan;Liang Tianquan(School of Physical Science and Information Engineering,Liaocheng University,Liaocheng 252000;School of Geography and Environment,Liaocheng University,Liaocheng 252000)

机构地区：[1]聊城大学物理科学与信息工程学院,聊城252000 [2]聊城大学地理与环境学院,聊城252000

出　　处：《现代计算机》2022年第22期24-30,共7页Modern Computer

基　　金：国家自然科学基金(31800367);山东省自然科学基金(ZR2021MD090、ZR2018BD008)。

摘　　要：为了解决传统暗通道先验算法易产生图像偏暗、大气光值选取不准确等问题,提出了一种融合暗通道先验和带色彩恢复的多尺度视网膜增强算法(Multi-Scale Retinex with Color Restoration,MSRCR)的图像去雾方法。首先建立大气散射模型,引入四叉树分解方法估计大气光值,并利用导向滤波算法细化透射率,利用优化的暗通道先验算法对图像进行清晰化处理得到复原图像;其次利用MSRCR算法对雾天图像进行色彩增强处理得到增强图像;然后根据复原图像和增强图像的特点对两者进行融合,最终得到去雾图像。实验表明,该算法在主观感知和客观指标上均优于现阶段的去雾算法。To address the problems that the traditional dark channel prior algorithm is prone to image darkness and inaccu⁃rate selection of atmospheric light values,an image defogging method that combines dark channel prior and Multi-Scale Retinex with Color Restoration(MSRCR)is proposed.Firstly,an atmospheric scattering model is established,a quadtree decomposition method is introduced to estimate the atmospheric light value,and a guided filtering algorithm is used to refine the transmittance,and an optimized dark channel prior algorithm is used to clarify the image to obtain the recovered image;secondly,the MSRCR al⁃gorithm is used to color enhance the foggy sky image to obtain the enhanced image;then the two are fused according to the characteristics of the recovered image and the enhanced image,and finally the defogged image.Experiments show that the algorithm is superior to the current dehazing algorithm in both subjective perception and objective indicators.

关 键 词：图像去雾 引导滤波 图像融合 暗通道先验 四叉树分解 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]