一种基于MALLAT算法的图像去雾方法  

An Image Defogging Method Based on MALLAT Algorithm

作　　者：贺欢 吐尔洪江·阿布都克力木[1] 何笑 HE Huan;TUERHONGJIANG·Abudukelimu;HE Xiao(School of Mathematical Sciences,Xinjiang Normal University,Urumqi,Xinjiang,830017,China)

机构地区：[1]新疆师范大学数学科学学院,新疆乌鲁木齐830017

出　　处：《新疆师范大学学报（自然科学版）》2020年第1期23-27,共5页Journal of Xinjiang Normal University(Natural Sciences Edition)

基　　金：国家自然科学基金资助项目(11261061,61362039,10661010);新疆维吾尔自治区自然科学基金资助项目(200721104);新疆师范大学数学教学资源开发重点实验室招标课题(XJNUSY082017B03)等资助。

摘　　要：文章针对户外雾天场景下采集到的图像中包含大面积浓雾,影响图像细节信息的提取等问题,提出了一种基于小波变换的图像去雾方法。首先将含雾图像转换为灰度图像,再使用二维离散小波变换对灰度图像进行一次分解,对分解后得到的低频子图使用灰度线性变换处理,对三个高频子图分别使用灰度非线性变换处理,最后将处理后的四个子图进行二维离散小波逆变换重构,得到目标图像。仿真实验表明,文章所提方法与其他增强函数处理相比,去雾效果较好。In this paper,a method of image defogging based on wavelet transform is proposed to solve the problem that the image collected in the scene of outdoor fog contains a large area of fog and affects the extraction of imagedetails.First,the fog image is transformed into gray image,then the gray image is decomposed by two-dimensional discrete wavelet transform,the low-frequency subgraph is processed by gray linear transformation,the three high-frequency subgraphs are processed by gray nonlinear transformation,and the four subgraphs are reconstructed by two-dimensional discrete wavelet inverse transformation to obtain the target image.Simulation results show that the proposed method and other enhancement function processing compared,the defogging effect is better.

关 键 词：小波变换 MALLAT算法 灰度线性变换 非线性变换 

分 类 号：TN957.52[电子电信—信号与信息处理]