大气光幕融合的去雾新方法  

Novel Haze Removal Method Based on Atmospheric Veil Fusion

作　　者：刘子兵 戴声奎[1] LIU Zibing;DAI Shengkui(College of Information Science and Engineering,Huaqiao University,Xiamen 361021,China)

机构地区：[1]华侨大学信息科学与工程学院,福建厦门361021

出　　处：《华侨大学学报（自然科学版）》2018年第2期246-251,共6页Journal of Huaqiao University(Natural Science)

基　　金：福建省科技计划重点项目(2013H0030);中央高校基本科研业务基金资助项目(JB-ZR1145)

摘　　要：针对雾天场景成像设备采集的图像存在对比度低、细节不清晰的问题,提出一种结合大气光幕融合的雾天场景复原算法.首先,基于不同场景深处大气光幕的物理特性和光学成像特性,获取远景雾气分布的近似估计.其次,通过像素级融合和滤波的方法得出准确的全局大气光幕.最后,通过反演大气散射模型得到复原图像,并进行亮度和色调调整.该方法可以有效避免过度去雾现象和光晕效应等不足,能快速复原场景的对比度和颜色.实验结果表明:该算法简单高效,具有较强场景适应能力,并保证实时性.In order to enhance contrast and detail information of haze image captured by the imaging system,a haze scene restoration algorithm based on atmospheric veil fusion is proposed.Firstly,the haze density distribution of distant area is roughly estimated based on the physical properties of the atmospheric veil in different depth and the optical reflectance imaging.Then,the global atmospheric veil is accurately estimated by using image fusion and filtering on the pixel level.Finally,the restored image is obtained through the inversion of the atmospheric scattering model,and the brightness and chroma of the images are adjusted via tone mapping.This method can avoid over defogging or halo artifacts and achieving a satisfactory restoration for better contrast and color fidelity.The experimental results yield that the proposed method is not only very simple and efficient,but also has robust scene adaptability and achieves real time performance.

关 键 词：图像复原 图像去雾 大气散射模型 大气光幕 图像融合 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]