基于大气光值图的自适应去雾算法

摘要

To restore interesting information from the blurring image taken in the bad weather such as fog and haze,an adaptive algorithm of image haze removal based on airlight map was proposed.The raw atmospheric veil was got using a low-pass filter which based on Laplace.Mean square error was used to refine the raw atmospheric veil adaptively and remove haze from input image using atmospheric scattering model.Experimental results demonstrate that the proposed method achieves better quality restoration and higher processing speed than classic dehazing algorithm.It is more successful in recovering different intensities of haze.Hence,it has high quality in human vision.%针对雾霾天气状况下获取的图像存在对比度低和细节信息缺失的问题,提出一种基于大气光值图的自适应去雾算法.基于大气光幕的原理,运用拉普拉斯算子构建低通滤波器,估算粗略的大气光幕图,使用局部均方误差运算实现对粗大气光幕图的自适应修正,依据大气散射模型进行去雾.仿真结果表明,该算法可有效复原退化场景的对比度和清晰度,在处理速度上比传统去雾算法有一定提升,对不同浓度的雾区域实现了自适应处理,去雾结果整体视觉效果更佳.