基于大气散射模型和Retinex理论的雾霾图像清晰化算法

Image Defogging based on Atmospheric Scattering Model and Retinex Theory

根据大气散射模型,提出一种基于Retinex理论的单幅雾霾图像清晰化算法。该算法首先将降质图像转换至hsv色彩空间,然后针对亮度通道调整大气散射模型表达式,使用基于双边滤波的Retinex算法提高图像对比度,并进行图像饱和度的指数调整,最后组合各通道处理结果获得清晰化图像。实验结果表明该方法能增强雾天图像对比度,达到图像去雾的目的。

Based on the atmospheric scattering model,a method for color image defoggingbased on Retinex theory was presented.Firstly,the degraded image was converted into the hsv color space,the expression of the scattering model was adjusted for the luminance channel,and the image contrast was improved using Retinex algorithm based on bilateral filtering,and then an index adjustment of the image saturation was performed.Finally,the processing results of each channel were combined to obtain a clear image.The experimental results show that the method can enhance the contrast of foggy images and achieve the purpose of image defogging.