基于颜色空间的单幅图像去雾算法

Algorithm for single image defogging based on color space

为改善雾天图像对比度差、能见度低的特点,本文结合雾天成像模型和暗原色先验规律,在颜色空间的基础上提出了一种去雾新算法。首先,在RGB颜色空间,根据暗原色先验规律估计出空气光,然后将图像从RGB颜色空间转换到HSI和HSV颜色空间,再对HSV空间下的明度分量运用大气散射模型进行去雾处理,最后再对HSI空间下的饱和度分量进行校正,最终得到去雾之后的图像。通过该算法能得到清晰化的图像,并且该算法较之传统的单幅图像去雾方法,速度更快、效果更自然。

In order to improve the low visibility and poor contrast of fog images,a new defogging algorithm based on color space is proposed in combination with the atmospheric scattering model and the dark channel prior.Firstly,the atmospheric veil is estimated by using dark channel prior in RGB space,the images is converted from RGB color space to HSI and HSV color spaces,and then the defogging processing for lightness in HSV colour space is conducted by using the atmospheric scattering model of foggy weather.The defogged image is obtained by correction of the saturation component in HSI space.In this way,a natural image can be obtained.In comparison with the traditional single image defogging methods,the algorithm is faster and has more natural effectiveness.