基于亮度对比度增强与饱和度补偿的快速图像去雾算法

Fast Image Defogging Algorithm Based on Luminance Contrast Enhancement and Saturation Compensation

针对传统去雾方法存在的恢复质量差和耗时较长等问题,提出一种基于亮度对比度增强与饱和度补偿的快速图像去雾算法.首先提出一种基于颜色畸变因子的偏色检测方法,并通过基于完美反射的白平衡方法对存在偏色的图像进行色彩校正;然后在Lab色彩空间的亮度通道内,结合基于天空检测的大气光值估计方法和基于暗通道先验的透射率估计方法,提出基于大气散射模型的亮度通道对比度增强方法;最后为了改善图像画面的苍白感,在HSV色彩空间的饱和度通道内提出基于大气散射模型的饱和度补偿方法,对初步去雾结果进行饱和度补偿,获得色彩鲜艳的无雾图像.与有代表性的去雾方法进行实验的结果表明,该算法可以有效地提高去雾速度,并在避免颜色畸变的同时获得高对比度和高饱和度的清晰复原结果.

Aiming at the disadvantages of low image quality and long response time in the existed image defogging methods,we propose a fast image defogging algorithm based on luminance contrast enhancement and saturation compensation.Firstly,a color cast detection method based on color distortion factor is put forward,and correction of the images with color cast is realized through a white balance method on the basis of perfect reflection.Then,in the luminance channel of Lab color space,a luminance contrast enhancement method based on atmospheric scattering model is raised by the combination of the sky detection based atmospheric light estimation as well as the dark channel prior based transmission estimation.Finally,to improve the image quality and get color-bright and haze-free effects,we design a new saturation compensation method based on the atmospheric scattering model in the saturation channel of HSV color space,to compensate the saturation of the preliminary defogging results.Compared with the state-of-the-art defogging methods,our algorithm can effectively improve the defogging speed.Moreover,high-contrast,and clear restoration results can be obtained without color distortion.