针对雾图能见度低和去雾图像亮度偏暗的问题,提出一种基于大气散射模型的双阶段去雾算法。首先使用线性变换估算复原图像亮度,使用拉伸方法估算复原图像饱和度,根据复原图像亮度、饱和度估算其最小通道,联合雾图最小通道获取粗糙透射率。在不同阶段分别使用双梯度代价函数、导向滤波优化粗糙透射率,依据大气散射模型复原图像和增强亮度。实验结果表明,所提算法复原图像更清晰明亮；图像综合质量、峰值信噪比和运行时间等客观指标均值优于所有比较算法,其中图像综合质量最少提高1.55倍,运行速度最少加速1.50倍。所提算法有效地增强了雾图的能见度和明亮度。

Aiming at the problem of low visibility of hazy images and dim brightness of dehazed image,a two-stage dehazing algorithm based on atmospheric scattering model was proposed.Firstly,the minimum channel of hazy-free image was derived by linearly transforming of brightness of hazy image and stretching saturation of hazy image, which was used to calculate rough transmittance combining with the minimum channel of the hazy image.The rough transmittance were optimized separately by double gradient cost function or guided filtering in different stages.Finally,the hazy-free image was restored by the atmospheric scattering model that would be used again to improve brightness for dehazed image.The experimental results show that the restored image become clearer and brighter after hazy removal;The average of objective indicators such as image comprehensive quality, peak signal-to-noise ratio and running time are superior to compared algorithms, in which the image comprehensive quality is improved by at least 1.55 times and the running speed is accelerated by at least 1.50 times on average.The proposed algorithm effectively improves the visibility and brightness of hazy image.

贵州省教育厅青年科技人才成长项目(黔教合 KY 字[2018]165)和贵州省教育厅创新群体项目(黔教合KY字[2021]015)资助项目