基于暗通道补偿和透射率修正的图像去雾算法  

IMAGE DEFOGGING ALGORITHM BASED ON DARK CHANNEL COMPENSATION AND TRANSMITTANCE CORRECTION

作　　者：容杰雄 张鸿[2] Rong Jiexiong;Zhang Hong(College of Computer Science and Technology,Wuhan University of Science and Technology,Wuhan 430065,Hubei,China;Hubei Province Key Laboratory of Intelligent Information Processing and Real-Time Industrial System,Wuhan University of Science and Technology,Wuhan 430065,Hubei,China)

机构地区：[1]武汉科技大学计算机科学与技术学院,湖北武汉430065 [2]武汉科技大学智能信息处理与实时工业系统湖北省重点实验室,湖北武汉430065

出　　处：《计算机应用与软件》2023年第4期166-171,262,共7页Computer Applications and Software

基　　金：国家自然科学基金项目(61373109)。

摘　　要：针对传统暗通道去雾算法应用于含有天空区域的有雾图像时出现的色彩失真以及细节缺失的问题,提出一种基于暗通道补偿和透射率修正的图像去雾算法。使用均值滤波替代最小值滤波,对暗通道进行补偿,利用补偿的暗通道对大气光值的预估方法进行优化;根据像素点与大气光点距离的远近对图像的初始透射率进行了修正,并使用快速引导滤波进一步细化图像的透射率;通过大气散射模型将透射率图像恢复为无雾图像。实验结果表明,该算法得到的复原图像具有较好的色彩保真和细节保持能力,在峰值信噪比以及结构相似性上都取得了提升,同时该算法在执行时间上缩短为原来的一半。Aimed at the problem of color distortion and lack of detail when the traditional dark channel defogging algorithm is applied to the foggy image containing the sky area,an image defogging algorithm based on dark channel compensation and transmittance correction is proposed.The mean filter was used to replace the minimum filter to compensate the dark channel.The compensated dark channel was used to optimize the estimation method of the atmospheric light value.The initial transmittance of the image was corrected according to the distance between the pixel point and the atmospheric light point,and the fast guided filter was used to further refine the transmittance of the image.The transmittance image was restored to a fog-free image through the atmospheric scattering model.The experimental results show that the restored image obtained by this algorithm has good color fidelity and detail retention ability,and the peak signal-to-noise ratio and structural similarity are improved.At the same time,the execution time of this algorithm is reduced to half of the original.

关 键 词：暗通道去雾 均值滤波 暗通道补偿 透射率修正 快速引导滤波 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]