期刊论文

Fast algorithm for image defogging by eliminating halo effect and preserving details

中文

计算机应用研究

1001-3695

2019

36

4

1228-1231

10.19734/j.issn.1001-3695.2017.11.0776

国家自然科学基金资助项目(61501198) 武汉市青年科技晨光计划资助项目(2014072704011248) 湖北省自然科学基金面上项目(2014CFB461) 华中师范大学中央高校基本科研业务费资助项目(CCNU14A05017).

本校为第一机构

针对现有的基于大气散射物理模型的图像去雾算法在去雾过程中大多无法避免地会产生光晕效应和细节丢失,提出了一种消除光晕效应和保持细节信息的图像快速去雾算法。首先运用四叉树子矩阵划分的分层遍历方法得到更精确的大气光值,再通过分析大气散耗函数,利用融合梯度信息的改进引导滤波得到精确估计的大气散耗函数,并自适应地获取最小值图像与大气光平均值的阈值,求解出透射图;最后反演复原出无雾图像,并对复原后的图像进行亮度调整。对多组有雾图像进行了实验,该算法能有效地抑制去雾过程中产生的光晕效应,较多地保留了图像的细节信息,且运行时间大约减少了一倍。融合梯度信息的改进引导滤波不...

Due to the existing image defogging algorithm based on atmospheric scattering model, the defogging process would inevitablely have a halo effect and details were lost, this paper proposed a fast defogging algorithm for eliminating halo effect and keep the detail information of the image. The algorithm first used space division level method based on four tree sub to get more accurate atmospheric light value. Through analysis of the atmospheric dispersion consumption function, it used the improved gradient information fusion filter to get the accurate estimation of atmospheric dispersion consump...