基于透射率修正的暗通道去雾改进方法  

Improved Method of Dark Channel Defogging Based on Transmittance Correction

作　　者：许烁[1] 王栋 何俊红 

机构地区：[1]上海大学机电工程与自动化学院,上海200072

出　　处：《工业控制计算机》2020年第7期113-115,共3页Industrial Control Computer

摘　　要：针对暗通道先验去雾算法在景深突变处和天空区域透射率估计不准确,从而造成景深突变处产生光晕现象和天空区域色彩失真,提出利用最小值像素暗通道和局部窗口暗通道分别求取透射率,然后在景深突变区域使用两者的加权值修正透射率,在天空区域采用容差机制修正透射率。实验结果表明,该去雾算法能够有效缓解光晕现象和天空区域的色彩失真。Aiming at the problem of the inaccurate estimation of the transmittance of the dark channel prior defogging algorithm in the region of depth abrupt change and sky areas,causing halo phenomenon and color distortion in the sky area,this paper proposes to uses the minimum pixel dark channel and the local window dark channel to obtain the transmittance separately,and then uses the weighted value of the two to modify the transmittance in the region of depth abrupt change,and use a tolerance mechanism to modify transmittance in sky areas.

关 键 词：暗通道图像去雾 透射率加权融合 容差机制 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]