摘要
针对颜色衰减先验去雾算法存在的透射率估计不准确、整体颜色偏暗及去雾效果欠佳等问题,提出一种基于改进动态大气散射系数函数的图像去雾方法。首先,定义大气散射系数,其为一种图像景深与图像景深指数函数乘积形式的函数。其次,利用平均梯度和信息熵相结合的归一化综合评价参数(CEP)对随机选取的非均匀雾图像进行实验,确定动态大气散射系数函数的两个最佳参数分别为1.3和0.5。最后,借助天空区域透射率纠正算法来校正动态大气散射系数函数。实验结果表明,所提方法能够有效地解决原去雾算法存在的问题,从而进一步提升了图像去雾效果。
This study proposes an image-dehazing method based on an improved dynamic atmospheric scattering coefficient function to solve the problems of inaccurate transmittance estimation,overall dark color,and poor dehazing effect in color-attenuation prior dehazing algorithm.First,the atmospheric scattering coefficient is defined as a function of the product of image depth-of-field and image depth-of-field exponential function.Next,normalized comprehensive evaluation parameter(CEP)is obtained by combining the average gradient and information entropy to test randomly selected non-uniform hazing images.The two optimum parameters of dynamic atmospheric scattering coefficient function are determined as 1.3 and 0.5.Finally,the dynamic atmospheric scattering coefficient function is corrected by the sky region transmittance correction algorithm.The experimental results show that the proposed method can effectively solve the problems of the original dehazing algorithm,thus,the image-dehazing effect is further improved.
作者
郑毅
高强
王笛
胡辽林
Zheng Yi;Gao Qiang;Wang Di;Hu Liaolin(School of Mechanical and Precision Iustrument Emgineering,Xi'an University of Technology,Xi'an,Shaanai 710048,China)
出处
《激光与光电子学进展》
CSCD
北大核心
2020年第18期63-70,共8页
Laser & Optoelectronics Progress
基金
陕西省自然科学基金(2014JM7273)。
关键词
图像处理
动态大气散射系数
颜色衰减先验
图像去雾
天空区域
image processing
dynamic atmospheric scattering coefficient
color attenuation prior
image dehazing
sky region
