摘要
在雾、霾等天气条件下,大气粒子的散射作用导致图像严重降质。本文提出一种简单快速的基于物理模型的图像去雾新算法,对大气散射模型进行化简,得到新的去雾模型。然后,利用暗原色先验方法估计大气光值A,并代入新的简化模型,得到去雾图像。实验表明,该算法在处理速度和去雾效果上都优于现有算法。
In the fog or haze weather, the scattering of atmospheric particles leads to serious image degradation. This paper proposes a simple and fast algorithm based on physical model to simplify the atmospheric scattering model and obtain a new defogging model. Then it uses the dark channel prior method to estimate the atmospheric light value A, and puts A into a new simplified model to get the fog image. Experiments show that the algorithm is superior to the existing algorithm in processing speed and defogging effect.
作者
张阳
王璟
肖梦
徐国庆
Zhang Yang Wang Jing Xiao Meng Xu Guoqing(School of Computer Science and Engineering, Wuhan Institute of Technology, Wuhan 430205, Hubei Computer Science and Engineering Experimental Classes, Wuhan Institute of Technology, Wuhan 430205, Hubei)
出处
《电脑与电信》
2017年第6期7-9,共3页
Computer & Telecommunication
基金
湖北武汉工程大学计算机科学与工程学院院长基金项目资金支持,项目编号:2017045
国家自然科学基金,项目编号:61502354
关键词
图像去雾
物理模型
暗原色先验
大气光值A
image defogging
physical model
dark channel prior
atmospheric light value A
