Abstract: In order to remove the effect of weather in degraded image, this paper proposes an improved image defogging algorithm by minimum distortion. Imporoved algorithm divides up the area of sky automatically and gets the optimal estimate of original signal based on minimum distortion rule in information theory, and acquires the dichromatic atmosphere scattering model of all the required parameters automatically. It realizes a fog degraded image de-fogged successfully. Simulation results in the Matlab platform demonstrate that improved algorithm has good real-time and self-adaptability.

Key words: image restoration, minimum distortion, foggy image

摘要： 为了从降质图像中去除天气的影响，提出改进的最小失真意义下图像去雾算法。改进算法基于二色大气散射模型，通过对天空区域的自动划分以及从信息论角度求得最小失真意义下原始图像的最优估计，从而一次性自动获取所需复原参数，实现单幅雾天降质图像复原。在Matlab平台上的仿真结果表明，改进算法具有较强的实时性和自适应性。

关键词: 图像复原, 最小失真, 雾天图像

CLC Number: 

• TP391.9