Single Image Dehazing Using Improved Gray World Theory and Dark Channel Prior

被引：0

作者：

Zhang, Haopeng ^{[1
,3
,4
]}

Dong, Bo ^{[2
]}

Jiang, Zhiguo ^{[1
,3
,4
]}

机构：

[1] Beihang Univ, Image Proc Ctr, Sch Astronaut, Beijing 100191, Peoples R China

[2] Beijing Radio Measurement Res Inst, Beijing 100854, Peoples R China

[3] Minist Educ, Key Lab Spacecraft Design Optimizat & Dynam Simul, Beijing 100191, Peoples R China

[4] Beijing Key Lab Digital Media, Beijing 100191, Peoples R China

来源：

NEW FRONTIERS IN ARTIFICIAL INTELLIGENCE (JSAI-ISAI 2018) | 2019年 / 11717卷

基金：

中国国家自然科学基金;

关键词：

Image dehazing; Gray world theory; Dark channel prior; WEATHER;

D O I：

10.1007/978-3-030-31605-1_5

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

The images captured outdoor are usually influenced by inclement weather conditions severely, bringing a great deal of inconvenience to the automatic data processing system. The widely used dehazing method based on dark channel prior (DCP) is not effective on the image with color distortion. In order to solve this problem, we improve the gray world theory (GWT) and propose a single image dehazing method using our improved gray world theory and the dark channel prior. Experiments show that our method can restore the hazy image with color distortion effectively and outperforms the state-of-art results.

引用

页码：67 / 73

页数：7