计算机工程与应用2016，Vol.52Issue(7)：166-170,5.DOI:10.3778/j.issn.1002-8331.1412-0227
基于暗原色先验的区域自适应图像去雾方法
Regional adaptive image haze removal method based on dark channel prior
摘要
Abstract
Images taken in foggy weather are seriously degraded due to the scattering of atmospheric particles. A simple and effective haze removal algorithm from a single image is proposed. Firstly, a halo evaluator is designed to detect halo zone. Then, precise transmission rat is obtained by weight fusion of single pixel based rate and block area based one, both taking the prior of dark channel. The weight is determined according to the halo evaluator. Finally, a parameter is added for image recovery to limit the low transmission and to protect the sky area. Experiments show that compared with other methods, more vivid and natural images can be recovered by the proposed method, especially at the edges of the fore-ground and the background and in the sky area.关键词
去雾/暗原色先验/加权/透射率Key words
dehazing/dark channel prior/weighting/transmission分类
计算机与自动化引用本文复制引用
刘冬冬,陈莹..基于暗原色先验的区域自适应图像去雾方法[J].计算机工程与应用,2016,52(7):166-170,5.基金项目
国家自然科学基金（No.61104213）；江苏省自然科学基金（No.BK2011146）。 （）
