摘要
图像去雾是计算机视觉的重要研究方向之一。传统图像去雾算法存在去雾不彻底、去雾图像对比度低、halo效应、色彩失真等问题。针对上述问题,提出了一种改进的非局部先验的图像去雾算法。该算法使用基于中智学的模糊C均值聚类算法和一种混合暗通道先验的透射率优化方法,改进了传统非局部先验的图像去雾算法在大气光估计、雾线定位和透射率优化过程中存在的问题。结果表明,与几种常用的图像去雾算法相比,提出的算法在去雾图像的大气光估计、客观分析和主观分析等方面均有一定的优势。
Image dehazing is an important research direction of computer vision.The traditional image dehazing algorithms have some problems,such as incomplete fog removal,low contrast,block effect and color distortion.In order to solve these problems,an improved non-local image dehazing algorithm is proposed The fuzzy C-means clustering algorithm based on neutrosophy and a hybrid dark channel prior’s transmission estimation map optimization method are used to improve the traditional non-local prior image dehazing algorithm.The results show that the algorithm proposed has a lot of advantages in atmospheric light estimation,objective analysis and subjective analysis compared with several commonly used image defogging algorithms.
作者
于坤
焦青亮
刘子龙
蒋依芹
刘玉芳
YU kun;JIAO Qingliang;LIU Zilong;JIANG Yiqin;LIU Yufang(Henan Key Laboratory of Infrared Materials&Spectrum Measures and Applications,School of Physics,Henan Normal University,Xinxiang 453007,China;Division of optics,National Institute of Metrology,Beijing 100013,China)
出处
《光学技术》
CAS
CSCD
北大核心
2020年第4期476-482,共7页
Optical Technique
基金
国家自然科学基金(61875180)
国家重点研发项目(2016YFC0103700,2017YFF0205103)
河南师范大学研究生科研创新基金(YL201801)。
关键词
信息光学
图像去雾
非局部先验
中智学
混合透射率优化
Information optics
Image dehazing
Non-local Prior
Neutrosophy
Hybrid transmittance optimization