摘要
针对现有去雾后图像质量评价算法少、针对性弱和有效性差等问题,本文提出一种基于分类学习的去雾后图像质量评价算法.该算法通过分析去雾后图像本身所蕴含的质量特征,提取出基于图像增强、图像复原、统计先验以及人类视觉系统(Human visual system,HVS)的度量指标;并在本文数据库基础上,利用支持向量机(Support vector machine,SVM)将质量评价问题转换为分类问题.实验结果表明,该算法与已有评价方法相比,在获得高效分类评价结果的同时,具有较好的实用性和主观一致性.
Since existing quality assessment methods suffer from poor pertinence and low efficiency, a novel quality assessment method based on classified learning for dehazed images is proposed. In this paper, firstly the metrics interms of image enhancement, image restoration, statistical prior, and human visual system are extracted by analyzing qualitative characteristics of images after haze removal. Then the quality assessment problem is converted to the classification problem by means of support vector machine using our database. Experimental results demonstrate that compared with other state-of-the-art methods the proposed method is highly efficient and practical with subjective and objective consistency.
出处
《自动化学报》
EI
CSCD
北大核心
2016年第2期270-278,共9页
Acta Automatica Sinica
基金
国家自然科学基金(61372167
61379104)资助~~
关键词
图像去雾
图像质量评价
支持向量机
暗通道先验
人类视觉系统
Image dehazing
image quality assessment
support vector machine（SVM）
dark channel prior（DCP）
human visual system（HVS）
