图像去雾算法清晰化效果客观评价方法  

Objective Assessment Method for the Clearness E?ect of Image Defogging Algorithm

作　　者：郭璠[1,2] 蔡自兴[1] 

机构地区：[1]中南大学信息科学与工程学院,长沙410083 [2]先进控制与智能自动化湖南省工程实验室,长沙410083

出　　处：《自动化学报》2012年第9期1410-1419,共10页Acta Automatica Sinica

基　　金：国家自然科学基金(90820302;60805027);国家博士点基金(200805330005);湖南省院士基金(2009FJ4030)资助~~

摘　　要：针对目前去雾效果评价方法少和已有评价方法存在局限性等问题,提出了两种图像清晰化效果评价方法.一种借助由环境渲染或光路传播图所模拟的雾环境图像,采用全参考方式评估算法的去雾效果;一种从人类视觉感知的角度出发,采用无参考方式构建综合评价体系以全面衡量算法的去雾性能.实验证明两种方法均能有效地评价各算法的清晰化效果,且评估结果与人眼的主观感受相一致.本文所提评价方法分别从构建模拟雾环境和人类视觉感知两方面考虑,与已有评价方法相比,在获得全方面评估结论的同时,具有较好的实用性和可靠性.Since there is lack of methodology to assess the performance of defogging algorithm and the existing assessment methods have some limitations, two new methods for assessing the clearness effect of image defogging algorithm are proposed in this paper. One is using synthetic foggy image simulated by environment rendering or transmission map to assess defogging effect in full-reference way. The other is constructing assessment system from the perspective of human visual perception to assess the algorithm performance comprehensively in no-reference way. Experiments show that both methods can assess the effect effectively, and the evaluation results are consistent with our subjective perception. Compared with other existing methods, our proposed methods can assess the effect from the construction of simulated environment and human visual perception, respectively. The new methods can obtain a comprehensive assessment results as well as provide a good practicability and reliability.

关 键 词：图像去雾算法 清晰化效果 图像评价 模拟雾环境 人类视觉感知 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]