一种基于图像细节保持的空间色域映射算法框架  

A Spatial Gamut Mapping Framework Based on Detail Preservation

作　　者：朱明 焦会敏 赵兴运 杨晶琨 ZHU Ming;JIAO Hui-min;ZHAO Xing-yun;YANG Jing-kun(School of Materials and Chemical Engineering,Henan University of Engineering,Zhengzhou,Henan 450007,China;School of Light Industry and Engineering,Qilu University of Technology,Jinan,Shandong 250353,China)

机构地区：[1]河南工程学院材料与化学工程学院,河南郑州450007 [2]齐鲁工业大学轻工科学与工程学院,山东济南250353

出　　处：《北京理工大学学报》2020年第3期290-297,共8页Transactions of Beijing Institute of Technology

基　　金：国家自然科学基金资助项目(61301231);2017年度河南省高等学校青年骨干教师培养计划(2017GGJS154);2019年河南省高等学校大学生创新训练计划项目(S201911517005)。

摘　　要：鉴于逐点色域映射在图像细节保持方面及空间色域映射在光晕抑制方面的缺陷,本文利用自引导滤波在图像细节提取和光晕抑制方面的特性,提出了一种既能有效保护图像细节又能消除滤波光晕的空间色域映射算法框架.详细分析了引导滤波参数对映射图像细节再现的影响,进而确定了算法框架中的最佳滤波参数设置,还针对分层映射的特点修正了框架中的CUSP和HpMinDE逐点色域裁剪算子.评价实验证明:新算法无论在图像偏好性还是复制准确性方面,都要明显强于其他两种同类型算法以及经典的优化迭代类算法.通过对光晕测试标准图像的映射试验,比较了本文算法与其他细节保持类算法所产生的光晕效果,验证了新算法在光晕抑制方面的优越性.Considering the defects of point wise gamut mapping algorithms(GMAs)in detail-preservation and the halos introduced by spatial gamut mapping algorithms(SGMAs),a SGMA framework was proposed to effectively preserve the image detail and inhibit spatial filtering halos.Firstly,the influences of different Guided-filtering parameters on detail-reproduction were analyzed in detail to determine the optimal filtering parameters.Then,two point wise gamut mapping operators“CUSP”and“HpMinDE”were modified for the proposed SMGA framework according to the characteristics of layered gamut mapping.Finally,some evaluating experiments were carried out.The results show that,in aspects of image preference and reproduction accuracy,the proposed SGMA is obviously better than the other two classical algorithms of the same type and the classical optimization iterative algorithm.The superiority of the proposed SGMA in halos-inhibition were also tested and verified by means of the standard halos-testing image.

关 键 词：空间色域映射 自引导滤波 细节保持 光晕 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TP37[自动化与计算机技术—计算机科学与技术]