首页 | 本学科首页   官方微博 | 高级检索  
     

一种基于结构相似性的图像质量评价指标
引用本文:何洁,李蕴奇.一种基于结构相似性的图像质量评价指标[J].长春理工大学学报,2014(3):105-108.
作者姓名:何洁  李蕴奇
作者单位:吉林省经济信息中心,长春130061
摘    要:图像质量评价方法是控制传输、压缩和处理对图像质量影响的关键技术,也是多媒体系统中不可或缺的一部分。提出一种基于结构相似理论的图像质量评价指标,首先将待评价的图像分割成若干互不相交的图像块;然后分别计算图像块的显著性和结构相似度;最后以每个图像块的相对显著性作为权重,计算整幅图像的结构相似度。该模型时间复杂度低,同时实验表明该指标的性能明显优于传统指标。

关 键 词:图像  质量评价  结构相似  谱冗余

Image Quality Evaluation Index based on Structural Similarity
HE Jie,LI Yunqi.Image Quality Evaluation Index based on Structural Similarity[J].Journal of Changchun University of Science and Technology,2014(3):105-108.
Authors:HE Jie  LI Yunqi
Affiliation:(Jilin Economic Information Center, Changchun 130061)
Abstract:Image quality assessment is one of the key technologies to control the effect of transmission, compression and processing on the quality of images,and it is also an integral part of multimedia systems. In this paper,an evalua-tion measure for image quality based on structural similarity theory is presented. First of all, the evaluated images are divided into several disjoint image blocks;and then the saliency and structural similarity of image blocks are calculated respectively. Finally,the structural similarity of the whole image with the saliency of each image block as weight is cal-culated. The time complexity of the model is low and the experiments show that the measure is superior to traditional ones.
Keywords:images  quality evaluation  structural similarity  spectral residual theory
本文献已被 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号