摘要
为了考察误差图像空间尺寸和形状对视觉感受的影响,提出了将亮度加权扩展和空间相似分解相结合的图像质量评价模型,考察了图像的频率误差特性,体现了图像的结构误差特性。通过调整实验参数,将相似分解、亮度加权、多通道相似分解及传统客观性噪比4种图像质量评价方法做了对比,分析和实验结果表明,该模型与亮度加权扩展、空间相似分解2个独立模型相比,图像质量评价结果与视觉感受进一步吻合,稳定性明显提高。
To investigate the influence of error in image space size and shape on the subjective perception,a model by combing brightness weighted network and spatial similarity decomposition was presented.In the experiment,by adjusting the experimental parameters,space similarity decomposition,brightness weighted,multi-channel similar decomposition and the traditional peak signal to noise ratio of four kinds of image quality evaluation methods were compared.Experimental results showed that brightness weighted spatial similarity decomposition could be consistent with the subjective perception further compared with brightness weighted network model or space similarity decomposition model and the stability was improved obviously.
出处
《计算机工程与设计》
CSCD
北大核心
2014年第6期2056-2059,共4页
Computer Engineering and Design
基金
国家自然科学基金项目(60872162)
关键词
亮度加权
多通道分解
相似分解
空间结构
图像质量
brightness weighted network
multi-channel decomposition
similarity decomposition
space structure
image quality
