质降参考图像质量评价方法研究  

A Survey on Quality Assessment Approaches for Reduced Reference Image

作　　者：王旭[1] 彭宗举[1,2] 杨铀[2] 邵枫[1] 蒋志迪[1] 

机构地区：[1]宁波大学电路与系统研究所,浙江宁波315211 [2]中国科学院计算技术研究所,北京100080

出　　处：《宁波大学学报（理工版）》2009年第4期506-510,共5页Journal of Ningbo University：Natural Science and Engineering Edition

基　　金：浙江省教育厅科研计划项目(20070954;20070957;20070962)

摘　　要：图像质量评价是图像处理研究中的热点问题之一,而质降参考图像质量评价方法作为它的一个重要研究方向,可实际应用在码率控制等领域中.对质降参考评价模型进行相关研究,根据采用特征参数提取方法以及质量分数度量方式详细讨论了现有模型的特点和局限性,并对比较经典的评价模型进行了综述,以及对质降参考质量评价方法的发展做了预测和展望.Image quality assessment has been drawing the keenest research interests in the field of image processing,and reduced reference image quality assessment（RRIQA） can be applied in many areas such as rate control.RRIQA is studied in this paper,and characteristics and limitations of current image assessment metrics are reviewed and analyzed based on the adopted feature extraction and quality pooling methods.Some typical metrics are given more particular attention in the analysis.The paper ends with discussing the possible future study trend for RRIQA.

关 键 词：质降参考 人类视觉系统 特征提取 数字水印 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]