基于相位的结构相似度图像质量评价模型  

A QUALITY ASSESSMENT MODEL FOR IMAGES WITH STRUCTURE SIMILARITY BASED ON PHASE

作　　者：李爱华[1] 

机构地区：[1]南昌大学抚州医学分院,江西抚州344000

出　　处：《计算机应用与软件》2014年第11期233-236,共4页Computer Applications and Software

摘　　要：基于结构相似度的图像质量评价算法忽视了图像自身的相位特征信息,不能较好地评测某些失真类型的图像质量。针对这一问题,提出一种基于相位的结构相似度图像质量评价模型(PSSIM)。该模型在结构相似度的基础上,进一步考虑了图像的相位一致性特征,设计了一种图像质量评价函数的计算方法。实验仿真结果表明,PSSIM是一种有效的图像质量评价方法,同时该方法和人类主观评价感受具有良好的一致性。The image quality assessment algorithm based on structural similarity ignores the phase feature information of the image its own,thus fails in well assessing and measuring the quality of some images in distortion type.Aiming at this issue,in this paper we propose a phase-based quality assessment model for images with structural similarity（ PSSIM）.Based on the structural similarity,the model further considers phase congruency feature of the image,and a computation method of assessment function for image quality is designed.Experimental simulation results demonstrate that PSSIM is an effective image quality assessment method.Meanwhile,it has good consistency with the subjective assessment feelings of human beings.

关 键 词：图像质量评价 结构相似度 相位特征 人类视觉系统 

分 类 号：TP302[自动化与计算机技术—计算机系统结构]