基于人类视觉的感知立体图像质量评价方法

为了实现对不同失真类型的立体图像进行质量评 价,提出了一种基于人类视觉的立体图像 质量客观评价方法,分别从图像清晰度与立体感两方面进行评价。图像清晰度方面,将原始 与失真立体图 像分解为5个带通图像后利用对比度敏感度函数(CSF)优化各失真带通图像,并模拟掩 盖效应,通过整合各原始 带通图像,综合感知误差,构造信噪比(SNR)作为评价图像 清晰度的性能指标；立体感方面,对绝对差值图像进 行视觉感知模拟,建立SNR指标评价立体感的优劣。实验结 果表明,对不同失真类型立体图像的评价 结果表明,Pearson线性相关系数(PLCC)与Spearman等级 相关系数(SRCC)均优于现有评价方法。

A perceptual stereoscopic image quality assessment method based on human visual system

In order to achieve quality assessment for the distorted stereo images with different types of distortions,a stereo image quality objective assessment method based on human vi sual system (HVS) is proposed in this paper.This method assesses the stereo image in two aspects,imag e fidelity and stereo sense.In view of the image quality,the original and distorted stereo images ar e divided into five bandpass images.The contrast sensitivity function plays an important role in op timizing the distorted bandpass images and masking effect is simulated.The original bandpass images an d the perceptive errors are synthesized and the signal-to-noise ratio as the performance index of evaluating stereo image quality.As for the stereo sense,this paper focuses on the absolute dispa rity images and simulates them.At last,the signal-to-noise ratio is calculated.The experimental results show that the Pearson linear correlation coefficient (PLCC) index and the Spearman rank correction coefficient (SRCC) index by using the pr oposed method are both superior to existing assessment methods in the assessment results for stereo images with different ty pes of distortions.