Home           Contact us           FAQs           
     Journal Home     |     Aim & Scope    |    Author(s) Information      |     Editorial Board     |     MSP Download Statistics
2014 (Vol. 7, Issue: 14)
Article Information:

Visual Perception Based Objective Stereo Image Quality Assessment for 3D Video Communication

Gangyi Jiang, Xiangyin Mao, Mei Yu, Zongju Peng and Feng Shao
Corresponding Author:  Gangyi Jiang 

Key words:  3D video communication, human visual system, image quality assessment, stereo image processing, visual perception , ,
Vol. 7 , (14): 2827-2837
Submitted Accepted Published
November 29, 2012 March 07, 2013 April 12, 2014

Stereo image quality assessment is a crucial and challenging issue in 3D video communication. One of major difficulties is how to weigh binocular masking effect. In order to establish the assessment mode more in line with the human visual system, Watson model is adopted, which defines visibility threshold under no distortion composed of contrast sensitivity, masking effect and error in this study. As a result, we propose an Objective Stereo Image Quality Assessment method (OSIQA), organically combining a new Left-Right view Image Quality Assessment (LR-IQA) metric and Depth Perception Image Quality Assessment (DP-IQA) metric. The new LR-IQA metric is first given to calculate the changes of perception coefficients in each sub-band utilizing Watson model and human visual system after wavelet decomposition of left and right images in stereo image pair, respectively. Then, a concept of absolute difference map is defined to describe abstract differential value between the left and right view images and the DP-IQA metric is presented to measure structure distortion of the original and distorted abstract difference maps through luminance function, error sensitivity and contrast function. Finally, an OSIQA metric is generated by using multiplicative fitting of the LR-IQA and DP-IQA metrics based on weighting. Experimental results shows that the proposed method are highly correlated with human visual judgments (Mean Opinion Score) and the correlation coefficient and monotony are more than 0.92 under five types of distortions such as Gaussian blur, Gaussian noise, JP2K compression, JPEG compression and H.264 compression.
Abstract PDF HTML
  Cite this Reference:
Gangyi Jiang, Xiangyin Mao, Mei Yu, Zongju Peng and Feng Shao, 2014. Visual Perception Based Objective Stereo Image Quality Assessment for 3D Video Communication.  Research Journal of Applied Sciences, Engineering and Technology, 7(14): 2827-2837.
    Advertise with us
ISSN (Online):  2040-7467
ISSN (Print):   2040-7459
Submit Manuscript
   Current Information
   Sales & Services
Home  |  Contact us  |  About us  |  Privacy Policy
Copyright © 2015. MAXWELL Scientific Publication Corp., All rights reserved