2007 Fiscal Year Final Research Report Summary
Development of user-centered quality on-demand home theater delivery system
| Project/Area Number |
18500078
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Media informatics/Database
|
|---|
| Research Institution | University of Toyama |
|---|
Principal Investigator |
HORITA Yuukou University of Toyama, Graduated school of science and technology, Professor (80209303)
|
|---|
| Project Period (FY) |
2006 – 2007
|
| Keywords | Av media / Quality evaluation / Video coding / Home theater / MPEG-2 / Semantic Differential method / Realistic / 視覚優位 |
|---|
| Research Abstract |
1) About the quality evaluation for coded video
In this research, we have analyzed the relationship between subjective video quality and subjective frame quality of coded video by using the result of subjective quality assessment. Consequently, subjective video quality could estimate with high accuracy from subjective frame quality of coded video. However, the value of outlier ratio were varying according to scene contents and bit rates. In consideration of overestimation and underestimation of the estimated value of video quality, we found out that the variation of outlier ratio was connected with the influence of motion information such as camera work and local motion as the temporal degradation.
2) About the quality evaluation for coded audio-visual media
In this research, we have investigated an interaction between video quality and audio quality in order to develop an objective multimedia quality assessment model of high-quality audio-visual media (high-definition video and multi-channel audio) according to ITU-T Recommendation J.148. For that purpose, we have analyzed the interaction by using the subjective quality assessment test of high-definition video and multi-channel audio. As a result, we have concluded that the following three procedures should be considered for the development of the objective multimedia quality assessment model of high-quality audio-visual media: (1) estimation of Aq and Vq from psychophysical quantity of video and audio media, (2) estimation of Aq(Vq) and Vq(Aq) from Aq and Vq, (3) estimation of MMq from Aq, Vq, Aq(Vq) and Vq(Aq) as compliant contetns.
|
