In this paper, we propose an adaptation system that targets delivery of mobile scalable video contents in heterogeneous network environment. This system relies on a novel objective perceptual quality metric based on subjective quality in order to achieve video adaptation according to human visual perception, thus guaranteeing to maximize the quality of service (QoS).
QoS is getting on a core issue for service reliability and user gratification. For do so, Iit is extremely important how to assess the degree of the video quality in order to ensure the optimal QoS. There have been a number of studies on video quality measure. The majority have been restricted to quality measure including temporal and/or SNR scalability because of codec characteristics.
In this paper, we at first introduce a new objective perceptual video quality metric with full scalability for mobile H.264/AVC Scalable Video Coding (SVC) contents. SVC supports full scalability, each of which shows different visual effects. The metric concentrates upon the effect of the change of some parameters such as frame rate, SNR, spatial resolution, and motion characteristics through a subjective quality assessment. To demonstrate the efficiency of the proposed system, a test bed was implemented. As mobile video applications are specifically targeted in this paper, the spatial resolution of the video varies between QCIF and CIF resolution. Further, some description tools of MPEG-21 Digital Item Adaptation(DIA) are used to describe the characteristics of a particular mobile usage environment, as well as to communicate the properties of the scalable video content.
The experimental results demonstrate that the proposed metric has a high correlation to subjective quality and it is useful in determining the video quality of SVC. The results also show that the proposed system can provide clients video content with a higher subjective quality than other methods such as the conventional maximum-bit-allocation...