MPEG-21-based scalable bitstream adaptation using medium grain scalability

Abstract

H.264/AVC scalable video coding (SVC) aims at providing high video quality to users in heterogeneous multimedia usage environments. In order to create an adapted bitstream in an optimal way, an adaptation decision-taking algorithm is needed. This algorithm has to take into account the capabilities and the constraints of the targeted usage environment, as well as the properties of the scalable bitstream. In this paper, description tools part of MPEG-21 digital item adaptation (DIA) are used to characterize usage environments and scalable bitstreams, such as usage environment description (UED), universal constraints description (UCD), and adaptation quality of service (AdaptationQoS). Further, we also propose an adaptation decision-taking algorithm for the exploitation of medium grain scalability (MGS) in SVC. This algorithm was implemented in an adaptation decision-taking engine (ADTE). Our experimental results demonstrate the efficiency of the proposed adaptation decision-taking algorithm for the exploitation of MGS in SVC.