Optimized Rate-Distortion Extraction with Quality Layers
Isabelle Amonou, Nathalie Cammas, S. Kervadec, Stéphane Pateux · 2006
In this paper, we present the concept of quality layers that has been introduced in Scalable Video Coding (SVC) amendment of MPEG4-AVC. The quality layers are used to evaluate and signal the impact on rate and distortion of the various enhancement information pieces. Using this quality layers information allows to gain up to 0.5 dB with respect to the basic standard Sofware Verification Model (SVM) extractor that was proposed initially in SVC. Thanks to the signaling of this information in the header of the Network Abstraction Layer (NAL) units, the rate adaptation can be performed with a simple parser, e.g. for rate adaptation in an intelligent network node.