摘要

In this paper, a scalable multi-view video coding (SMVC) scheme is proposed to efficiently realize region of interest (ROI) support in SMVC. In addition, in order to further improve the coding efficiency, we develop a novel bit allocation model which can implement reasonable bit allocation among views. Extensive experiments are conducted on the typical multi-view video (MVV) sequences, and the efficiency of the presented codec is demonstrated. Meanwhile, the addressed SMVC architecture can easily provide temporal, spatial, SNR, view, and content scalabilities.