An AVS- and object- based approach to scalable coding of plenoptic videos

An AVS- and object- based coding approach to support scalable coding of plenoptic videos (PVs) is proposed. PVs are simplified dynamic light fields in which the videos are taken at regularly spaced locations along line segments, which can provide a continuum of view point using view synthesis. In previous works, we have explored the object-based coding of PVs using the MPEG-4 object coding framework so as to facilitate its rendering and processing. This paper further studies the object-based coding of PVs based on the more efficient Audio Video Standard of China (AVS) and incorporate scalability for supporting users with different bandwidths. AVS is China´s national standard and it adopts an MPEG-like hybrid compression framework. The proposed AVS-based object coding framework extend AVS by employing temporal and spatial prediction to support multi-view coding for better coding efficiency. Reduced inter-coding dependency is also utilized to facilitate the issue of random access and selective transmission. Moreover, object-based temporal scalability based on a hierarchical-P structure is also illustrated.