A comparative study of 3D transmission formats for 4K auto-stereoscopic displays

In this contribution some 3D plus depth (3D+2Z) schemes currently discussed at JCT-3V (ITU/MPEG) for the coding and transmission of 3D content are compared to the well-known and simpler 2D plus depth (2D+Z) approach, i.e. the transmission of only one view and its related depth map. The results of this comparison are presented with regard to coding performance and quality of the base and virtual views. In the light of these results, the 2D+Z format turns out to be competitive over a broad range of requirements for the delivery of consumer auto-stereoscopic 3D TV. Such a format is much easier to implement and to compress than competing 3D+2Z solutions. It can better deal with low quality depth maps and it doesn't need to bother with left-right consistencies.