Bit-Rate Adaptive Downsampling for the Coding of Multi-View Video with Depth Information

摘要: In this paper, the potential for improving compression efficiency of multi-view video coding with depth information is explored. The proposed technique uses downsampling prior to encoding, arbitrary views and maps. A bit-rate adaptive downscaling-ratio decision approach certain maps encoding. Colour videos are considered separately due their different characteristics effects on synthesized free view-point videos. inter-view references, if present, downsampled same resolution as input be coded. results several sequences indicate that using mixed spatial both can achieve savings in bit-rate, compared full fixed depth-to-colour ratio when quality viewpoints considered. computational complexity encoder significantly reduced at time, since number blocks coded reduced, hence block mode decisions carried out reduced.