Asymmetrically frame-compatible depth video coding

To achieve efficient coding while maintaining the quality of rendered views, a frame-compatible coding scheme for two-view depth video is proposed. The idea is to remove the redundancy within the inter-view depth video before packing and encoding. The content in two views is analysed after performing view warping from the primary view to the secondary view. The uncovered region is defined as the missing region where no corresponding information can be found in the primary view. Only the uncovered region is retained in the secondary view and will be combined with the primary view to form a frame-compatible video. The experimental results indicate that the proposed scheme outperforms the conventional frame-compatible technology in providing better objective and subjective quality.