Neighboring block disparity vector derivation in 3D video coding

摘要

In one example of the disclosure, a method of coding video data comprises coding video data using texture-first coding, and performing an NBDV derivation process for a block of the video data using a plurality of neighboring blocks. The NBDV derivation process comprises designating a motion vector associated with a neighboring block of the plurality of neighboring blocks coded with a block-based view synthesis prediction (BVSP) mode as an available disparity motion.

主权项

1. A method of coding a block of video data, the method comprising:
performing a neighboring block-based disparity vector (NBDV) derivation process to derive a disparity vector for the block of video data; and based on a sub-region of the block of video data being determined to be coded using block-based view synthesis prediction (BVSP) mode, the method further comprising:
refining, using BVSP mode, the disparity vector produced by the NBDV derivation process for the sub-region to create a refined disparity vector; andcoding the sub-region of the block of video data using BVSP mode and the refined disparity vector, wherein refining the disparity vector includes at least one of:
selecting a depth value from one or more depth pixels of a reference block identified by the disparity vector produced by the NBDV derivation process for the sub-region of the block of video data, the one or more depth pixels being in a depth view component of the reference block, but not full pixels within a proximity to a center of a depth block in a reference depth view which is located by the disparity vector; orsetting a vertical component of the disparity vector produced by the NBDV derivation process to be zero.