Method and apparatus of disparity vector derivation in 3D video coding

A method for processing three-dimensional or multi-view video, the method comprising:

• receiving pixel data associated with a current texture block of a current texture frame corresponding to a dependent view;
• determining a depth block for the current texture block based on a coded depth frame in a reference view;
• and
• determining a derived disparity vector for a partition of the current texture block based on a maximum value of two or more depth samples of a partition of the depth block such that interview residual prediction is applied to the pixel data utilizing the derived disparity vector,
• wherein the partition of the current texture block corresponds to the partition of the depth block, and
• wherein the derived disparity vector corresponds to a maximum value of four corner samples of the depth block.

An apparatus comprising:

• at least one circuit configured for:
• receiving pixel data associated with a current texture block of a current texture frame corresponding to a dependent view;
• determining a depth block for the current texture block based on a coded depth frame in a reference view;
• and
• determining a derived disparity vector for a partition of the current texture block based on a subset of depth samples of a partition of the depth block,
• wherein the derived disparity vector corresponds to a maximum value of four corner samples of the depth block, and
• wherein the partition of the current texture block corresponds to the partition of the depth block,
• the subset contains less depth samples than the partition of the depth block and
• the subset excludes a single-sample subset corresponding to a center sample of the partition of the depth block
• such that inter-view motion predictive encoding or decoding is applied to the pixel data utilizing the derived disparity vector.