Derivation of depth map estimate

摘要

In some example techniques for generating and updating depth map estimates used for inter-view motion prediction and/or inter-view residual prediction for coding multiview video data, each of a plurality of dependent views is associated with a dependent depth view estimate, which may be generated or updated based on coding of the texture data of the dependent view relative to a base view. In such examples, each of the dependent depth map estimates may be warped to the base view to produce a respective one of a plurality of base depth map estimates. Each dependent depth map estimate and the respective base depth map estimate for a depth map estimate pair associated with the respective one of the plurality of dependent views.

主权项

1. A method of coding video data, the video data comprising at least one base view and a plurality of dependent views, wherein the plurality of dependent views depend on the base view, the method comprising:
for each of the dependent views, coding texture data for the dependent view for an access unit using disparity motion information for the dependent view relative to texture data for the base view for the access unit; generating, independent of any other dependent depth map estimate, a respective dependent depth map estimate for each respective dependent view of the plurality of dependent views for the access unit based on the disparity motion information for the respective dependent view; and warping each of the respective dependent depth map estimates to the base view to produce a respective one of a plurality of base depth map estimates for the access unit, wherein each of the dependent depth map estimates and the respective base depth map estimates form a respective one of a plurality of depth map estimate pairs for the access unit, and wherein each of the plurality of depth map estimate pairs for the access unit is associated with a respective one of the plurality of dependent views.