VIDEO CODING WITH COMPOSITION AND QUALITY ADAPTATION BASED ON DEPTH DERIVATIONS

摘要

Techniques for coding video data estimate depths of different elements within video content and identify regions within the video content based on the estimated depths. One of the regions may be assigned as an area of interest. Thereafter, video content of a region that is not an area of interest may be masked out and the resultant video content obtained from the masking may be coded. The coded video content may be transmitted to a channel. These techniques permit a coding terminal to mask out captured video content prior to coding in order to support coding policies that account for privacy interests or video composition features during a video coding session.

主权项

1. A method, comprising:
estimating depth of different elements within video content, identifying regions within the video content based on the estimated depths, assigning a region to be an area of interest, masking video content of a region that is not an area of interest, coding resultant video content obtained from the masking, and transmitting the coded video content to a channel.