Encoding of computer-generated video content

摘要

An example method for encoding computer-generated video content may comprise: generating, based on a geometric model, a sequence of video frames. The example method may further comprise determining, based on the geometric model, a velocity map comprising a plurality of velocity values for a plurality of pixels, wherein each velocity value represents a difference between a viewport position of a pixel in a current frame and a previous frame. The example method may further comprise determining, based on the velocity map, a plurality of motion vectors, wherein each motion vector corresponds to a macroblock comprising one or more pixels of a video frame of the sequence of video frames. The example method may further comprise producing, by a processing device, an output bit stream representative of the sequence of video frames, by encoding, using the motion vectors, the sequence of video frames.

主权项

1. A method, comprising:
determining, based on a geometric model employed for generating a sequence of a video frames, a velocity map comprising a plurality of velocity values for a plurality of pixels, each velocity value representing a difference between a viewport position of a pixel in a current video frame and a previous video frame of the sequence of a video frames; determining, based on the velocity map, a plurality of motion vectors, wherein each motion vector of the plurality of motion vectors represents a displacement form a first location, within the current video frame, of a macroblock comprising one or more pixels of the current video frame relatively to a second location of the macroblock within the previous video frame; and producing, by a processing device, an output bit stream representative of the sequence of video frames, by encoding, using the motion vectors, the sequence of video frames.