TECHNIQUES FOR PARALLEL VIDEO TRANSCODING

摘要

Various embodiments are directed to techniques for dividing compressed video data into portions for load balancing or redundancy among multiple video transcoders. A device includes an analysis component to analyze NAL units of compressed video data representing a motion video to identify a GOP refresh, and to compare a period of time represented by frames of the motion video following the GOP refresh to a maximum; a division component to divide the compressed video data at a first point coincident with the GOP refresh to define a start of a compressed video portion, and at a second point to define an end of the compressed video portion to include frames of the motion video representing a shorter period of time between the first and second points than the maximum; and an assignment component to assign the compressed video portion to a transcoding device. Other embodiments are described and claimed.

主权项

1. A device to transcode video imagery comprising:
an analysis component to analyze network abstraction layer (NAL) units of compressed video data that represents a motion video to identify a first group-of-picture (GOP) refresh, and to compare a period of time represented by frames of the motion video following the first GOP refresh to a predetermined maximum period of time; a division component to divide the compressed video data at a first point coincident with the first GOP refresh to define a start of a first compressed video portion, and at a second point to define an end of the first compressed video portion based on the comparison to include a first quantity of frames of the motion video that represents a shorter period of time between the first and second points than the predetermined maximum period of time; and an assignment component to assign the first compressed video portion to a first transcoding device to transcode the first quantity of frames.