| 摘要 |
A video decoder includes a mode of operation for reverse playback that includes dividing a sequence of coded video data into groups of pictures (GOPs) and further subdividing the groups of pictures into a plurality of spans. GOPs extend in coding order from a first key frame to a terminal frame that precedes another key frame in coding order. Each span is processed in reverse coding order by decoding the first key frame and any reference frames that precede the currently processed span in coding order, selecting a plurality of frames from the currently processed span for decode and display and further selecting for decode but not display any reference frames from the currently processed span that were omitted by the first selection. All selected frames are decoded but only those marked for display are rendered at a display device. The method accommodates control parameters that define the number of spans selected from each GOP and the number of frames selected from each span for decoded and display. Via these control parameters, operation of the method may be tuned to local decoding environments for which processing resources and decoded picture buffer sizes may be unknown.
|
