| 主权项 |
1. A method comprising:
identifying a plurality of frames of interest from a primary stream of encoded frames; selecting at least one of the frames of interest for inclusion in a separate redundant stream, wherein frame selection differentiates between IDR frames, I frames, and anchor frames based on respectively assigned levels, wherein the anchor frames may comprise an I frame, an IDR frame, or a forward predicted frame dependent only on a most recently decoded anchor frame; assigning a first level to an anchor frame only if a video encoder uses relative lower quantization values resulting in a higher number of bits; assigning a second level to a reference frame that is not an anchor frame, wherein the reference frame references one or more anchor frames; assigning a third level to an anchor frame that is a forward predicted frame; assigning a fourth level to a reference frame which only references anchor frames; producing another encoded instance of the at least one selected encoded frame; associating the another encoded instance of the at least one selected encoded frame in the separate redundant stream with the corresponding selected frame in the primary stream; and providing the primary stream, the separate redundant stream, and a common clock reference to a decoder. |
