| 摘要 |
A digital compression apparatus and method uses a first compression encoding step associated with a first encode circuit to produce a first statistical data set from a first compression encoding of an initial digital video bit stream representing a group of pictures (GOP). The initial bit stream of the GOP is also stored in a first memory, while the first compression encoding is performed. Concurrently with, or following the first compression of the GOP data by the first encoder, a companion processor uses the first statistical data set and optional additional information to compute a set of filter and encoder control settings. The initial digital video bit stream of the GOP data previously stored in the first memory is retrieved from the memory after the first encoding step and is input as a time delayed version of the initial digital video bit stream through a set of filters and subsequently through a second compression encode circuit to produce a second compressed video bit stream of the GOP data. The filters are configured to provide dynamically variable filter characteristics that are responsive to the filter control settings from the companion processor. The filter characteristics can be changed optionally by the companion processor on a block, frame or GOP basis. The second encode circuit also has encoder characteristics responsive to the encoder control settings. The filter and encoder control settings selected by the companion enable the second encoder to provide enhanced compression encoding performance second compression encoding of the GOP data video bit stream relative to the compression of the first encode circuit. In particular, the second compressed bit stream may have improved performance in either or both bit rate and total bits relative to the compression of the first encode circuit.
|
