| 主权项 |
1. A video-encoder-device-implemented method for encoding a set of impairments-compensation-filter coefficient values for an encoded video frame, the method comprising:
by the video encoder device, obtaining an unencoded frame of video; by the video encoder device, encoding the unencoded frame of video into an encoded bitstream by differencing a block of unencoded video in the unencoded frame with a first impairments-compensated predication signal block from a motion compensation process to determine a residual block, transforming the residual block into a block of coefficients, quantizing the block of coefficients into a block of quantized coefficients, and encoding the block of quantized coefficients into the bitstream by an entropy coder; by the video encoder device, reverse quantizing and reverse transforming the block of quantized coefficients into a locally decoded block, assembling the locally decoded block into a locally assembled frame and deblock filtering the locally assembled frame into a deblocked frame; obtaining, by the video encoder device, a codebook including a plurality of sets of impairments-compensation-filter coefficient values suitable for configuring an impairments compensation filter of a video decoder to process frames of encoded videos having a variety of characteristics, wherein the impairments compensation filter compensates for at least one coding artifact produced by a motion compensated transform coding process; by the video encoder device, computing a target set of impairments-compensation-filter coefficient values according to the deblocked frame; during encoding of an unencoded frame of video to generate an encoded bitstream, the video encoder device adaptively selecting from the codebook a set of impairments-compensation-filter coefficient values and forming an impairments-compensated reconstructed frame, which impairments-compensated reconstructed frame is used as a reference frame by the motion compensation process to prepare a second impairments-compensated prediction signal block to improve how faithfully the video decoder reproduces at least a portion of the unencoded frame of video when the set of impairments-compensation-filter coefficient values is used to configure the impairments compensation filter to process at least the portion of the encoded frame; encoding the unencoded frame of video; decoding the encoded frame in a local decoder loop by the video encoder device; and computing a target set of impairments-compensation-filter coefficient values according to the decoded frame and the unencoded frame and forming an impairments-compensated reconstructed frame of video, which impairments-compensated reconstructed frame of video is used as a reference frame for the motion compensated transform coding process; and including in the encoded bitstream generated by the video encoder device an indicator enabling the video decoder to select the selected set of impairments-compensation-filter coefficient values from the video decoder's copy of the codebook. |
