Hybrid video coding

Patent metadata

Field	Value
Publication number	US-11070839-B2
Application number	US-201213463837-A
Country	US
Kind code	B2
Filing date	May 4, 2012
Priority date	Nov 6, 2009
Publication date	Jul 20, 2021
Grant date	Jul 20, 2021

Abstract

A further coding efficiency increase is achieved by, in hybrid video coding, additionally predicting the residual signal of a current frame by motion-compensated prediction using a reference residual signal of a previous frame. In other words, in order to further reduce the energy of the final residual signal, i.e. the one finally transmitted, and thus increase the coding efficiency, it is proposed to additionally predict the residual signal by motion-compensated prediction using the reconstructed residual signals of previously coded frames.

First claim

The invention claimed is:

1. Hybrid video decoder comprising a computer programmed to, or a microprocessor configured to:
predict a reference frame of a video by intra prediction or motion-compensated prediction to obtain a prediction signal of the reference frame;
predict a residual signal of the reference frame, which relates to a prediction error of the prediction signal of the reference frame, by motion-compensated prediction from a reference residual signal of a further reference frame of the video to obtain a residual prediction signal of the reference frame;
entropy decode a final residual signal of the reference frame;
reconstruct the reference frame by summing
the prediction signal of the reference frame;
the residual prediction signal of the reference frame; and
the final residual signal of the reference frame,

predict a predetermined frame of the video by intra prediction or motion-compensated prediction to obtain a prediction signal of the predetermined frame;
predict a residual signal of the predetermined frame, which relates to a prediction error of the prediction signal of the predetermined frame, by motion-compensated prediction from a reference residual signal of the reference frame to obtain a residual prediction signal of the predetermined frame;
entropy decode a final residual signal of the predetermined frame; and
reconstruct the predetermined frame by summing
the prediction signal of the predetermined frame;
the residual prediction signal of the predetermined frame; and
the final residual signal of the predetermined frame.

2. Hybrid video decoding method comprising:
predicting a reference frame of a video by intra prediction or motion-compensated prediction to obtain a prediction signal of the reference frame;
predicting a residual signal of the reference frame, which relates to a prediction error of the prediction signal of the reference frame, by motion-compensated prediction from a reference residual signal of a further reference frame of the video to obtain a residual prediction signal of the reference frame;
entropy decoding a final residual signal of the reference frame, and
reconstructing the reference frame by summing
the prediction signal of the reference frame;
the residual prediction signal of the reference frame; and
the final residual signal of the reference frame;

predicting a predetermined frame of the video by intra prediction or motion-compensated prediction to obtain a prediction signal of the predetermined frame;
predicting a residual signal of the predetermined frame, which relates to a prediction error of the prediction signal of the predetermined frame, by motion-compensated prediction from a reference residual signal of the reference frame to obtain a residual prediction signal of the predetermined frame;
entropy decoding a final residual signal of the predetermined frame, and
reconstructing the predetermined frame by summing
the prediction signal of the predetermined frame;
the residual prediction signal of the predetermined frame; and
the final residual signal of the predetermined frame.

3. Hybrid video decoding method according to claim 2 , further comprising building the reference residual signal of the reference frame by a sum of the final residual signal of the reference frame and the residual prediction signal of the reference frame.

4. Hybrid video decoding method according to claim 2 , further comprising:
selecting, for the predetermined frame, to
build the reference residual signal of the reference frame by the final residual signal of the reference frame, or
build the reference residual signal of the reference frame by a sum of the final residual signal of the reference frame and the residual prediction signal of the reference frame; and

selecting, for the reference frame, to
build the reference residual signal of the further reference frame by a final residual signal of the further reference frame, or
build the reference residual signal of the further reference frame by a sum of the final residual signal of the further reference frame and a residual prediction signal of the further reference frame.

5. Hybrid video decoding method according to claim 2 , further comprising:
building a first candidate reference residual signal of the reference frame by the final residual signal of the reference frame and inserting the first candidate reference residual signal in a decoded picture buffer,
building a second candidate reference residual signal of the reference frame by a sum of the final residual signal of the reference frame and the residual prediction signal of the reference frame, and inserting the second candidate reference residual signal into the decoded picture buffer, and
using the first or second candidate reference residual signal as the reference residual signal of the reference frame depending on a signalization within a bitstream.

6. Hybrid video decoding method according to claim 2 , further comprising:
entropy decoding information on residual prediction motion parameters for the predetermined frame, and
using the residual prediction motion parameters in predicting the residual signal of the predetermined frame.

7. Hybrid video decoding method according to claim 6 , further comprising:
entropy decoding information on video prediction motion parameters for the predetermined frame, and
predicting the predetermined frame by motion-compensated prediction using the video prediction motion parameters to obtain the prediction signal of the predetermined frame, a prediction error of which the residual signal of the predetermined frame relates to.

8. Hybrid video decoding method according to claim 2 , further comprising:
decoding one or more syntax elements for the predetermined frame; and
applying the prediction of the residual signal of the predetermined frame to a predetermined set of first sets of samples of the predetermined frame, the predetermined set being defined by the one or more syntax elements.

9. Hybrid video decoding method according to claim 8 , further comprising:
applying the prediction of the predetermined frame resulting in the prediction signal of the predetermined frame to second sets of samples of the predetermined frame,
decoding one or more syntax elements for each of the second sets of samples, and
using the one or more syntax elements for each of the second sets of samples to identify the predetermined set of the first sets of samples out of the second sets of samples.

10. Hybrid video decoding method according to claim 2 , further comprising:
decoding one or more syntax elements for the predetermined frame and applying the prediction of the residual signal of the predetermined frame to a predetermined set of first sets of samples of the predetermined frame, the predetermined set being defined by the one or more syntax elements, and
applying an intra prediction of the predetermined frame partially forming the prediction signal of the predetermined frame to a predetermined first set of second sets of samples of the predetermined frame, and a motion-compensated prediction of the predetermined frame partially forming the prediction signal of the predetermined frame, to a predetermined second set of the seconds sets of samples.

11. Hybrid video decoding method according to claim 2 , further comprising extracting a residual reference frame index indexing the reference frame, from a bitstream.

12. Hybrid video decoding method according to claim 2 , further comprising:
inferring a residual reference frame index indexing the reference frame; wherein
the reference frame is the one based on which the prediction signal of the

Assignees

• Bross Benjamin
• Wiegand Thomas
• Schwarz Heiko
• Fraunhofer Ges Forschung

Inventors

• Bross Benjamin
• Wiegand Thomas
• Schwarz Heiko

Classifications

• H04N19/523 Primary

  with sub-pixel accuracy · CPC title

• H04N19/19

  using optimisation based on Lagrange multipliers · CPC title

• H04N19/109

  among a plurality of temporal predictive coding modes · CPC title

• H04N19/174

  the region being a slice, e.g. a line of blocks or a group of blocks · CPC title

• H04N19/196

  being specially adapted for the computation of encoding parameters, e.g. by averaging previously computed encoding parameters (processing of motion vectors H04N19/513) · CPC title