Video coding methods and apparatus
AssigneeSony Computer Entertainment Inc.
InventorJason N. Wang
Life of this patent
Estimating the term…
Abstract
In methods for encoding and decoding digital pictures certain prediction parameters may be omitted from the output on the encoder side or the input on the decoder side. An encoder can identify prediction parameter values that can be omitted by determining whether a decoder emulator can reproduce the prediction parameter values from other information, such as predicted pixel and prediction error values of one or more previously decoded sections from the same or a different picture.
Claims35
A method for encoding a current section of a digital picture, comprising:
- a) generating a set of transform coefficients from a set of prediction errors for a current section of a picture using an encoder device;
- b) determining a first set of one or more prediction parameter values for the current section of the digital picture using the encoder device;
- c) determining a second set of one or more prediction parameter values for the current section of the digital picture with a decoder emulator using predicted pixel and prediction error values of one or more previously decoded sections, wherein the determining the second set includes, with the decoder emulator, applying an inverse transform to the set of transform coefficients to re-generate the set of prediction errors for the current section;
- d) comparing the first set of prediction parameters with the second set of prediction parameters to identify one or more prediction parameter values common to the first set and the second set;
- z) identifying a third set of prediction parameter values, the third set of prediction parameter values being selected from the first set but omitting the prediction parameter values common to the first set and the second set;
- and
- e) transmitting, into a compressed stream for a decoder, the set of transform coefficients and the third set of prediction parameter values with the prediction parameter values common to the first set and the second set omitted from the compressed stream.
An apparatus for encoding a current section of a digital picture, comprising:
- a processor;
- a memory;
- and
- computer coded instructions embodied in the memory and executable by the processor, wherein the computer coded instructions are configured to implement a method for encoding/decoding a current section of a digital picture using three dimensional motion compensation, the method comprising:
- a) generating a set of transform coefficients from a set of prediction errors for a current section of a picture using an encoder;
- b) determining a first set of one or more prediction parameter values for the current section of the digital picture using the encoder;
- c) determining a second set of one or more prediction parameter values for the current section of the digital picture with a decoder emulator using predicted pixel and prediction errors of one or more previously decoded sections, wherein the determining the second set includes, with the decoder emulator, applying an inverse transform to the set of transform coefficients to re-generate the set of prediction errors for the current section;
- d) comparing the first set of prediction parameters with the second set of prediction parameters to identify one or more prediction parameter values common to the first set and the second set;
- z) identifying a third set of prediction parameter values, the third set of prediction parameter values being selected from the first set but omitting the prediction parameter values common to the first set and the second set;
- and
- e) transmitting, into a compressed stream for a decoder, the set of transform coefficients and the third set of prediction parameter values with the prediction parameter values common to the first set and the second set omitted from the compressed stream.
A computer program product comprising:
- a non-transitory computer-readable storage medium having computer readable program code embodied in said medium for, the program code including computer-executable instructions that, when executed implement a method for encoding a current section of a digital picture, said method comprising:
- a) generating a set of transform coefficients from a set of prediction errors for a current section of a picture using an encoder;
- b) determining a first set of one or more prediction parameter values for the current section of the digital picture using the encoder;
- c) determining a second set of one or more prediction parameter values for the current section of the digital picture with a decoder emulator using predicted pixels and prediction errors of one or more previously decoded sections, wherein the determining the second set includes, with the decoder emulator, applying an inverse transform to the set of transform coefficients to re-generate the set of prediction errors for the current section;
- d) comparing the first set of prediction parameters with the second set of prediction parameters to identify one or more prediction parameter values common to the first set and the second set;
- z) identifying a third set of prediction parameter values, the third set of prediction parameter values being selected from the first set but omitting the prediction parameter values common to the first set and the second set;
- and
- e) transmitting, into a compressed stream for a decoder, the set of transform coefficients and the third set of prediction parameter values with the prediction parameter values common to the first set and the second set omitted from the compressed stream.
A picture decoding method for decoding one or more data packets containing data representing one or more coded sections of one or more digital pictures, wherein the one or more data packets include a set of transform coefficients and a partial set of motion vectors, the method comprising:
- a) applying an inverse transform to the transform coefficients with a decoder device to generate a set of prediction error values;
- b) estimating one or more estimated motion vectors with the decoder device from a set of predicted pixel and prediction error values of one or more previously decoded sections of the one or more digital pictures;
- and
- c) replacing one or more of the estimated motion vectors with the decoder device with one or more corresponding motion vectors from the partial set of motion vectors to produce an output set of motion vectors;
- and
- d) generating one or more decoded sections with the decoder device using the set of prediction error values and the output set of motion vectors.
An apparatus for decoding a current section of a digital picture using three dimensional motion compensation, comprising:
- a processor;
- a memory;
- and
- computer coded instructions embodied in the memory and executable by the processor, wherein the computer coded instructions are configured to implement a method for decoding one or more data packets containing data representing one or more coded sections of one or more digital pictures, wherein the one or more data packets include a set of transform coefficients and a partial set of motion vectors, the method comprising:
- a) applying an inverse transform to the transform coefficients with a decoder to generate a set of prediction error values;
- b) estimating one or more estimated motion vectors with the decoder from a set of predicted pixel and prediction error values of one or more previously decoded sections of the one or more digital pictures;
- and
- c) replacing one or more of the estimated motion vectors with the decoder with one or more corresponding motion vectors from the partial set of motion vectors to produce an output set of motion vectors;
- and
- d) generating one or more decoded sections with the decoder using the set of prediction error values and the output set of motion vectors.
A computer program product comprising:
- a non-transitory computer-readable storage medium having computer readable program code embodied in said medium, the program code including computer-executable instructions that, when executed, implement a method for decoding one or more data packets containing data representing one or more coded sections of one or more digital pictures, wherein the one or more data packets include a set of transform coefficients and a partial set of motion vectors, the method comprising:
- a) applying an inverse transform to the transform coefficients with a decoder to generate a set of prediction error values;
- b) estimating one or more estimated motion vectors with the decoder from a set of predicted pixel and prediction error values of one or more previously decoded sections of the one or more digital pictures;
- and
- c) replacing one or more of the estimated motion vectors with the decoder with one or more corresponding motion vectors from the partial set of motion vectors to produce an output set of motion vectors;
- and
- d) generating one or more decoded sections with the decoder using the set of prediction error values and the output set of motion vectors.
Dependencies are inferred from the claim text.