Back to results
US 2022/0191497 A1Published application
Decomposition of residual data during signal encoding, decoding and reconstruction in a tiered hierarchy
AssigneeV-Nova International Limited
InventorsLuca Rossato, Guido Meardi
Life of this patent
Estimating the term…
Abstract
Computer processor hardware receives a first set of adjustment values. The first set of adjustment values specify adjustments to be made to a predicted rendition of a signal generated at a first level of quality to reconstruct a rendition of the signal at the first level of quality. The computer processor hardware processes the first set of adjustment values and derives a second set of adjustment values based on the first set of adjustment values and a rendition of the signal at a second level of quality. The second level of quality is lower than the first level of quality.
Claims15
A decoder comprising:
- a calculator of a prediction for a first level of quality to receive a first decoded image at a bottom level of quality and to produce a first predicted image at a first level of quality with operations as specified by a first set of corresponding parameters for operations;
- a first residual decoder to receive reconstruction data and produce first residual data, the first residual data being of the same spatial resolution as the first predicted image;
- a first summation component to receive and combine, on an element-by-element basis, the first predicted image and the first residual data to produce a second decoded image;
- a calculator of a prediction for a second level of quality to receive the second decoded image at the first level of quality and to produce a first predicted image at a second level of quality with operations as specified by a second set of corresponding parameters for operations;
- a second residual decoder to receive reconstruction data and produce second residual data, the second residual data being of the same spatial resolution as the second predicted image;
- and
- a second summation component to receive and combine, on an element-by-element basis, the second predicted image and the second residual data to produce a second decoded image;
- wherein the decoder is a video decoder for decoding a video signal,
- wherein said decoded images comprise planes of pixels corresponding to frames of the video signal,
- wherein the levels of quality correspond to different spatial resolutions,
- wherein the residual data for a given level of quality comprises a difference between an original plane of pixels for the video signal at the given level of quality and a predicted plane of pixels for the video signal at the given level of quality.
8 more elements
An encoder comprising:
- a first downsampler to receive an input image at a first resolution and to downsample the input image to a second resolution that is lower than the first resolution;
- a first residual generator to compute a difference between the input image at the first resolution and a predicted image at the first resolution to generate first residual data;
- a first generator of reconstruction data to receive the first residual data and to output first reconstruction data;
- a first simulated decoder to receive the first reconstruction data at the second resolution and to decode the reconstruction data to output the predicted image at the first resolution;
- a second residual generator to compute a difference between the input image at the second resolution and a predicted image at the second resolution to generate second residual data, the predicted image at the second resolution being generated by a second simulated decoder at a lowest resolution, the predicted image at the second resolution being derived from a simulated decoding of an encoding of the input image at the lowest resolution;
- a second generator of reconstruction data to receive the second residual data and to output second reconstruction data;
- wherein the encoder encodes the first and second reconstruction data to generate a multiscale data stream for communication to a decoder for decoding, the multiscale data stream encoding a plurality of input images to generate an encoding of multiple frames of a video signal, the decoder generating a reconstruction of the video signal by decoding the first and second reconstruction data to reconstruct first and second residual data for combination with a decoded image at the lowest resolution.
4 more elements
Dependencies are inferred from the claim text.