Parallel 4:4:4 Video Encoder Common Mode Selection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video encoding methods face inefficiencies in parallel processing of 4:4:4 video signals, particularly in selecting common block partitions and spatial prediction modes, which can lead to increased decoder complexity and reduced compression performance compared to independent channel methods.

Innovation Solution

A video encoder that selects a common block partition and spatial prediction mode by concurrently evaluating all color components in parallel, using a hybrid serial-parallel approach to optimize spatial prediction for intra-coded pictures, thereby enhancing encoding efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If a common block partition and common spatial prediction mode are used for all color channels, then decoder complexity is reduced and compression performance is improved, but encoding speed is reduced due to serial evaluation of each channel

Engineering Contradiction:
Improvedecoder complexityVSAvoidencoding speed
Core Design Contradiction:
Device complexityVSProductivity

Solution Approach 1:

The patent segments the evaluation process into two independent parallel streams: one for evaluating all spatial prediction modes across all color channels simultaneously, and another for selecting the common block partition. This segmentation allows the encoder to evaluate modes in parallel without having to serially process each channel, thereby maintaining fast encoding speed while still selecting optimal common modes that reduce decoder complexity and improve compression.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary parallel evaluation of all spatial prediction modes for all color channels before final selection. By pre-evaluating all modes in parallel and storing their cost functions, the system prepares all necessary information in advance, allowing the final common mode selection to be made efficiently without serial processing, thus maintaining high encoding speed.

Inventive Principle:
Principle #10Preliminary action

2Productivity

If independent block partition and spatial predictors are selected for each color channel, then encoding speed is increased through parallel processing, but decoder complexity increases and compression performance deteriorates

Engineering Contradiction:
Improveencoding speedVSAvoiddecoder complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent merges the mode selection process across all color channels by selecting a single common spatial prediction mode that is applied to all channels (Y, Cb, Cr). This merging reduces decoder complexity because the decoder only needs to handle one mode per block rather than three separate modes, while the encoding process maintains speed by evaluating all channels in parallel to determine the common mode.

Inventive Principle:
Principle #5Merging (Combining)

3Productivity

If independent block partition and spatial predictors are selected for each color channel, then encoding speed is increased through parallel processing, but the number of bits required to encode spatial prediction information increases

Engineering Contradiction:
Improveencoding speedVSAvoidbits to encode spatial prediction information
Core Design Contradiction:
ProductivityVSLoss of information

Solution Approach 1:

The patent merges the spatial prediction mode representation across all color channels by encoding a single common mode index that applies to all channels. Instead of encoding separate mode indices for Y, Cb, and Cr channels, the system encodes one common mode that is reused for all channels, significantly reducing the number of bits required to represent spatial prediction information while maintaining parallel evaluation for speed.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentEP1972149B1Methods and apparatus for parallel implementations of 4:4:4 coding
Publication Date: 2017.12.13 THOMSON LICENSING SA
  • EP1972149B1 patent drawingFigure 1
  • EP1972149B1 patent drawingFigure 2
  • EP1972149B1 patent drawingFigure 3

AI summary

There are provided methods and apparatus for parallel implementations of 4:4:4 coding. A video encoder for encoding video signal data for an image block includes an encoder (100) for encoding all color components of the image block by selecting a common block partition and a common spatial prediction mode. The common block partition and the common spatial prediction mode are selected by concurrently evaluating all of the color components in parallel.