Constrained Offset Compensation for Error-Resilient Video Decoding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video compression technologies lack error-resilient mechanisms for offset compensation and loop filtering, leading to degradation of image quality due to errors in compressed video bit streams during transmission.

Innovation Solution

Implementing constrained offset compensation and loop filtering by using encoding parameters of target blocks and neighboring blocks, including sample adaptive offset (SAO) compensation and adaptive loop filter (ALF) indicators and parameters, to determine the application of these filters during encoding and decoding.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If offset compensation or loop filter is applied to minimize distortion in reconstructed video, then image quality is improved, but errors in compressed bit streams are propagated temporally or spatially, degrading image quality

Engineering Contradiction:
Improveimage qualityVSAvoiderror propagation resistance
Core Design Contradiction:
Manufacturing precisionVSReliability

Solution Approach 1:

The patent divides the video picture into multiple blocks and applies offset compensation and loop filtering selectively to each block based on its encoding mode. Intra-encoded blocks receive full filtering, while inter-encoded blocks have filtering constrained to prevent error propagation from neighboring blocks. This segmentation approach allows the system to maintain high image quality in reliable regions while preventing error spread in error-prone regions.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies different filtering characteristics to different regions of the video picture based on their encoding properties. The constrained offset compensation and loop filtering are applied locally to intra-encoded blocks, while inter-encoded blocks receive limited or no filtering. This local differentiation ensures that filtering enhances quality where it is safe to do so while avoiding error propagation in regions that should be treated with caution.

Inventive Principle:
Principle #3Local quality

2Reliability

If constrained offset compensation and loop filtering are applied to prevent error propagation, then reliability is improved, but the complexity of encoding and decoding processes increases

Engineering Contradiction:
Improveerror propagation resistanceVSAvoidencoding and decoding complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent implements dynamic filtering control where the filtering operation is enabled or disabled for each block based on its encoding mode and the presence of errors. The system adapts its behavior in real-time during decoding, applying full filtering to safe blocks and constrained or no filtering to potentially erroneous blocks. This dynamic approach balances reliability enhancement with processing complexity by only activating complex filtering operations when necessary.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP3965427A1Method for encoding and decoding images based on constrained offset compensation and loop filter, and apparatus therefor
Publication Date: 2022.03.09 ELECTRONICS & TELECOMM RES INST
  • EP3965427A1 patent drawingFigure 1
  • EP3965427A1 patent drawingFigure 2
  • EP3965427A1 patent drawingFigure 3

AI summary

Provided are a method and apparatus for encoding and decoding images based on constrained offset compensation and a loop filter. The image decoding apparatus: receives, from an encoder, a first indicator indicating whether a sequence, a picture, a frame, a slice, a coding unit (CU), a prediction unit (PU), and/or a transform unit (TU) supports constrained offset compensation; receives, from the encoder, a second indicator indicating whether constrained sample adaptive offset (SAO) compensation or an adaptive loop filter (ALF) is applied; receives a parameter from the encoder; and applies the SAO compensation or the ALF to pixels of a restored image on the basis of the second indicator and the parameter.