Adaptive In-Loop Filtering for Video Distortion Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current video encoding/decoding techniques face challenges in minimizing distortion between original and reconstructed pictures, particularly with high-resolution videos, as existing filtering methods like deblocking and sample adaptive offset (SAO) are limited in reducing artifacts and optimizing rate-distortion.

Innovation Solution

The implementation of an adaptive in-loop filtering method that uses adaptive parameter sets, including chroma and luma adaptive in-loop filters, to filter coding tree blocks based on specific filtering criteria, such as clipping flags and indices, to reduce distortion and improve video encoding/decoding efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Object-affected harmful factors

If deblocking filtering is performed on block boundaries, then blocking artifacts are reduced, but distortion between original and reconstructed pictures cannot be minimized

Engineering Contradiction:
Improveblocking artifactsVSAvoiddistortion between original and reconstructed pictures
Core Design Contradiction:
Object-affected harmful factorsVSManufacturing precision

Solution Approach 1:

The patent applies different filtering strategies to different regions (luma and chroma components) based on their specific characteristics. Luma components receive one type of filtering while chroma components receive another, allowing each region to be optimized for its specific artifact reduction needs while minimizing overall distortion.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent introduces adaptive parameters including clipping flags and indices that dynamically adjust filtering strength and characteristics based on local image characteristics. This allows the filtering process to adapt to different content types and regions, minimizing distortion while reducing artifacts.

Inventive Principle:
Principle #35Parameter changes

2Manufacturing precision

If sample adaptive offset is applied to reduce ringing artifacts, then distortion is reduced to some degree, but there is a limit when the difference between original and reconstructed picture is large

Engineering Contradiction:
Improvedistortion between original and reconstructed picturesVSAvoideffectiveness under large differences
Core Design Contradiction:
Manufacturing precisionVSAdaptability or versatility

Solution Approach 1:

The patent implements dynamic adaptive in-loop filtering that adjusts filtering parameters based on the actual difference between original and reconstructed pictures. When differences are large, the system automatically strengthens filtering; when differences are small, it reduces filtering intensity, maintaining adaptability across various quality scenarios.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The filtering process incorporates feedback mechanisms that continuously monitor the reconstructed picture quality and adjust filtering parameters accordingly. This feedback loop enables the system to optimize distortion reduction in real-time based on actual image content and reconstruction accuracy.

Inventive Principle:
Principle #23Feedback

3Manufacturing precision

If high-resolution, high-quality video is transmitted or stored, then video quality is improved, but transmission or storage cost increases

Engineering Contradiction:
Improvevideo qualityVSAvoidtransmission or storage cost
Core Design Contradiction:
Manufacturing precisionVSQuantity of substance

Solution Approach 1:

The patent optimizes the balance between video quality and data量 by dynamically adjusting filtering parameters and compression settings. This allows high-quality video transmission at reduced data rates, lowering storage and transmission costs while maintaining acceptable quality levels.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentUS20220248006A1Adaptive in-loop filtering method and apparatus
Publication Date: 2022.08.04 ELECTRONICS & TELECOMM RES INST
  • US20220248006A1 patent drawing
  • US20220248006A1 patent drawing
  • US20220248006A1 patent drawing

AI summary

Disclosed herein is a video decoding method including acquiring adaptation parameter sets including an adaptive in-loop filter (ALF) set including a plurality of ALFs, determining an adaptation parameter set applied to a current picture or slice and including an ALF set applied to the current picture or slice, from among the adaptive parameter sets, determining an adaptation parameter set applied to a current coding tree block (CTB) and including an ALF set applied to the current CTB included in the current picture or slice, from the adaptation parameter set applied to the current picture or slice, and filtering the current CTB based on the ALF set of the determined adaptation parameter set applied to the current CTB, wherein the acquired adaptation parameter sets include chroma ALF number information, and wherein the ALF set includes chroma ALFs, the number of which is indicated by the chroma ALF number information.