Reconstructed Block Filtering for Accurate Video Intra-Prediction

Solution Overview

Problem

Existing video encoding and decoding methods face challenges in achieving accurate prediction, particularly in intra prediction, leading to prediction errors and reduced video compression efficiency due to difficulties in predicting pixels away from reference pixels and distortion between original and restored videos.

Innovation Solution

An intra prediction method involving in-loop filtering is employed, which includes generating a residual block through inverse quantization and transformation, forming a prediction block, performing in-loop filtering on the summed residual and prediction blocks, and storing the filtered block for subsequent encoding, while adjusting filtering based on prediction direction and block characteristics.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If existing intra prediction methods are used, then encoding complexity is reduced, but prediction accuracy deteriorates for pixels away from reference pixels

Engineering Contradiction:

Improveprediction accuracyVSAvoidencoding complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies in-loop filtering to the reconstructed block before it is used as a reference for subsequent predictions. This preliminary filtering action removes distortion and improves the quality of reference pixels, thereby enhancing prediction accuracy for future blocks without increasing the complexity of the prediction algorithm itself.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The in-loop filter acts as an intermediary between the reconstruction process and the prediction process. It processes the reconstructed block to eliminate artifacts and distortion, providing cleaner reference data to the prediction unit, which then achieves better prediction accuracy without modifying its own complexity.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Reliability

If existing intra prediction methods are used, then encoding process is simpler, but distortion between original and restored video increases

Engineering Contradiction:

Improvevideo qualityVSAvoidencoding process complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent applies in-loop filtering to the reconstructed block before it is used as a reference for subsequent predictions. This preliminary filtering action removes distortion and improves the quality of reference pixels, thereby enhancing prediction accuracy for future blocks without increasing the complexity of the prediction algorithm itself.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The in-loop filter creates a feedback mechanism where the reconstructed block is filtered and then fed back into the prediction process. This feedback loop continuously improves the quality of reference data, reducing distortion accumulation across multiple prediction operations while maintaining a manageable encoding process.

Inventive Principle:

Principle #23Feedback

3Loss of information

If existing prediction methods are used, then data transmission amount is higher, but prediction error is reduced

Engineering Contradiction:

Improvedata transmission amountVSAvoidprediction error

Core Design Contradiction:

Loss of informationVSMeasurement precision

Solution Approach 1:

The patent applies in-loop filtering to the reconstructed block before it is used as a reference for subsequent predictions. This preliminary filtering action removes distortion and improves the quality of reference pixels, thereby enhancing prediction accuracy for future blocks without increasing the complexity of the prediction algorithm itself.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS20250287045A1Method and apparatus for encoding/decoding images using a prediction method adopting in-loop filtering

Publication Date: 2025.09.11 DOLBY LABORATORIES LICENSING CORP

AI summary

Disclosed is a prediction method adopting in-loop filtering. According to the present invention, a prediction method for encoding and decoding video comprises the following steps: generating a residual block of the current block through an inverse quantization and inverse transform; generating a prediction block of the current block through an intra-prediction; performing in-loop filtering on the current block in which the residual block and the prediction block are combined; and storing the current block, on which the in-loop filtering is performed, in a frame buffer for an intra-prediction of the next block to be encoded. As described above, prediction is performed using an in-loop filter during processes for encoding and decoding video, thereby improving the accuracy of prediction and reducing errors in prediction, thus improving the efficiency of video compression and reducing the amount of data to be transmitted.