Intra Prediction Refinement Using Offset Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding standards, such as AV1, face challenges in accurately predicting intra-predicted samples due to discontinuities between neighboring samples, which affect prediction accuracy and coding efficiency.
Innovation Solution
The proposed solution involves refining intra-predicted samples by generating an offset value based on neighboring samples and applying a filter to these samples, specifically using sub-block level refinement and filtering techniques to improve prediction accuracy, particularly for boundary samples in modes like DC, smooth, horizontal, and vertical predictions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If intra prediction is used to compress video data, then coding efficiency is improved, but discontinuities between neighboring samples reduce prediction accuracy
Solution Approach 1:
The patent applies different filtering operations to different regions of the prediction block. Boundary samples that exhibit discontinuities are subjected to specific filtering treatments, while interior samples use different processing. This localized approach addresses prediction accuracy issues at boundaries without unnecessarily processing all samples, thus maintaining coding efficiency while improving local prediction accuracy.
Solution Approach 2:
The patent modifies prediction parameters by introducing offset values to corrected samples. These offset values adjust the prediction parameters locally where discontinuities occur, allowing the system to maintain the overall prediction structure while making targeted improvements to problematic areas, thereby balancing coding efficiency and prediction accuracy.
2Measurement precision
If filtering is applied to reduce discontinuities, then prediction accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent divides the prediction block into boundary samples and interior samples, applying filtering operations only where necessary. By segmenting the processing scope to only include areas with discontinuities, the system reduces overall computational complexity while still improving prediction accuracy in the problematic regions.
Solution Approach 2:
The patent applies filtering operations partially - only to boundary samples where discontinuities are detected, rather than to all samples in the prediction block. This partial action approach provides sufficient correction to improve prediction accuracy without the excessive computational cost of processing entire blocks, thus resolving the contradiction between accuracy improvement and complexity reduction.
Data Source
AI summary
The present disclosure provides methods and apparatuses for applying intra prediction refinement to intra predicted samples. An exemplary method includes: determining a filter based on neighboring samples of intra predicted samples of a picture; generating an offset value based on the neighboring samples; refining the intra predicted samples by adding the offset value; and applying the filter to the intra predicted samples.