Position-Dependent Intra Prediction for Low-Complexity Artifact Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding systems face challenges in achieving high compression efficiency while minimizing computational complexity, particularly in intra prediction processes that result in blocking artifacts due to discontinuities at block boundaries.
Innovation Solution
A post-processing stage for position-dependent intra prediction combination (PDPC) that modifies predicted sample values based on a weighted average of left or top reference samples, determined by the intra prediction angle, to smooth out these discontinuities, using simplified computational methods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-generated harmful factors
If traditional intra prediction methods are used, then compression efficiency is achieved, but blocking artifacts occur due to discontinuities at block boundaries
Solution Approach 1:
The patent applies parameter changes by modifying the predicted sample values through a post-processing stage that adjusts prediction parameters based on position and reference samples. The predicted value is changed from a simple intra prediction result to a combined value that incorporates weighted contributions from both the original prediction and reference samples, thereby reducing blocking artifacts while maintaining prediction accuracy.
2Manufacturing precision
If complex post-processing methods are applied to reduce blocking artifacts, then prediction accuracy improves, but computational complexity increases
Solution Approach 1:
The patent applies local quality by position-dependent processing where different weights are assigned to reference samples based on their spatial position relative to the current block. The processing is localized to the block boundary regions where blocking artifacts are most prominent, applying PDPC only where needed rather than uniformly across the entire image, thus reducing overall computational complexity while maintaining prediction accuracy.
Solution Approach 2:
The patent applies partial action by implementing a post-processing stage that selectively processes only the predicted sample values at block boundaries where discontinuities occur. Rather than reprocessing the entire prediction, the method applies a targeted correction using reference samples only where blocking artifacts are generated, reducing computational complexity compared to full reprocessing.
3Productivity
If position-dependent intra prediction combination is applied, then computing efficiency is improved, but compression performance may be affected
Solution Approach 1:
The patent applies feedback by using reference samples from already-processed neighboring blocks to improve the prediction of current blocks. The reference samples provide feedback information about the actual pixel values at boundaries, which is then used to adjust the intra prediction values. This feedback mechanism ensures that compression performance is maintained or improved while enabling computing efficiency through the use of previously decoded information.
Data Source
AI summary
Described are video coding systems comprising a post-processing stage using position dependent intra prediction combination where a predicted sample is modified based on a weighting between a value of a left or top reference sample and the obtained predicted value for the sample, wherein the left or top reference sample is determined based on the intra prediction angle. This provides better computing efficiency while keeping identical compression performance. Encoding method, decoding method, encoding apparatus and decoding apparatus based on this post-processing stage are proposed.