Adaptive Boundary Filtering for Intra Prediction in Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding and decoding technologies face challenges in efficiently filtering boundaries for intra prediction, particularly in handling multiple reference layers which can lead to increased computational complexity and reduced efficacy of boundary smoothing.
Innovation Solution
The proposed method employs a gradient-based post-filtering approach and adaptive boundary filtering techniques that utilize multiple reference layers to improve boundary smoothing in intra prediction, specifically for angular modes and diagonal directions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If multiple reference layers are used for boundary filtering in intra prediction, then compression efficiency is improved, but computational complexity increases
Solution Approach 1:
The patent divides the boundary filtering process into distinct segments: gradient calculation for each reference layer, separate filtering applications for different boundary regions, and selective application based on prediction mode. This segmentation allows the system to manage computational complexity by processing different regions independently rather than applying a single complex filter to the entire boundary.
Solution Approach 2:
The patent applies different filtering strategies to different local regions of the boundary based on the prediction mode and reference layer characteristics. For example, angular modes receive gradient-based filtering while other modes may use different approaches. This local quality principle ensures that computational resources are concentrated where they provide the most benefit, improving compression efficiency without uniformly increasing complexity across all regions.
2Manufacturing precision
If gradient-based post-filtering is applied to smooth boundaries in intra prediction, then boundary smoothing efficacy is improved, but computational complexity increases
Solution Approach 1:
The patent calculates gradients from reference samples before applying the filtering operation. By pre-computing the gradient values based on reference layer data, the system prepares the necessary information in advance, which then enables more efficient application of the gradient-based filter during the actual prediction process. This preliminary action reduces the computational burden during the main filtering operation.
Solution Approach 2:
The patent applies gradient-based post-filtering selectively rather than uniformly to all prediction modes and all boundary regions. The filtering is applied partially based on the specific prediction mode (e.g., angular modes) and boundary characteristics, which reduces the overall computational complexity while maintaining boundary smoothing efficacy where it is most needed.
Data Source
Figure 1
Figure 2~3
Figure 4A~4B
AI summary
After initial intra prediction based on projection along an intra prediction direction, there may exist discontinuities between the predicted block and neighboring blocks. To smooth the block boundary, boundary filtering may be used with intra prediction. In particular, different methods may be used for boundary filtering. The particular filtering method to be used may be selected adaptively based on the intra prediction mode. When gradient-based filtering is selected, the gradient at a reference sample, that is at the same row or column as the current sample, in a reference layer is calculated as the difference between the reference sample and the predicted value for the reference sample using the intra prediction mode. Additional gradient at another reference sample from another reference layer may also be used. Then the initially predicted value for the sample may be adjusted based on the gradient(s) and a decay function.