Deblocking Filter Boundary Strength for Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in reducing visual artifacts caused by block-based coding, particularly at the edges between video blocks, which affect the quality of decoded video data.
Innovation Solution
A deblocking filter is applied to the edges between video blocks based on specific boundary strength values, depending on whether the blocks are intra-predicted or not, and additional conditions, to reduce the visibility of blocking artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If deblocking filters are applied to all edges between video blocks, then blocking artifacts are reduced, but processing complexity increases
Solution Approach 1:
The patent applies different boundary strength values (0, 1, or 2) to different edges based on local characteristics such as intra-prediction mode and block type. This selective approach ensures filtering is applied only where needed, reducing overall processing complexity while maintaining artifact reduction effectiveness in critical areas.
Solution Approach 2:
The patent changes the parameter of boundary strength (Bs) to control filtering intensity dynamically. By setting Bs=0 to disable filtering, Bs=1 for weak filtering, and Bs=2 for strong filtering, the system adapts the filtering strength to local edge characteristics, optimizing the balance between artifact reduction and processing load.
2Object-affected harmful factors
If boundary strength is increased for all edges, then artifact reduction is improved, but processing time increases
Solution Approach 1:
The patent determines boundary strength locally for each edge based on specific conditions: Bs=2 when either block is intra-predicted, Bs=1 when blocks are inter-predicted with certain conditions, and Bs=0 otherwise. This localized assessment prevents unnecessary processing time on edges that don't require strong filtering.
Solution Approach 2:
The patent applies partial filtering action by setting Bs=0 for edges that don't meet specific criteria, effectively skipping filtering for those edges. This selective partial action reduces processing time while maintaining sufficient artifact reduction for edges that genuinely need it.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A video coder associates a first boundary strength value with an edge in response to determining that a first video block or a second video block is associated with an intra-predicted coding unit (CU), where the edge occurs at a boundary between the first video block and the second video block. The video coder may associate a second or a third boundary strength value with the edge when neither the first video block nor the second video block is associated with an intra-predicted CU. The video coder may apply one or more deblocking filters to samples associated with the edge when the edge is associated with the first boundary strength value or the second boundary strength value. The third boundary strength value indicates that the deblocking filters are turned off for the samples associated with the edge.