Adaptive Deblocking Filter Using Local Activity Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding techniques, such as HEVC, often balance computational complexity and coding efficiency by oversimplifying deblocking schemes, which can result in poorer coding efficiency and visible artifacts due to inadequate consideration of local activities between blocks.
Innovation Solution
The proposed method determines the activity in neighboring blocks along boundaries, calculates an average and difference in activity, and uses this classification to select and apply appropriate filters, enhancing deblocking by considering more factors and improving filter selection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If deblocking schemes are oversimplified to balance computational complexity, then computational complexity is reduced, but coding efficiency deteriorates and visible artifacts increase
Solution Approach 1:
The patent applies different filtering strengths (strong, normal, or zero) to different boundary segments based on local activity characteristics. By classifying each boundary segment independently using activity metrics from neighboring blocks, the system adapts the filtering intensity to local requirements, avoiding uniform oversimplification while maintaining computational efficiency through localized decision-making.
2Device complexity
If deblocking schemes are oversimplified to balance computational complexity, then computational complexity is reduced, but visible artifacts increase
Solution Approach 1:
The patent dynamically changes filtering parameters (filtering strength) based on computed activity metrics. By calculating activity values from neighboring blocks and using them to select appropriate filtering levels, the system adapts to local image characteristics, reducing visible artifacts in regions where they occur while avoiding unnecessary processing in regions where they don't exist.
3Manufacturing precision
If activity classification is performed to improve filter selection, then coding efficiency is improved, but device complexity increases
Solution Approach 1:
The patent divides the boundary into multiple segments and performs independent activity classification on each segment using neighboring block information. This segmentation approach allows parallel processing of different boundary regions, improving coding efficiency through localized optimization while managing device complexity by breaking down the overall task into smaller, independent units that can be processed efficiently.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Deblocking is a step in video coding for removing distortions that may result from dividing a video frame into blocks, and encoding the video frame based on the blocks. Techniques described herein can include determining the activity in neighboring blocks along the boundary of the blocks, where the activity measures smoothness or complexity of pixels in the boundary area. An average of the activity can then be determined, as well a difference in the activity between the left block and the right block. The average activity and the difference in activity can then be used to determine a classification for the boundary area. The classification can further be used to select a filter to apply to the pixels in the boundary area. Once the filter have been applied, the blocks can be added to a reconstructed video frame.