Extra-Long Deblocking Filters for Large-Block Video Smoothing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing deblocking filter processes in video coding standards like VVC are limited by the fixed length of filters, which do not effectively handle the increased transform sizes, leading to suboptimal video quality and efficiency.
Innovation Solution
Introduce extra-long deblocking filters with extended filter lengths, such as 15, 23, and 31 taps, and adaptive quantization parameters to enhance the filtering process, improving the subjective quality of video sequences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If fixed-length deblocking filters are used in VVC, then device complexity is reduced and processing is simplified, but video quality deteriorates due to inability to effectively handle increased transform sizes
Solution Approach 1:
The patent implements dynamic filter length selection where the deblocking filter adapts its length (7, 15, or 31 taps) based on block size and smoothness criteria. This allows the system to use longer filters for large blocks that benefit from extended smoothing while using shorter filters for small blocks to avoid excessive complexity and processing overhead.
Solution Approach 2:
The patent applies different filter lengths to different regions of the video data based on local characteristics. By evaluating block smoothness and size locally, the system selects appropriate filter lengths for each boundary segment, providing enhanced quality where needed (large smooth blocks) while maintaining efficiency elsewhere (small or textured blocks).
2Manufacturing precision
If longer deblocking filters are applied to all blocks, then video quality improves through better artifact reduction, but processing time and computational load increase
Solution Approach 1:
The patent applies the principle of partial action by selectively using longer filters only when beneficial. The smoothness evaluation criteria determine whether extended filtering (15 or 31 taps) is warranted, avoiding unnecessary computational expenditure on blocks where longer filters would not provide meaningful improvement.
Solution Approach 2:
The patent segments the filtering process into multiple decision stages. First, a quick smoothness check determines if any filtering is needed. Then, based on block size and smoothness metrics, the process segments into appropriate filter length selections (7, 15, or 31 taps), allowing efficient processing paths for different block types.
Data Source
AI summary
A method for deblocking is provided. The method includes checking whether a current block is smooth. The method includes checking whether an adjacent block is smooth. The method includes determining that a criteria is true. The criteria comprises (1) one or more of the current block and the adjacent block being smooth and (2) first and second differences being less than a threshold. The first difference is a difference between a first average value of the current block and a second average value of the adjacent block and the second difference is a difference between a third average value of the current block and a fourth average value of the adjacent block. The method includes, as a result of determining that the criteria is true, applying a long-tap deblocking across the boundary of the current block. A length of the long-tap deblocking is greater than or equal to 15.