Adaptive Deblocking Filters for Content-Aware Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing deblocking filters in video coding technologies apply uniform parameters across blocks, leading to inefficiencies in computational resources and visual quality, particularly in high-resolution, high-bit-depth, and high-dynamic-range videos, as they fail to adapt to content and viewing characteristics.
Innovation Solution
Adaptive deblocking filtering methods that determine the number of borders and filter strength based on content complexity and viewing characteristics, using β and tc threshold parameters adjusted by content and viewing conditions to optimize filtering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If uniform deblocking filter parameters are applied across all blocks, then the filtering process is simple and fast, but visual quality deteriorates and computational resources are wasted on high-resolution content
Solution Approach 1:
The patent applies different deblocking filter parameters to different blocks based on their content characteristics. Blocks are classified into types (e.g., smooth, texture, edge) and assigned appropriate filter strengths and border counts. This local adaptation improves visual quality in complex regions while maintaining simplicity in uniform regions, resolving the contradiction between operational simplicity and visual precision.
Solution Approach 2:
The patent dynamically adjusts filter parameters based on block content and viewing characteristics. The filter strength and number of borders to be filtered are not fixed but adapt according to the detected content complexity and viewing conditions. This dynamic adjustment allows the system to optimize between processing simplicity and visual quality in real-time.
2Manufacturing precision
If strong filtering is applied to remove blockiness, then visual quality improves, but computational complexity and processing time increase
Solution Approach 1:
Instead of applying strong filtering uniformly, the patent applies filtering strength locally based on block characteristics. Blocks with high blockiness and smooth content receive stronger filtering, while texture-rich blocks receive weaker or no filtering. This localized approach improves visual quality where needed without unnecessarily increasing computational complexity across the entire image.
Solution Approach 2:
The patent applies filtering selectively only to the extent necessary. By determining the exact number of borders to filter and the appropriate filter strength based on content analysis, the system avoids excessive filtering in regions where it is not needed, thereby reducing unnecessary computational complexity while maintaining sufficient visual quality improvement.
3Productivity
If adaptive filtering based on content complexity is applied, then visual quality and coding efficiency improve, but device complexity increases
Solution Approach 1:
The patent segments the image into blocks and further classifies them into content types based on local characteristics. This segmentation allows the system to apply different filtering strategies to different segments independently. By breaking down the complex task of adaptive filtering into manageable block-level decisions, the system improves coding efficiency without proportionally increasing overall device complexity.
Solution Approach 2:
The patent changes filter parameters (strength, border count) based on detected content complexity and viewing characteristics. By using parameter tables and lookup mechanisms that map content types to appropriate filter settings, the system achieves adaptive filtering efficiency. This parameter-based adaptation improves coding efficiency while keeping the complexity management through pre-defined parameter relationships rather than complex real-time calculations.
4Manufacturing precision
If filtering is applied to high-resolution high-dynamic-range videos, then visual quality improves, but computational resources are exhausted
Solution Approach 1:
The patent applies filtering locally based on block content characteristics rather than uniformly across the entire high-resolution image. By identifying and filtering only the blocks that actually exhibit blockiness and smooth content, the system improves visual quality in critical regions while conserving computational resources in regions where filtering is not necessary, thus resolving the contradiction between visual quality improvement and computational resource consumption.
Solution Approach 2:
The patent applies filtering only to the extent necessary for high-resolution content, avoiding excessive computational processing in regions where the human visual system would not perceive improvement. By selectively filtering based on content complexity and viewing characteristics, the system achieves sufficient visual quality improvement for high-resolution videos without exhausting computational resources.
Data Source
AI summary
Methods and systems for in-loop filtering may comprise receiving a video comprising at least one frame. The frame may comprise at least one block of pixels. A content complexity of the block of pixels may be determined. A viewing characteristic of the video may be determined. A number of borders to be filtered may be determined based on at least one of the content complexity or the viewing characteristic. A deblocking filter strength may be determined based on at least one of the content complexity or the viewing characteristic. The number of borders of the block of pixels may be filtered according to the deblocking filter strength.