Deblocking Filter Parallel Processing for Image Quality
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing technique for deblocking filters in image processing, as proposed in JCTVC-A119, does not effectively enable parallel processing between vertical and horizontal block boundaries within a macro block or between macro blocks, leading to processing delays and increased data rates due to high processing demands.
Innovation Solution
An image processing device and method that includes a decoding section, a determination section, a filtering section, and a control section, which determine whether to apply a deblocking filter to neighboring blocks using pixels of a reconstructed image as reference pixels, allowing for parallel processing of vertical and horizontal block boundaries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If the deblocking filter is applied to prevent image quality degradation due to block distortion, then image quality is improved, but processing time and computational load increase significantly
Solution Approach 1:
The patent segments the deblocking filter application into independent vertical and horizontal boundary processing units. By dividing the filtering operation into separate vertical boundary processing and horizontal boundary processing stages, the system can process different boundaries in parallel without mutual dependency, significantly reducing processing time while maintaining image quality.
Solution Approach 2:
The patent introduces a temporal dimension by processing vertical and horizontal boundaries in separate sequential passes rather than simultaneously. This dimensional separation allows independent processing of vertical boundaries in one pass and horizontal boundaries in another pass, enabling parallelization across different spatial locations while maintaining dependency resolution.
2Reliability
If the deblocking filter processes vertical and horizontal block boundaries sequentially, then processing dependencies are resolved, but processing delay increases
Solution Approach 1:
The patent segments the boundary processing into independent vertical and horizontal phases. By separating the processing of vertical boundaries from horizontal boundaries into distinct phases, the system resolves dependencies between different boundary types while enabling parallel processing within each phase, thus reducing overall processing delay.
Solution Approach 2:
The patent performs preliminary processing of vertical boundaries before horizontal boundaries. By completing the vertical boundary filtering first and storing the results, the system prepares the image data in advance for the subsequent horizontal boundary processing, eliminating the need for sequential processing and reducing delay.
3Manufacturing precision
If the deblocking filter is applied to all block boundaries, then image quality degradation is prevented, but data rate increases due to processing requirements
Solution Approach 1:
The patent segments the filtering operation into vertical and horizontal components that can be processed independently. This segmentation allows the system to apply the deblocking filter to all boundaries while reducing the computational complexity per boundary type, thereby controlling data rate requirements while maintaining comprehensive image quality improvement.
Data Source
AI summary
An image processing device including a decoding section configured to decode an image from an encoded stream, a determination section configured to perform determination processes of determining whether to apply a deblocking filter to neighboring blocks neighboring across a block boundary within an image to be decoded by the decoding section, a filtering section configured to apply a deblocking filter to neighboring blocks to which the determination section has determined to apply a deblocking filter, and a control section configured to allow the determination section to perform the determination processes for a vertical block boundary and a horizontal block boundary using pixels of the neighboring blocks of a reconstruct image as reference pixels.


