Non-local Adaptive Loop Filter for Video Denoising
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Video compression techniques, such as block-based motion compensation and quantization, introduce compression noise leading to artifacts like blocking, ringing, and blurring in reconstructed pictures, which existing in-loop filters struggle to fully mitigate.
Innovation Solution
A non-local adaptive loop filter (NL-ALF) is employed in video coding systems to denoise reconstructed pictures by selecting the most suitable non-local denoising technology from candidates like non-local means, low-rank, block matching, and three-dimensional discrete cosine transform technologies based on patch characteristics, and applying these technologies in parallel processing regions to form patch groups.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If multiple non-local denoising technologies are combined and adaptively selected, then picture quality is improved, but device complexity increases
Solution Approach 1:
The reconstructed picture is divided into multiple patches, and different non-local denoising technologies are selectively applied to different patch groups based on their characteristics. This segmentation allows the system to manage complexity by processing smaller regions with appropriate algorithms rather than applying a single complex system to the entire picture.
Solution Approach 2:
The filter system dynamically selects from multiple candidate non-local denoising technologies based on patch characteristics such as texture complexity and distortion metrics. This dynamic adaptation allows the system to optimize picture quality for each region while managing overall complexity through intelligent algorithm selection rather than constant use of the most complex methods.
2Manufacturing precision
If non-local adaptive loop filter processes each patch group individually, then picture quality is improved, but processing time increases
Solution Approach 1:
The picture is divided into multiple patch groups that can be processed in parallel. By segmenting the processing task across multiple independent patch groups, the system maintains high picture quality through individualized processing while reducing overall processing time through parallel execution of multiple smaller tasks simultaneously.
3Adaptability or versatility
If multiple candidate non-local denoising technologies are provided, then adaptability is improved, but device complexity increases
Solution Approach 1:
The system provides multiple candidate non-local denoising technologies and dynamically selects the most appropriate one for each patch group based on characteristics such as texture complexity and distortion metrics. This dynamic selection mechanism improves adaptability across different picture regions while managing device complexity through intelligent algorithm choice rather than simultaneously maintaining all algorithms active.
Data Source
AI summary
Aspects of the disclosure provide a method for denoising a reconstructed picture in a video coding system. The method can include providing two or more candidate non-local denoising technologies, selecting a target non-local denoising technology from the two or more candidate non-local denoising technologies for a patch group, and denoising the patch group of the reconstructed picture with the target non-local denoising technology. Besides, two parallel processing methods are provided for forming patch groups with predictor-based searching algorithms.