In-Loop Filtering Across Virtual Boundaries in Image Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing demand for high-resolution, high-quality image/video data, particularly in immersive media formats like VR and AR, leads to higher transmission and storage costs due to the increased amount of information, necessitating a more efficient compression technology.
Innovation Solution
Implementing in-loop filtering across virtual boundaries using a sequence parameter set (SPS) flag to enable efficient deblocking and adaptive loop filtering (ALF), enhancing image coding efficiency and subjective/objective visual quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If in-loop filtering is performed across virtual boundaries, then visual quality and compression efficiency are improved, but hardware resource consumption increases
Solution Approach 1:
The image is divided into virtual boundaries (tiles, slices, or bricks) that segment the filtering process into manageable regions. This segmentation allows the filtering to be performed in parallel across different regions, improving visual quality while enabling efficient hardware utilization through distributed processing.
Solution Approach 2:
The filtering process dynamically adapts to virtual boundaries by adjusting filter application based on boundary characteristics. The system determines whether to apply filtering across boundaries based on content complexity and boundary types, optimizing the balance between visual quality improvement and hardware resource consumption.
2Manufacturing precision
If high-resolution, high-quality image/video data is transmitted, then image quality is improved, but transmission and storage costs increase
Solution Approach 1:
In-loop filtering is performed during the encoding process before transmission, preliminarily improving image quality by reducing artifacts and enhancing details. This preliminary action ensures that high-resolution data maintains superior quality throughout transmission and storage without requiring additional bandwidth or storage capacity.
Solution Approach 2:
The filtering process modifies parameters such as filter strength, kernel size, and boundary handling based on local image characteristics. By adaptively changing these parameters, the system optimizes compression efficiency and visual quality while minimizing the impact on transmission and storage requirements.
3Productivity
If deblocking and adaptive loop filtering are applied, then compression efficiency is improved, but processing complexity increases
Solution Approach 1:
The filtering process is segmented into distinct stages: deblocking filtering applied first to reduce block artifacts, followed by adaptive loop filtering for further quality enhancement. This segmentation allows each filter to be optimized independently and applied only where needed, improving compression efficiency while managing processing complexity through modular design.
Data Source
AI summary
In relation to an in-loop filtering procedure described in the present document, a virtual boundary is defined so as to further increase the subjective/objective visual quality of a restored picture and the in-loop filtering procedure can be applied across the virtual boundary. The virtual boundary can include, for example, a discontinuous edge such as a 360-degree image, a VR image, or picture in picture (PIP). For example, the virtual boundary can be at a predetermined appointed position, and the existence and/or the position thereof can be signaled. Embodiments of the present document present a method for efficiently signaling virtual boundary-related information.