Sub-Picture Boundary Filtering Control in Video Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Video coding systems face issues with filter-related errors when sub-pictures are independently extracted due to unexpected filtering operations across sub-picture boundaries, leading to noise and artifacts in decoded images.
Innovation Solution
Implementing a flag to control filtering operations at the sub-picture level, allowing filters to be applied or omitted across sub-picture boundaries based on the flag's setting, ensuring accurate decoding and display of sub-pictures without errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If filtering operations are applied across sub-picture boundaries, then coding artifacts are removed and image quality is improved, but errors occur when sub-pictures are independently extracted
Solution Approach 1:
The filtering operation dynamically adapts its behavior based on the extraction mode. When sub-pictures are encoded for independent extraction, the filter is configured to operate only within sub-picture boundaries. When sub-pictures are encoded for display as a group, the filter operates across boundaries. This dynamic configuration resolves the contradiction by making the filtering behavior context-dependent.
Solution Approach 2:
The filtering operation applies different processing quality and scope to different regions. Within sub-picture interiors, full filtering is applied regardless of mode. At sub-picture boundaries, the filtering scope changes based on the encoding mode - restricted to intra-sub-picture regions for extraction mode, and extended across boundaries for group display mode. This local differentiation resolves the contradiction between artifact removal and decoding accuracy.
2Adaptability or versatility
If sub-pictures are encoded for independent extraction, then extraction flexibility is improved, but filter-related errors occur at sub-picture boundaries
Solution Approach 1:
The filtering operation is segmented into independent regions based on sub-picture boundaries. When extraction mode is enabled, the filter is restricted to operate only within each sub-picture's interior, creating independent filtering segments that do not cross boundaries. This segmentation allows independent extraction without filter-related errors while maintaining extraction flexibility.
Solution Approach 2:
The system preemptively prevents filter-related errors by detecting the extraction mode beforehand and configuring the filter accordingly. Before filtering operations are applied, the system checks whether sub-pictures are encoded for independent extraction and restricts filter boundaries in advance. This preliminary anti-action prevents the harmful effect of cross-boundary filtering errors before they can occur.
3Manufacturing precision
If filtering operations are applied across tile boundaries, then coding artifacts are removed, but errors occur when sub-pictures are independently extracted
Solution Approach 1:
The filtering operation dynamically adjusts its spatial scope based on the encoding mode. When independent extraction is enabled, the filter dynamically restricts its operation to remain within sub-picture boundaries even when tiles cross those boundaries. When group display is enabled, the filter operates across tile boundaries normally. This dynamic adaptation eliminates filtering errors while maintaining image quality.
Solution Approach 2:
The filtering operation applies different processing rules to different spatial locations. At locations where tiles cross sub-picture boundaries, the filter behavior changes based on extraction mode - restricted to intra-sub-picture regions for extraction mode, and extended across boundaries for group display mode. This local quality differentiation prevents filtering errors at critical boundary locations.
Data Source
AI summary
A video coding mechanism is disclosed. The mechanism includes receiving a bitstream comprising a flag and a picture including a sub-picture with a sub-picture boundary. The sub-picture is decoded. Filtering operations are performed across the sub-picture boundary when the flag is set to a first value. The filtering operations across the sub-picture boundary are omitted when the flag is set to a second value. The sub-picture is forwarded for display as part of a decoded video sequence.