Picture Header NAL Ordering for Efficient Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face inefficiencies in determining the starting point of a picture within a multi-layer bitstream, leading to unnecessary bit usage and reduced coding efficiency.
Innovation Solution
Ensuring that a picture header network abstraction layer unit precedes a first video coding layer unit in a picture unit, eliminating the need for a picture header identifier, thereby reducing bits in the bitstream and improving coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If a picture header identifier is signaled in the picture header to identify the associated picture, then picture association can be determined, but additional bits are consumed in the bitstream
Solution Approach 1:
The patent extracts and removes the picture header identifier field from the picture header structure. By taking out this unnecessary identification element, the patent eliminates the trade-off between picture association information and bitstream size, since picture association can be determined through the sequential structure of NAL units rather than explicit identifiers.
Solution Approach 2:
The patent makes the picture header structure universal across different picture types and coding scenarios by removing picture-specific identifiers. The same picture header format can be used for all pictures in the bitstream, with picture association inferred from the contextual position and structure of NAL units rather than explicit identification fields.
2Reliability
If picture header identifiers are included in each picture header, then picture associations can be explicitly identified, but coding efficiency is reduced due to increased bit usage
Solution Approach 1:
The patent enables the bitstream structure to self-identify picture associations through its inherent organizational structure. The sequential arrangement of NAL units, where each picture header precedes its associated VCL NAL units, allows the system to automatically determine picture associations without external identifiers, thereby maintaining reliability while improving coding efficiency.
Solution Approach 2:
The patent changes the fundamental parameter of picture header identification from explicit numerical identifiers to implicit structural positioning. By transforming the identification mechanism from content-based (identifiers) to structure-based (positioning), the patent achieves both reliable picture association and improved coding efficiency.
Data Source
AI summary
A method of decoding is provided. The method includes receiving a video bitstream including a picture unit (PU), where the PU includes a picture header (PH) network abstraction layer (NAL) unit preceding a first video coding layer (VCL) NAL unit, and decoding a coded picture from the PU to obtain a decoded picture.