Shared Picture Header Parameters to Reduce Slice Signaling Redundancy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently compressing and decompressing video data without sacrificing image quality, particularly in limited bandwidth scenarios, leading to redundancy and inefficiencies in signaling syntax elements across slices.
Innovation Solution
Signaling syntax elements corresponding to an entire picture in a picture header instead of each slice header, reducing redundancy and improving coding efficiency by ensuring these elements are only signaled once per picture.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If syntax elements are repeated in each slice header, then each slice can be independently decoded, but redundancy increases and coding efficiency decreases
Solution Approach 1:
The picture is divided into multiple slices that can be independently decoded, allowing parallel processing and resilient decoding. Each slice contains slice header syntax elements that enable independent decoding while sharing common picture parameters signaled at the picture header level.
Solution Approach 2:
Common picture parameters such as picture order count, picture type, and scanning information are merged into a single picture header that is shared across all slices, eliminating redundant signaling while maintaining the ability to independently decode each slice.
2Adaptability or versatility
If syntax elements are signaled in each slice header, then slice-level control is enabled, but bitstream size increases
Solution Approach 1:
Slice-specific control information is extracted into separate slice header syntax elements, while common picture parameters are taken out of the slice headers and placed in a picture header, reducing the overall bitstream size while preserving slice-level control capabilities.
Solution Approach 2:
Different parts of the bitstream have different levels of detail: the picture header contains common parameters at a higher level, while slice headers contain only the necessary slice-specific control information, optimizing the balance between control flexibility and data efficiency.
3Loss of information
If picture parameters are repeated for each slice, then complete information is available in each slice, but redundancy increases
Solution Approach 1:
The picture header serves multiple functions by providing common parameters that are universally applicable to all slices in the picture, eliminating the need to restate these parameters in each slice header while ensuring information completeness.
Solution Approach 2:
Redundant parameter signaling is discarded by moving common parameters to the picture header level, and information completeness is recovered through the picture parameter set (PPS) that references both the picture header and slice headers, ensuring all necessary information is available without repetition.
Data Source
AI summary
A method of decoding is provided. The method includes receiving a video bitstream including a picture unit (PU) having a picture header (PH) with a PH syntax structure, where the PH syntax structure includes the following syntax elements common for all slices of a coded picture associated with the PH syntax structure: a non-reference picture flag, a picture parameter set (PPS) identifier (ID), a picture order count (POC) least significant bit (LSB), a no output of prior pictures flag, a recovery point POC, and a picture output flag; and decoding the coded picture from the PU to obtain a decoded picture.