Scalable Video Coding Reference Picture List Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image encoding and decoding technologies face challenges in providing scalability to accommodate diverse user requirements and terminal capabilities, particularly in handling high-definition and ultra-high-definition video content, as they struggle to efficiently manage reference images and optimize encoding/decoding rates across various networks and devices.
Innovation Solution
The method and apparatus for scalable video coding support multiple layers by analyzing and managing layer dependencies through video parameter sets, determining reference picture lists, and applying new layer dependencies to enhance encoding and decoding efficiency, allowing for spatial, temporal, and image quality scalability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple layers are supported for scalability, then adaptability to diverse user requirements is improved, but device complexity increases
Solution Approach 1:
The patent segments the reference picture management into multiple layers, where each layer has its own reference picture list and dependency relationships. This allows independent management of reference images at different scalability levels (spatial, temporal, quality) without interfering with each other, thus enabling scalability while maintaining manageable complexity through structured organization.
Solution Approach 2:
The patent introduces a layer dimension to the reference picture management system, organizing reference images not just in temporal or spatial dimensions but across multiple quality and resolution layers. This dimensional extension enables scalable video coding by allowing the decoder to selectively process references from different layers based on capability and requirement.
2Manufacturing precision
If layer dependency analysis is performed for each slice, then encoding precision is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary analysis of layer dependency relationships at the slice level during the encoding setup phase. By pre-determining which layers depend on which reference layers and organizing reference picture lists in advance, the actual encoding process can proceed more efficiently without repeated complex analysis, thus achieving high encoding precision while reducing processing time during active encoding.
Solution Approach 2:
The patent applies different layer dependency configurations to different slices locally. Each slice can have its own reference picture list and layer dependency settings optimized for its specific content characteristics, allowing high encoding precision for each local region while avoiding the need for complex global analysis of the entire video stream.
3Manufacturing precision
If reference picture lists are constructed based on layer dependency, then image quality is improved, but memory requirements increase
Solution Approach 1:
The patent creates reference picture lists that serve multiple functions simultaneously: they maintain layer dependency relationships for quality scalability, provide temporal reference for motion compensation, and support spatial reference for inter-layer prediction. This multi-functionality allows the system to achieve high image quality through comprehensive reference management without proportionally increasing memory requirements, as the same reference structures serve multiple purposes.
Data Source
AI summary
This document discloses a method and apparatus for image encoding/decoding which support multiple layers. The method for image decoding comprises analyzing a video parameter set (VPS) extension for layer dependency on a current layer, analyzing slices for layer dependency on a current slice by determining whether not to apply layer dependency to the current slice or to apply new layer dependency to the current slice, the new layer dependency determining within a range of layer dependency analyzed from the VPS extension, and constructing a reference picture list about the current layer based on at least one of layer dependency analyzed from the VPS extension and layer dependency analyzed from the slice.