Region of Interest Scalability in SHVC Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies, such as Scalable HEVC, enhance entire picture resolution and quality, which is not necessary for applications like traffic and security monitoring, leading to inefficient resource utilization.
Innovation Solution
Implementing Region of Interest (ROI) scalability in SHVC, where scalability is applied only to specific parts of the picture, allowing for independent enhancement of pixels within defined ROIs, using methods like explicit and implicit signaling of ROI maps, and differential encoding to optimize resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If scalability is applied to the entire picture, then higher resolution and quality are achieved, but resource utilization becomes inefficient
Solution Approach 1:
The patent applies different quality levels to different regions of the picture by defining Regions of Interest (ROIs) that receive enhanced resolution and quality through scalability, while other regions maintain baseline quality. This resolves the contradiction by concentrating resources only where high quality is needed rather than uniformly across the entire picture.
Solution Approach 2:
The patent segments the picture into multiple regions with different quality requirements by introducing ROI markers and slice segment structures. This allows the video coding system to process different regions with appropriate resource allocation, improving overall efficiency while maintaining high quality in critical areas.
2Measurement precision
If scalability is applied to the entire picture, then higher resolution is provided, but computational load increases
Solution Approach 1:
The patent reduces computational load by applying high-resolution scalability only to specific ROI regions rather than the entire picture. The encoder and decoder can skip or simplify processing in non-ROI areas, significantly reducing computational requirements while maintaining high resolution where needed.
Solution Approach 2:
The patent implements partial scalability by applying enhancement only to necessary regions (ROIs) rather than the complete picture. This partial action approach achieves the required resolution improvement with reduced computational effort compared to full-picture scalability.
3Measurement precision
If enhancement is applied to the entire picture, then quality is improved, but resource allocation becomes inefficient
Solution Approach 1:
The patent improves resource allocation efficiency by directing enhancement resources specifically to ROI regions that require higher quality, such as areas with important visual information. Non-ROI regions receive standard quality processing, optimizing the overall resource utilization while maintaining acceptable quality across the entire picture.
Solution Approach 2:
The patent changes the quality parameter selectively across different picture regions by using ROI signaling mechanisms and differential encoding. This allows dynamic adjustment of resource allocation based on content importance, improving productivity by matching resource投入 with actual quality requirements of different regions.
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
Region of Interest (ROI) scalability with SHVC is able to be implemented where scalability is used for part of a picture but not the whole picture. Applications of ROI scalability include traffic monitoring, security monitoring and tiled streaming.