Scalable Video Encoding Dead Sub-stream for Quality Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing scalable video coding techniques, such as the MPEG-21 SVM model, have limitations in achieving optimal bit rate points for decoding, often resulting in suboptimal image reconstruction quality, especially for lower spatial resolutions, and fail to utilize refinement information effectively during extraction.
Innovation Solution
The proposed method introduces a 'dead sub-stream' in the encoding process, which is not used for prediction of higher layers but allows for higher quality reconstruction at specific resolution levels, and includes additional information to identify this sub-stream and its quality, enabling selective quality adaptation and optimal extraction paths during decoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If scalable video coding techniques (MPEG-21 SVM model) are used to encode video streams with hierarchical layers, then adaptability to different bit rates and resolutions is improved, but image reconstruction quality at specific resolution levels deteriorates due to suboptimal bit rate points
Solution Approach 1:
The patent segments the video data stream into multiple resolution levels (base layer and enhancement layers), where each layer can be independently decoded at optimal bit rate points. This segmentation allows the system to maintain adaptability across different resolutions while ensuring each level achieves optimal reconstruction quality through separate optimization of bit rate points.
Solution Approach 2:
The patent introduces the concept of optimizing bit rate points as a separate parameter for each resolution level. By allowing different bit rate points for different layers (e.g., base layer at one bit rate point, enhancement layer at another), the system can simultaneously maintain adaptability and achieve optimal image reconstruction quality at each specific resolution level.
2Productivity
If refinement information is used for prediction in higher layers, then coding efficiency is improved, but the ability to achieve optimal quality at specific resolution levels deteriorates due to constrained extraction paths
Solution Approach 1:
The patent introduces dynamic selection of extraction paths based on the target resolution level. The extraction mechanism can dynamically choose whether to use refinement information from lower layers for prediction, depending on whether the target is optimal coding efficiency or optimal quality at a specific resolution. This dynamic approach resolves the contradiction by making the prediction strategy adaptable to the specific quality requirements.
Solution Approach 2:
The patent performs preliminary encoding of multiple layers with different bit rate optimizations. During decoding, the system can selectively extract and combine layers based on the desired outcome - either maximizing coding efficiency by using refinement information or maximizing quality at specific resolutions by selecting optimal bit rate points for each layer.
3Device complexity
If a single data stream is used to serve multiple resolution requirements, then device complexity is reduced, but the ability to optimize bit rate points for each resolution level deteriorates
Solution Approach 1:
The patent creates a universal data stream structure that serves multiple resolution requirements simultaneously. The hierarchical layering allows a single encoded stream to be decoded at different resolution levels, each with its own optimized bit rate point. This multi-functional approach reduces device complexity compared to maintaining separate encoding systems for each resolution while preserving the ability to optimize bit rate points for each level.
Data Source
AI summary
A method is provided for encoding a sequence of images generating a data stream in the form of a structure of embedded data layers of n successive levels each of which corresponds to the predetermined resolution of said images. Said method comprises an encoding stage, which encodes at least one of said layers of an n+1 level by prediction on the basis of said level n layer and encodes each data layer of the level n in the form of a base sub-stream and, optionally, in the form of at least one enhancement sub-stream enabling to obtain at least one version of the enhanced quality of said images. The method also encodes, for at least one data layer of the n level, at least one single sub-stream enabling to reconstruct the version of said images at said level n resolution with a quality higher than said enhanced qualities, wherein said single sub-steam of the level n layer is not used during the prediction encoding of the layers of a level other than the level n.