Upsampling Process for Scalable Video Coding Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In scalable video coding, existing methods face challenges in efficiently upsampling base layer video data to match the resolution of enhancement layers, particularly in achieving accurate alignment and phase correspondence between layers for effective prediction and encoding.
Innovation Solution
The proposed solution involves a three-module upsampling process that selects input samples from the base layer, chooses appropriate filters based on phase filtering, and uses these filters to recreate video data at the enhancement layer resolution, utilizing a set of fixed filters with different phase shifts to align sample positions between the base and enhancement layers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If upsampling filtering is applied to the base layer to match enhancement layer resolution, then spatial scalability is improved, but computational complexity and filtering requirements increase
Solution Approach 1:
The upsampling process is divided into three separate modules: (1) selecting input samples from the base layer video signal, (2) selecting appropriate filters from a set of fixed filters based on phase information, and (3) applying phase filtering to generate output samples. This segmentation allows each module to be optimized independently and simplifies the overall filtering complexity.
Solution Approach 2:
A set of fixed filters with different phase shifts is pre-computed and stored in memory. During upsampling, the appropriate filter is selected from this pre-prepared set based on the calculated phase, eliminating the need for real-time filter design and reducing computational complexity during actual processing.
2Measurement precision
If phase filtering is used to align sample positions between layers, then prediction accuracy is improved, but the complexity of filter selection and phase calculation increases
Solution Approach 1:
The phase parameter is calculated based on the resolution ratio between base layer and enhancement layer, and this phase value directly determines which pre-computed filter from the set should be used. By changing the filter selection criterion to a simple phase-based index, the system achieves high prediction accuracy without complex real-time filter optimization.
3Manufacturing precision
If multiple fixed filters with different phases are used for upsampling, then interpolation accuracy is improved, but memory requirements and filter management complexity increase
Solution Approach 1:
Instead of using a single complex adaptive filter or an excessively large number of filters, the invention uses a moderate set of fixed filters with specific phase shifts (e.g., 0, π/4, π/2, 3π/4). This partial approach provides sufficient interpolation accuracy for most practical cases while keeping memory requirements manageable.
Data Source
AI summary
A process for determining the selection of filters and input samples is provided for scalable video coding. The process provides for re-sampling using video data obtained from an encoder or decoder process of a base layer (BL) in a multi-layer system to improve quality in Scalable High Efficiency Video Coding (SHVC). In order to provide better alignment between layers, it is proposed that reference layer offset adjustment parameters be signaled.