Video Signal Residual Data Up-Sampling Using Motion Compensation Blocks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing Scalable Video Codec (SVC) schemes face inefficiencies in image quality degradation at low bit rates, particularly when up-sampling residual data for interlayer prediction, which complicates the encoding and decoding processes.
Innovation Solution
A method for efficiently up-sampling residual data based on motion compensation, using bi-linear interpolation for dyadic resolution ratios and 6 tap interpolation for non-dyadic ratios, with different filters applied to luminance and chrominance data to simplify the process and reduce computational load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If residual data is up-sampled using traditional methods (bi-linear interpolation for dyadic, 6 tap for non-dyadic) with transform block boundaries, then image quality is maintained, but computational complexity increases and processing becomes more difficult
Solution Approach 1:
The patent changes the fundamental parameter of up-sampling by abandoning traditional interpolation methods (bi-linear and 6 tap filters) in favor of a simple repetition method where each pixel value is repeated to fill the up-sampled block. This parameter change dramatically reduces computational complexity while maintaining acceptable image quality for video coding applications.
Solution Approach 2:
The patent employs a simple, computationally inexpensive up-sampling method that sacrifices some image quality precision compared to traditional filters, but gains significant computational efficiency. The repetition method is analogous to using a simple, disposable solution rather than a complex, expensive one.
2Measurement precision
If different up-sampling methods are applied to luminance and chrominance data, then image quality is improved, but processing time and computational load increase
Solution Approach 1:
The patent applies different up-sampling strategies to different components: luminance data uses the simple repetition method while chrominance data uses traditional bi-linear interpolation. This local differentiation optimizes processing by applying complex methods only where necessary (chrominance) and simple methods where sufficient (luminance), balancing quality and speed.
Solution Approach 2:
The patent changes the up-sampling parameter based on data type: using repetition for luminance and bi-linear interpolation for chrominance. This parameter adaptation allows the system to optimize between computational efficiency and image quality for different color components.
3Manufacturing precision
If transform block boundaries are respected during up-sampling, then encoding accuracy is improved, but the encoding process becomes more complex
Solution Approach 1:
The patent changes the approach to boundary handling by using motion compensation block boundaries instead of transform block boundaries for determining up-sampling regions. This parameter change simplifies the process because motion compensation blocks are already available from the decoding process, eliminating the need for additional transform block analysis.
Data Source
AI summary
Disclosed herein is a method of encoding video signals. The method includes creating a bit stream of a first layer by encoding the video signals, and creating a bit stream of a second layer by encoding the video signals based on the first layer. When residual data, corresponding to an image difference, within the first layer, is up-sampled and used for the encoding of the second layer, the residual data is up-sampled for each block that is predicted based on motion compensation.