Difference Domain Intra Prediction for Scalable Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Scalable video coding (SVC) faces challenges in achieving effective rate-distortion trade-off due to the high frequency nature of difference domain pixels, which weakens the correlation between current and neighboring difference pixels, leading to poor prediction results in DC mode and most probable mode determination.
Innovation Solution
The implementation of Zero DC Mode and DC Offset Mode in SVC, where the DC value is set to zero or offset by a value, respectively, without considering neighboring difference pixels for prediction, and bypassing transform and residual coding when it does not lead to better rate-distortion trade-off, along with defining a Most Probable Mode list based on characteristics of the difference domain.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If DC mode is used for intra prediction in difference domain, then prediction is performed using neighboring difference pixels, but prediction accuracy deteriorates due to weak correlation caused by high frequency components
Solution Approach 1:
The patent extracts and removes the problematic high frequency components from the difference domain pixels before performing DC mode prediction. By filtering out these high frequency components that cause weak correlation, the prediction accuracy is improved while maintaining the simplicity of DC mode.
Solution Approach 2:
The patent changes the state of the difference pixels by filtering out high frequency components, transforming them into a state more suitable for DC mode prediction. This parameter change (removing high frequency content) restores the correlation between current and neighboring pixels, thereby improving prediction accuracy.
2Productivity
If transform and residual coding are performed for all coding units, then compression is achieved, but rate-distortion trade-off deteriorates when it does not lead to better compression efficiency
Solution Approach 1:
The patent introduces a dynamic decision mechanism that adaptively determines whether to perform transform and residual coding for each coding unit based on specific criteria. This dynamic approach allows the system to skip unnecessary transform and residual coding operations, improving rate-distortion trade-off by avoiding redundant processing that does not contribute to compression efficiency.
Solution Approach 2:
The patent selectively discards (skips) transform and residual coding operations for coding units where they would not improve compression efficiency. By conditionally omitting these operations, the system avoids wasting computational resources and maintains better rate-distortion trade-off, while still achieving adequate compression through other means.
3Ease of operation
If neighboring difference pixels are used for DC prediction, then prediction is performed using available reference data, but prediction quality deteriorates due to high frequency noise
Solution Approach 1:
The patent converts the harmful effect of high frequency components (which cause noise and poor prediction quality) into a benefit by selectively removing them. The filtering process transforms the problematic high frequency content into a cleaner signal that maintains the simplicity of DC mode while improving prediction quality.
Data Source
AI summary
An apparatus for coding video information according to certain aspects includes a memory unit and a processor in communication with the memory unit. The memory unit stores difference video information associated with a difference video layer of pixel information derived from a difference between an enhancement layer and a corresponding base layer of the video information. The processor determines a DC prediction value for a video unit associated with the difference video layer while refraining from using pixel information from a neighboring area of the video unit, wherein the DC prediction value is equal to zero or is offset by an offset value. The DC prediction value is a prediction value used in intra prediction based at least on an average of neighboring video units of the video unit. The processor further determines a value of the video unit based at least in part on the DC prediction value.