Cross-Component Picture Prediction With Statistical Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards like H.266/VVC do not adequately consider the differences in statistical characteristics of various color components, leading to low prediction efficiency in cross-component prediction.
Innovation Solution
A method for picture prediction that involves obtaining an initial predicted value through a prediction model, filtering it, and adjusting it to balance the statistical characteristics of different color components, using techniques such as filtering, grouping, value modification, quantization, and de-quantization to improve prediction efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If cross-component prediction is performed without considering statistical characteristics differences, then the prediction process is simple, but prediction efficiency is low
Solution Approach 1:
The patent changes the statistical parameters of the predicted component by applying filtering operations (arithmetic filtering, geometric filtering, harmonic filtering) to transform the predicted values. This parameter transformation aligns the statistical characteristics of different color components, improving prediction efficiency while maintaining a relatively simple process structure.
Solution Approach 2:
The patent introduces filtering operations as an intermediary step between the initial prediction model and the final predicted values. This intermediary processing layer transforms the statistical characteristics without requiring a complete redesign of the prediction model, thus improving efficiency while controlling complexity.
2Measurement precision
If filtering operations are applied to balance statistical characteristics, then prediction accuracy improves, but processing complexity increases
Solution Approach 1:
The patent applies parameter transformation through filtering operations that modify the statistical characteristics of predicted values. By changing parameters such as mean and variance through arithmetic, geometric, or harmonic filtering, the prediction accuracy improves while the processing remains computationally manageable.
Solution Approach 2:
The filtering operations are applied locally to specific predicted values rather than transforming the entire prediction system. This localized approach improves accuracy at specific points without requiring global system redesign, thus controlling processing complexity.
3Productivity
If cross-component prediction is performed, then coding efficiency improves, but residual errors increase due to statistical differences
Solution Approach 1:
The patent transforms the statistical parameters of the predicted component through filtering operations, aligning them with the reference component's characteristics. This parameter alignment reduces residual errors by ensuring that the predicted values have compatible statistical properties with the actual values, thereby improving prediction precision while maintaining coding efficiency.
Data Source
AI summary
A method for picture prediction, an encoder, and a decoder are provided. The method includes the following. An initial predicted value of a colour component to-be-predicted of a current block in a picture is obtained through a prediction model. The initial predicted value is filtered and a target predicted value of the colour component to-be-predicted of the current block is obtained.