Picture Encoding and Decoding with Direct MIP Derivation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The current Versatile Video Coding (VVC) standard requires significant computational resources and storage space for deriving matrix-based Intra Prediction (MIP) blocks, which is inefficient.
Innovation Solution
A method is introduced to determine a suitable MIP size identifier when the width and height of a coding block are a positive integer power of 2, allowing direct derivation of an MIP prediction block of the same size as the current block, eliminating the need for an up-sampling process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If matrix-based Intra Prediction is used to achieve higher coding quality, then prediction accuracy is improved, but computational resources and storage space requirements increase significantly
Solution Approach 1:
The patent segments the prediction process by introducing a reduced set of reference samples (inSize) compared to the full block size. Instead of using all N² reference samples, the method uses a subset of size inSize, thereby reducing computational complexity while maintaining prediction accuracy through strategic sample selection and weighting.
Solution Approach 2:
The patent changes the parameter inSize to control the number of reference samples used in the prediction calculation. By adjusting inSize based on the coding block size and prediction mode, the system optimizes the balance between prediction accuracy and computational complexity, reducing the burden on processing resources.
2Device complexity
If matrix-based Intra Prediction uses a reduced set of reference samples, then computational complexity is reduced, but prediction accuracy may deteriorate
Solution Approach 1:
The patent applies local quality by using different weighting factors (mWeight[i][j]) for different reference samples based on their spatial position and importance. This allows the system to prioritize certain reference samples over others, maintaining prediction accuracy even when using a reduced set of samples by concentrating computational effort on the most informative samples.
Solution Approach 2:
The patent creates a composite prediction approach by combining multiple reference samples with different weighting factors in a unified prediction formula. This composite method integrates the strengths of individual reference samples while mitigating their individual limitations, achieving both reduced complexity and maintained accuracy through the weighted combination of inSize reference samples.
3Area of moving object
If up-sampling process is included in MIP derivation, then prediction block size is increased, but processing time and computational load increase
Solution Approach 1:
The patent inverts the conventional approach by directly deriving the prediction block at the desired output size without performing up-sampling. Instead of deriving a smaller prediction block and then up-sampling to the target size, the method directly computes the full-sized prediction block using the reduced reference sample approach, thereby eliminating the time-consuming up-sampling step while maintaining the required block size.
Data Source
AI summary
A method for decoding a picture, a method for encoding a picture, an encoder, and a decoder are provided. The method for encoding a picture includes (i) determining a width and a height of a coding block in the picture; (ii) if the width and the height are equal to N, where N is a positive integer power of 2, determining a matrix-based intra prediction (MIP) size identifier indicating that an MIP prediction size equal to N; (iii) deriving a group of reference samples of the coding block; and (iv) deriving an MIP prediction of the coding block based on the group of reference samples and an MIP matrix corresponding to the MIP size identifier.