Video Decoding Matrix Intra Prediction with Downsampling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Matrix intra prediction in video coding schemes faces challenges with increased data amount and calculation complexity due to the need for selecting appropriate matrices and referencing adjacent pixels, which limits prediction performance and efficiency.
Innovation Solution
A video decoding apparatus that includes a matrix reference pixel derivation unit for downsampling adjacent images, a mode derivation unit for candidate list generation, a prediction processing parameter derivation unit, and a matrix prediction image derivation unit to generate and interpolate prediction images, reducing data amount and calculation complexity while maintaining prediction performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the range of adjacent pixels is enlarged to improve prediction performance, then prediction image quality is improved, but calculation amount of matrix operation increases
Solution Approach 1:
The patent segments the reference image into multiple regions (first reference image region and second reference image region) and processes them differently. The first region undergoes downsampling while the second region is used directly, allowing the system to benefit from both reduced calculation and high-quality reference data in different areas.
Solution Approach 2:
Different parts of the reference image are assigned different processing treatments. The first reference image region is downscaled to reduce calculation load, while the second reference image region maintains original quality for accurate prediction. This local differentiation optimizes both calculation efficiency and prediction quality.
2Measurement precision
If multiple prediction modes are defined to improve prediction accuracy, then prediction performance is improved, but amount of coded data for selecting matrix increases
Solution Approach 1:
Instead of coding all possible prediction modes, the patent transcodes only a subset of prediction modes (first and second prediction modes) into a compact representation. This partial encoding reduces the coded data amount while still providing sufficient prediction accuracy for most cases.
Solution Approach 2:
The patent inverts the traditional approach by having the decoder derive prediction modes from trans-coded information rather than explicitly coding all modes. The decoder reconstructs the prediction modes (first and second prediction modes) from the trans-coded first and second prediction mode information, reducing bandwidth requirements.
Data Source
AI summary
A video decoding apparatus includes a matrix reference pixel derivation unit that derives, as a reference image, an image obtained by downsampling an image adjacent to a top side and a left side of a target block,a mode derivation unit that derives a candidate list of prediction modes used for a target block according to the reference image and a size of the target block, a prediction processing parameter derivation unit that derives a prediction processing parameter used to derive a prediction image according to the candidate list, a matrix intra prediction mode indicator, and the size of the target block, a matrix prediction image derivation unit that derives a prediction image based on an element of the reference image and the prediction processing parameter, anda matrix prediction image interpolation unit that derives the prediction image or an image obtained by interpolating the prediction image, as a prediction image. The mode derivation unit derives a candidate list having a number of elements equal to or less than half a total number of prediction modes defined for the size of the target block.