Chroma Intra Prediction With Compact LUT and Wide-Angle Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in achieving efficient compression of video data without sacrificing picture quality, particularly in handling chroma components with different subsampling ratios in horizontal and vertical directions.
Innovation Solution
A method and apparatus for chroma intra prediction that involves deriving a chroma intra prediction mode from a look-up table using the initial luma intra prediction mode, followed by wide-angle mapping to obtain modified intra prediction modes, and using the derived parameters to predict chroma component samples.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If chroma intra prediction mode is determined using a large look-up table, then prediction accuracy is improved, but memory usage and device complexity increase
Solution Approach 1:
The patent segments the chroma intra prediction mode determination into multiple steps: first deriving a preliminary mode from luma mode using a compact LUT, then applying wide-angle mapping to adjust for different subsampling ratios. This segmentation allows using a smaller LUT while achieving accurate predictions through subsequent refinement steps.
Solution Approach 2:
The patent performs preliminary derivation of chroma intra prediction mode from luma intra prediction mode using a compact look-up table before final adjustment. This preliminary action reduces the immediate memory requirement while maintaining the ability to achieve accurate final predictions through the wide-angle mapping step that follows.
2Measurement precision
If chroma intra prediction mode is determined using a large look-up table, then prediction accuracy is improved, but compression efficiency deteriorates
Solution Approach 1:
The patent segments the mode determination process to use a compact LUT followed by wide-angle mapping, reducing the bit rate overhead for transmitting chroma prediction modes while maintaining prediction accuracy. This segmentation improves compression efficiency by minimizing the information that needs to be encoded.
Solution Approach 2:
The patent performs preliminary derivation of chroma mode from luma mode using a compact LUT, reducing the number of bits needed to signal chroma prediction modes. This preliminary action improves compression efficiency by leveraging the correlation between luma and chroma modes while requiring less transmission overhead.
3Measurement precision
If different subsampling ratios are used in horizontal and vertical directions, then chroma component accuracy is improved, but processing complexity increases
Solution Approach 1:
The patent applies asymmetric wide-angle mapping that specifically accounts for different horizontal and vertical subsampling ratios. The mapping function adjusts the chroma intra prediction mode based on the aspect ratio and subsampling configuration, allowing accurate prediction for non-square chroma blocks while maintaining a relatively simple processing framework.
Solution Approach 2:
The patent changes the intra prediction mode parameter through wide-angle mapping based on the subsampling ratio and block aspect ratio. This parameter transformation allows the system to adapt to different subsampling configurations (4:2:0, 4:2:2, etc.) and block shapes without requiring completely different processing pipelines, thus managing complexity.
Data Source
AI summary
Devices and methods of directional intra prediction for chroma component of a picture are provided. The method includes obtaining an initial intra prediction mode of the chroma component based on a value of a luma intra prediction mode (intraPredModeY) by using a first look up table (LUT), deriving a chroma intra prediction mode (intraPredModeC) from a second LUT by using the initial intra prediction mode of the chroma component. The chroma component has different subsampling ratios in horizontal and vertical directions. The second LUT has 67 entries, indices for the second LUT are 0-66, and each index is associated with an angle of an intra prediction direction. The method further includes obtaining an angle parameter for the chroma component based on the chroma intra prediction mode intraPredModeC; and obtaining predicted samples of the chroma component based on the angle parameter.