Cross-component-based correction for intra predictor of chroma block

By correcting chroma prediction values using a cross-component model based on luma samples, the method enhances encoding efficiency and image quality in video data compression, addressing the challenges of increasing image sizes and frame rates.

WO2025234639A1 Publication Date: 2025-11-13HYUNDAI MOTOR CO LTD +2
5 Cites 0 Cited by

Patent Information

Application Number
PCT/KR2025/005249
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2024-05-08
Filing Date
2025-04-17
Publication Date
2025-11-13

AI Technical Summary

Technical Problem

Existing video compression technologies struggle to efficiently encode and decode video data with increasing image sizes, resolutions, and frame rates, necessitating improved encoding efficiency and image quality, particularly in handling chroma blocks.

Method used

A method is proposed to correct chroma prediction values using a cross-component model based on the correlation between luma and chroma samples, involving generating first and second chroma prediction values, deriving a weighted average, and encoding residual data to enhance encoding efficiency.

Benefits of technology

This approach improves encoding efficiency and image quality by leveraging the correlation between luma and chroma components, optimizing chroma block prediction in video data compression.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

A cross-component-based correction technique for an intra predictor of a chroma block is disclosed. A video encoder or a video decoder may: generate first chroma prediction values for chroma samples of a chroma block by using an intra prediction technique; derive a cross-component model on the basis of luma prediction values for luma samples of a luma block corresponding to the chroma block and the first chroma prediction values; and generate second chroma prediction values for the chroma samples of the chroma block from reconstructed values for the luma samples by using the cross-component model. The video encoder or the video decoder may generate modified chroma prediction values for the chroma samples of the chroma block on the basis of a weighted average of the first chroma prediction values and the second chroma prediction values, and encode or decode the chroma block on the basis of the modified chroma prediction values.
Need to check novelty before this filing date? Find Prior Art