Intra-frame prediction method and device

By using neural networks to generate multiple prior candidate intra prediction modes of the current block and their corresponding probability values ​​in the video encoding technology, the problems of low intra prediction accuracy, large errors and low rate distortion optimization efficiency in the prior art are solved, and more efficient video compression is achieved.

CN114286099BActive Publication Date: 2025-03-21HUAWEI TECH CO LTD
2 Cites 0 Cited by

Patent Information

Application Number
CN202011043931.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2020-09-28
Publication Date
2025-03-21
Estimated Expiration
2040-09-28

AI Technical Summary

Technical Problem

The existing video encoding technology has low accuracy, large errors in intra prediction, and low efficiency in rate distortion optimization, making it difficult to improve compression rate under limited network resources and high video quality requirements.

Method used

By obtaining the intra prediction mode or texture distribution of multiple reconstructed image blocks in the surrounding area of ​​the current block, a neural network is used to generate multiple prior candidate intra prediction modes of the current block and their corresponding probability values, calculate the weight factor and weight sum to obtain the predicted value of the current block.

Benefits of technology

The accuracy of intra prediction is improved, the error of intra prediction is reduced, the overall rate distortion optimization efficiency of intra prediction is improved, and the performance of video compression is improved.

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

Abstract

The present application provides an intra-frame prediction method and device. It relates to the field of video or image compression technology based on artificial intelligence (AI), and specifically to the field of video compression technology based on neural networks. The method includes: obtaining the intra-frame prediction mode or texture distribution of each of the P reconstructed image blocks in the surrounding area of ​​the current block; obtaining Q prior candidate intra-frame prediction modes and Q probability values ​​of the current block according to the intra-frame prediction mode or texture distribution of each of the P reconstructed image blocks; obtaining M weight factors corresponding to the M prior candidate intra-frame prediction modes according to the M probability values ​​corresponding to the M prior candidate intra-frame prediction modes; performing intra-frame predictions according to the M prior candidate intra-frame prediction modes to obtain M prediction values; and obtaining the prediction value of the current block by weighted summation of the M prediction values ​​and the corresponding M weight factors. The present application can improve the accuracy of intra-frame prediction, reduce the error of intra-frame prediction, and improve the RDO efficiency of intra-frame prediction.
Need to check novelty before this filing date? Find Prior Art