Loop filtering method, device and equipment in video encoding and decoding, and storage medium

Abstract

The invention provides a loop filtering method, device and equipment in video encoding and decoding and a storage medium, and relates to the technical field of video encoding and decoding. The methodcomprises the following steps: obtaining a reconstruction frame corresponding to a current frame in a video and residual distribution information, wherein the residual distribution information refersto information influencing or reflecting residual signal data distribution; performing fusion processing on the reconstruction frame and the residual distribution information to obtain fusion data; and performing filtering processing on the fused data through a loop filtering model to obtain a filtered reconstruction frame. According to the invention, the information capable of influencing or reflecting residual signal data distribution is added into input data of a loop filtering model, so the model can better learn the mapping (simulation quantization inverse process) from the quantized residual error to the residual error before quantization, so that quantization losses of different degrees are recovered or relieved, the model can be designed according to factors influencing the codingloss degree, the sensitivity of the loop filtering model to the quantization loss degree is improved, and the filtering quality of the model is improved.

Description

technical field [0001] The embodiments of the present application relate to the technical field of video coding and decoding, and in particular, to a method, device, device, and storage medium for loop filtering in video coding and decoding. Background technique [0002] In recent years, with the development of deep learning technology, many studies have attempted to use deep learning models to replace some modules in traditional video coding frameworks or propose end-to-end coding frameworks based on deep learning, especially in loop filtering tools. , which can significantly improve the encoding performance compared with traditional filtering tools. The main idea of ​​the loop filter tool based on deep learning is to learn the mapping relationship between the reconstruction image and the original image through the neural network, which is usually used to replace or help the traditional filter tool to repair the information loss caused by the lossy coding and suppress the c...

AI Technical Summary

Problems solved by technology

However, the current loop filtering tool based on deep learning does not fully consider the factors that affect the degree of coding loss in the model design, resulting in obvious room for improvement in the filtering quality.

Images