A Fast Judgment Method, System and Storage Medium for Multifunctional Video Coding

Abstract

The invention discloses a multi-functional video coding fast decision method, system and storage medium. The method includes: calculating the texture value of the input coding tree unit; performing quadtree division on the input coding tree unit to obtain 4 coding units ; Calculate the texture difference value of several sub-parts in each coding unit, and perform a skip judgment on the division mode of the coding unit according to the calculated texture difference value and a preset threshold; determine the division mode as The division modes not skipped are added to the candidate list; the coding unit is divided according to the division mode in the candidate list. The present invention utilizes the texture features of each sub-part in the coding unit to perform the skip judgment of the division mode of the coding unit, and can reduce the invalid division operation of QTMT by skipping the judgment and improve the coding efficiency; it is applicable to all partition tree types of VCC The QTMT partition scheme has wide applicability. The invention can be widely applied in the field of video coding.

Description

technical field [0001] The invention relates to the field of video coding, in particular to a fast decision method, system and storage medium for multifunctional video coding. Background technique [0002] H.266 / Versatile Video Coding (VVC for short, Versatile Video Coding) is a new generation of video coding standard, which is still in the development stage. Compared with the previous generation standard H.265 / HEVC, VVC retains HEVC's block prediction-based technical framework and introduces many new technologies and coding tools. For example, VVC introduces Quadtree with Nested Multi-type Tree Using Binary and Ternary Splits (QTMT), Position Dependent Intra Prediction Combination (PDPC), Intra Sub-Partitions (ISP), etc. [0003] The QTMT partitioning scheme in VVC is very different from the partitioning scheme in HEVC. HEVC only supports quadtree partitioning, so after partitioning, all sub-blocks have only a square shape. However, in VVC, the QTMT partition scheme allo...

AI Technical Summary

Problems solved by technology

These methods are only applicable to the quadtree partitioning scheme in H.265 / HEVC, and do not take into account the QTMT partitioning schemes currently adopted by VVC including quadtree, vertical binary tree, horizontal binary tree, vertical ternary tree or horizontal ternary tree division. So these methods no longer work

In addition, some scholars have proposed QTBT (Quadtree plus binary tree) fast partitioning algorithms based on convolutional neural networks (Convolutional Neural Networks, CNN) or machine learning (Machine Learning, ML), but these algorithms are only for quadtree and binary tree partitioning. , without considering the division of vertical ternary tree or horizontal ternary tree, the applicability is not strong

Images