Image encoding method, action recognition method and computer equipment

Abstract

The invention provides an image coding method, which comprises the following steps that: obtaining a plurality of groups of human skeleton data for executing a target action, each group of human skeleton data comprising joint point data when the target action is executed; Extracting a motion characteristic matrix corresponding to the plurality of groups of human skeleton data according to the joint point data in the plurality of groups of human skeleton data; And encoding the motion characteristic matrix to obtain a motion characteristic image. The motion feature matrixes corresponding to themultiple sets of human body skeleton data are extracted, and then the extracted motion feature matrixes corresponding to the multiple sets of human body skeleton data are encoded into the motion feature image. Because the data volume of the motion feature image is smaller than the data volume of the plurality of action feature vector sequences, the consumption of storage resources and computing resources is reduced.

Description

technical field [0001] The present disclosure relates to the technical field of artificial intelligence, and in particular to an image coding method, an action recognition method and computer equipment. Background technique [0002] With the development of artificial intelligence technology, action recognition technology based on artificial intelligence technology has a wide range of application scenarios in social life, including family child care robots, monitoring dangerous behaviors in public places, and human-computer interaction game development. Based on the motion recognition technology, by recognizing the user's motion, the user's dangerous motion can be warned in time to avoid dangerous events. [0003] In the process of recognizing the user's actions, it is necessary to collect multiple sets of human skeleton data through the acquisition device, and form the joint point features extracted from each set of human skeleton data into a sequence of action feature vecto...

AI Technical Summary

Problems solved by technology

[0004] However, due to the relatively large amount of human skeleton data, the number of multiple action feature vector sequences extracted from multiple sets of human skeleton data is also relatively large, resulting in a large consumption of storage resources and computing resources.

Images