Action recognition method, man-machine interaction method and device and storage medium

Abstract

The invention discloses an action recognition method, a man-machine interaction method and device and a storage medium, and belongs to the technical field of computers. The method comprises the following steps: extracting a plurality of target windows from a target video; obtaining first action feature information of each target window, when the maximum first similarity among the plurality of acquired first action feature information and preset feature information is greater than a first preset threshold value, and the difference value between the first similarity and the second similarity isgreater than a second preset threshold, determining that the dynamic action contained in the first target window corresponding to the first similarity is a preset dynamic action. The embodiment of theinvention provides a scheme for identifying a preset dynamic action. Action recognition is performed on the video frame in the target window, the problem of recognition errors caused by the influencebetween adjacent target windows is avoided, preset dynamic actions contained in the target window are recognized, the limitation that only static actions can be recognized in the prior art is brokenthrough, the application range is expanded, and the flexibility is improved.

Description

technical field [0001] The present invention relates to the field of computer technology, in particular to an action recognition method, a human-computer interaction method, a device and a storage medium. Background technique [0002] Action recognition technology is a technology that analyzes and processes action images to understand the actions contained in action images. With the development of computer technology, action recognition technology is widely used in video surveillance, somatosensory games, virtual reality, Internet medical care and other fields. , has become a popular research direction. [0003] In related technologies, the user makes a certain action within the shooting range of the electronic device, and the electronic device shoots the user to obtain an action image, and extracts body key points in the action image, such as wrist key points, elbow key points or shoulder key points. According to the distribution of the extracted body key points, the angle...

AI Technical Summary

Problems solved by technology

[0004] The inventor believes that the above solution can only recognize static actions based on static action images, so the scope of application is relatively limited

Images