A method, device and storage device for tracking, recording and broadcasting based on gestures

Abstract

The invention discloses a method, device and storage device for tracking, recording and broadcasting according to gestures. The method includes the steps of: performing deep learning on gesture pictures in advance to generate a trained gesture detection network model; The image is calculated using the trained gesture detection network model, and the gesture classification corresponding to the image to be detected is obtained, and the corresponding tracking recording and broadcasting control is performed according to the corresponding gesture classification. The present invention first establishes a gesture detection network model through deep learning, and then uses the trained gesture detection network model to calculate the real-time collected image to be detected, and obtains the gesture classification corresponding to the image to be detected, so that it can be recorded in real time according to the different characteristics of the recorded person. Adjusting the camera lens with gestures for tracking and recording not only reduces the cost of tracking and recording, but also greatly improves the accuracy of tracking and recording.

Description

technical field [0001] The present invention relates to the technical field of tracking recording and broadcasting, in particular to a method, device and storage device for tracking recording and broadcasting according to gestures. Background technique [0002] The recording and broadcasting system is to integrate and record the video, audio, and image signals of electronic equipment recorded on the spot to generate standardized streaming media files for external live broadcast, storage, post-editing, and on-demand broadcasting. When the recording and broadcasting system performs tracking recording and broadcasting, how to adjust the recording and broadcasting strategy according to the recorded behavior is very important. Most of the existing tracking recording and broadcasting is to detect the height by adding auxiliary cameras to determine whether the recorded person (such as a student) There is a standing action to adjust the camera accordingly, which is not only expensiv...

AI Technical Summary

Problems solved by technology

[0004] In view of the above-mentioned deficiencies in the prior art, the purpose of the present invention is to provide a method, device and storage device for tracking recording and broadcasting based on gestures, thereby overcoming the high cost of existing tracking recording and broadcasting methods, which are prone to errors, are not smart enough, and cannot The problem of adjusting the camera lens according to the different gestures of the recorded person

Images