Screen control system and method based on LeapMotion gesture recognition

Abstract

The invention relates to a LeapMotion gesture recognition-based screen control system and method, the system comprises a binocular camera, a LeapMotion device, a gesture recognition computer and a server which are connected in sequence; the server is connected to a screen computer, the screen computer is connected to a screen display, and the binocular camera is used for collecting a hand action image of a user; the LeapMotion equipment generates a corresponding 3D model according to the hand action image; the gesture recognition computer is used for performing gesture recognition on the 3D model and outputting a gesture action signal instruction; the server outputs a corresponding screen data picture skip instruction to the screen computer according to the gesture action signal instruction; and the screen computer controls switching of the data pictures on the screen display according to the screen data picture skipping instruction. Compared with the prior art, gesture action recognition can be quickly and accurately carried out on hand movement of a user, and the purpose of non-contact screen control is achieved.

Description

technical field [0001] The invention relates to the technical field of non-contact screen control, in particular to a screen control system and method based on LeapMotion gesture recognition. Background technique [0002] Shopping guide machines in shopping malls, bank inquiry machines, hospital self-service machines, tourist attraction guide machines and other equipment mostly use contact human-computer interaction screens, that is, "touch screens". After touching the screen with the user's hand, the touch screen is used as a physical medium, By capturing user gestures and gesture recognition, the switching of data screens on the screen can be controlled. However, in order to prevent the spread of infectious viruses through the touch screen, it is necessary to control the screen in a non-contact manner. [0003] The current non-contact gesture recognition mainly relies on data gloves and electromyographic signal analysis. Among them, data gloves have a high accuracy rate. C...

AI Technical Summary

Problems solved by technology

[0003] The current non-contact gesture recognition mainly relies on data gloves and electromyographic signal analysis. Among them, data gloves have a high accuracy rate. CyberGlove is a data sensor using an optical fiber sensor, which can accurately measure up to 22 joint angles. The official guarantees the error within 3 degrees, but the disadvantages are also obvious. The price is too expensive. Two gloves cost 300,000 yuan. If the budget is small, domestic brands that also use fiber optic sensors include WiseGlove, or Manes in the Netherlands and 5DT , the price is about 50,000 pieces, and its accuracy will be worse. In addition, there are thousands of data gloves in China, which use bending sensors or resistive sensors, and their accuracy is difficult to guarantee; EMG signal analysis is based on physiological Signal, by measuring the voltage signal generated at both ends of the muscle when the muscle contracts, and using deep learning methods to identify gesture intentions, but this electromechanical signal analysis method is still in the experimental research stage, and its practical application effect is unknown

Images