Development of an application for interaction with a computer using gestures using Computer Vision technologies

Abstract

The work investigates methods for contactless computer control using hand gestures based on computer vision and machine learning technologies. Examples of the application of technologies in various areas of life are given, where they show superior efficiency and convenience over traditional interfaces for interacting with the system. The tools and instruments used in the development of the software application are described. A modular architecture of the system is proposed that provides real-time gesture recognition, including image segmentation, tracking of key hand points and classification of movements. Functionality for moving the cursor, performing left and right clicks, double clicks, scrolling pages and adjusting the volume is implemented. The result of the work is a software product that supports moving the mouse cursor, left and right clicks, double clicks, dragging objects, scrolling and adjusting the volume. Both the functioning of the main application commands and the operation on different hardware configurations in real time were tested. Solutions for further improvement and expansion of the software functionality were proposed.