An Intelligent Human-Unmanned Aerial Vehicle Interaction Approach in Real Time Based on Machine Learning Using...

Area of Science:

• Human-computer interaction
• Robotics
• Machine Learning

Background:

• Civilian applications of unmanned aerial vehicles (UAVs) are rapidly expanding, necessitating intuitive control methods.
• Controlling UAVs in three-dimensional space presents significant challenges for human interaction.
• Existing human-UAV interaction methods often lack real-time responsiveness and accuracy.

Purpose of the Study:

• To develop an intelligent, real-time human-UAV interaction approach using machine learning and wearable smart gloves.
• To create a robust system capable of accurately interpreting complex hand gestures for UAV command.
• To enhance the efficiency and speed of human-UAV collaboration in civilian contexts.

Main Methods:

• Design and implementation of two real-time wearable smart gloves for data acquisition.
• Application of machine learning algorithms for processing glove sensor data and classifying hand gestures.
• Development of a multi-mode command structure and task scheduling algorithm for seamless interaction.

Main Results:

• The system demonstrated high accuracy, achieving approximately 98% with 49,000 data points.
• Real-time processing achieved performance in the milliseconds range.
• Validation was performed using 25 distinct hand gestures from 20 individuals.

Conclusions:

• The proposed machine learning-based approach enables accurate and efficient real-time human-UAV interaction.
• Wearable smart gloves offer a viable interface for intuitive control of UAVs in civilian settings.
• The system's robustness and high accuracy pave the way for advanced human-UAV collaboration.