Vision-Based Human Tracking Control of a Wheeled Inverted Pendulum Robot

Area of Science:

• Robotics
• Control Systems
• Computer Vision

Background:

• Wheeled inverted pendulum (WIP) robots require sophisticated control for stability and navigation.
• Tracking dynamic human targets presents challenges in robotics due to unpredictable movements.

Purpose of the Study:

• To design a vision-based adaptive control system for a WIP robot to track a moving human target.
• To enhance human target detection and tracking robustness using multisensor fusion.

Main Methods:

• Integration of OptiTrack and Kinect cameras for improved human detection and tracking.
• Development of robust adaptive control for WIP robot balancing and target following.
• Combination of leader-follower control, dynamic balance control, and visual tracking strategies.

Main Results:

• The proposed system achieves robust and efficient human target detection and tracking.
• The WIP robot successfully maintains balance while following the human target using visual feedback.
• Experimental studies validate the effectiveness of the integrated control strategies.

Conclusions:

• The developed vision-based adaptive control system enables effective human target tracking for WIP robots.
• Multisensor fusion enhances the robustness and performance of robotic systems in dynamic environments.
• The combined control approach successfully addresses both balancing and tracking requirements.