A HANDS-FREE HUMAN-MACHINE INTERFACE FOR ELECTRIC WHEELCHAIR CONTROL BASED ON FACE AND TORSO MOVEMENTS

As the global population ages, mobility impairments are becoming a significant challenge, increasing the need for innovative assistive technologies. Traditional electric wheelchairs require manual input, which may not be feasible for users with severe disabilities. This paper presents a novel hands-free Human-Machine Interface (HMI) designed to enhance electric wheelchair control using face and torso movements. The proposed system integrates a vision-based motion capture method with a shared control strategy that considers both the rider's. A stereovision setup detects and tracks head and torso movements, which are then processed to generate control commands for the wheelchair. The experimental results demonstrate the feasibility of this system, highlighting its potential to improve wheelchair maneuverability in confined indoor environments while providing an intuitive and non-invasive control method. Challenges related to lighting conditions and system robustness are also discussed, with future work aimed at refining control accuracy and adaptability for different users.