Wheelchair Controlled by Eye Movement Detection Device

Roee Sudman 1 Or Koren 2 Ofer Ziv 3
1Afeka - Tel Aviv Academic College of Engineering, Israel
2Afeka - Tel Aviv Academic College of Engineering, Israel
3Afeka - Tel Aviv Academic College of Engineering, Israel

Background: People with disabilities who are paralyzed from the neck down (ALS, spinal cord injuries) need another person to drive them while they are seated in a wheelchair. This project allows them to drive their wheelchair independently.

Methods: The system consists of two main parts: The first is an eye movement detection system and the second is a motion system. Images of the user`s pupils are captured, then translated into commands aimed to move the motors of the system. The device activates an image processing algorithm and a supervised learning algorithm that creates and analyzes a training data and produces an inferred function, which can be used for mapping future pupil detections. The motion system is a unique robotic driving system for a standard wheelchair. The system is constructed from a main frame, attached to two hub motors and four caster wheels. When the platform is connected to the wheelchair, it can drive the wheelchair using differential method that allows wide maneuvering.

Results: A feasibility model was built. This model represents the goal of the project by controlling a standard wheelchair using the eye movement detection algorithm. The model performs the main functions, moving forward and turning right and left, according to the eye movement detection algorithm which was implemented in the controller of the electrical motors.

Conclusion: The feasibility model implements the ability to move in flat and sloping surfaces and the correlation between the mechanical part of the wheelchair and the eye movement detection algorithm was proved.

Powered by Eventact EMS