AI Driven Eyeball Tracked Wheelchair
Riza Mariya Abraham, Rosmi Thomas, Jobson Sebastian, Merene Joseph · 2023
Quadriplegia is a condition of complete paralysis that affects a person. The main cause of quadriplegia is an injury to the spinal cord of the human body. Various techniques are designed to help the disabled with complete paralysis to carry out their daily activities easily and one of these is the smart wheelchair. So we propose an electric wheelchair that can be controlled by tracking the movement of eyeball. Four gazes of the eyeball are tracked to control the proposed wheelchair. The left, right and center position of eyeball controls the leftward, rightward and forward motion of the wheelchair and forced blinking of eye gives the stop signal to the wheelchair. The right or left oriented continuous eyeball movement controls the backward motion of wheelchair. Firstly, the camera captures the real-time image and sends to Raspberry Pi. The Raspberry Pi process the image using computer vision and machine learning techniques and the output signal is transmitted to the Bluetooth module in the wheelchair. Then the signal is detected by the A Tmega1608 and the corresponding instruction is given to the motor drive. Accordingly the gear motor moves in accordance with the programmed direction. OpenCV and MediaPipe library are used to detect facial landmarks and track the eyeball movement from the captured frames. This AI enabled wheelchair tracking system will outperform the existing techniques.