Eye Blink Controlled Low Cost Smart Wheel Chair Aiding Disabled People

In this paper, a smart wheelchair is proposed to help out physically disabled people by using their eye blink activities. A prototype is implemented by processing eye ball images of the subject to detect intentional eye blinks. An algorithm is proposed to produce wheelchair movement command from consecutive intentional eye blinks. To identify intentional eye blinks from eye ball images, advanced image processing techniques are applied using raspberry-pi. The commands are wirelessly transferred to the motor driver of the wheel chair. The HC-05 Buletooth receiver is connected to the wheelchair which receives transmitted data from raspberry-pi and motor driver implements the command according to predefined decision table. It can also sense obstacle using APDS 9960 proximity sensor to avoid collision with surroundings. The proposed smart wheelchair has shown False Command Rate (FCR) of 2.1 % to 13.2%. The experiment is performed on four subjects and each of them has attended three sessions. Overall, the eye blink controlled aiding wheelchair is going to create a humanitarian impact on physically disabled people.