부산대학교 도서관

This article proposes a system that aids people with disabilities. An Electric Eye Controlled Wheelchair System is built to help disabled people. With the designed system, disabled people can move effortlessly without support from others. The system uses image acquisition wherein the image of the eye is processed to find out the gaze direction of the eye using Haar cascade and gaze estimation algorithms and hence wheelchair moves according to the direction of eyeball movement. The gaze estimation algorithm is so precise and one single algorithm does the job of what two algorithms (Canny Edge detection, Hough Transform) are supposed to do and to execute the same task. With this technique, a disabled person can steer their wheelchair with their eye movement. The webcam is placed in Infront of the person which captures the live movements, and an image processing technique is used to track the position of the pupil in both eyes with the help of a raspberry pi processor. The image processing technique used here is Gaze tracking which uses Open CV. The gaze tracking tracks pupil movement and depicts its coordinates. According to pupil motion, the motor driver will be instructed to go forward, left, and right. A blink instruction is used to stop the wheelchair when the person blinks. Additionally, a front-mounted ultrasonic sensor that can detect obstructions and automatically halt wheelchair movement is mounted for safety reasons. The system is monitored by a Raspberry Pi device, which lowers the cost.