Wheelchair simulator control with eye blink movements using image processing techniques

Image processing is a computer work that can be used in many industrial and technological applications, integrated with the computer. By using the numerical data of the images transferred to the digital environment, improved or modified images can be obtained. At the same time, object recognition processes can be performed by using image processing techniques. The image processing system can also be used for counterfeiting protection in face recognition systems. Detection of blinks is important in systems that monitor the attention of operators. In this study, the data obtained by separating eye and eyebrow elements from the face region determined by using image processing techniques, using OpenCv and Dlib libraries were tried to be processed. Software algorithms were created with the Python programming language. Blink detection processes obtained from filtering have been tested on people who do not have any discomfort in the eye area. The effects of light environments with different light Lux values on blinking were calculated with the LDR sensor developed with Arduino. At the same time, the effects of the angle of the light source on facial features and blinking were observed. The effects of camera angle and distance on the calculation of blinks were calculated. The data obtained as a result of long experimental studies was sent to the Raspberry Pi module and the control of the mini robot, which will replace the wheelchair, was provided with the prepared algorithms.