DRIVER DROWSINESS DETECTION SYSTEM

DRIVER DROWSINESS DETECTION SYSTEM

Ms. JAYA DHARSHINI.G

Student, Department of Computer Science and Engineering, Ramco Institute of Technology

ABSTRACT
A technology for detecting driver sleepiness is intended to warn them, preventing accidents. If the driver is fatigued or not, the detection mostly relies on the outcome of the face/eye detection or the pupil of the eye. The majority of earlier research make the assumption that the driver does not wear glasses. However, spectacles can readily affect eye detection, which lowers the proper detection ratio. In this study, a tiredness detection approach with eyeglass removal is suggested in order to counteract the impact of the spectacles. First, the OpenCV library's functions are used to identify the facial area. Then, utilising morphological processes, the spectacles are taken out. Following the removal of spectacles, the eye regions are found using the OpenCV library's functions, and they are tracked by utilising a template matching technique. The horizontal projection and Kalman filter are used to achieve the binarization result of the ocular area. Next, the open/closed state of the eyes is established, and based on the series state of the eyes, weariness is established. The effectiveness of the suggested strategy is assessed using four testing movies. The fatigue detection rate can approach 100%, while the average correct detection ratio for eye state is 88.5 percent. The proposed method is workable, according to the preliminary results.

Keywords—Face detection, Eye detection, Eye tracking, Kalman filter, Template matching