DEVELOPMENT OF DROWSINESS DETECTION TECHNOLOGY BASED ON COMMAND LINE INTERFACE APPLICATION
The high number of accidents in Indonesia, of which is caused by drivers who experience fatigue and drowsiness while driving. Drowsiness and fatigue can be caused by the number of activities carried out, lack of sleep, long duration of driving, and an unfit body condition. Various indicators that are objective and subjective to assess the level of drowsiness and fatigue have been developed. However, objective indicators that can detect driver fatigue through sleep levels in real-time have not been widely developed, especially in Indonesia. Therefore, this study will develop drowsiness detection technology based on an intelligent system application, using three parameters, namely eye blinking detection, yawning mouth duration (yawn detection), and head pose detection, which is designed to detect driver drowsiness by receiving input in the form of image capture from the face and providing an output in the form of a warning signal. Facial area detection uses the facial landmark method that works by allocating points on the face to determine the biological shape of the human face. In processing facial images, Dlib facial landmarks and Open CV will be used, with the Python programming language. This research will be divided into several stages of technology development. In the first stage, a command line interface (CLI) application will be developed to simplify the development process. The results of the first stage of research are technology based on the Command Line Interface Application. The accuracy test of this application uses a video dataset of four drivers, where the test results state that the CLI application has an accuracy rate of 92.25%, which means that this application has a good performance in detecting driver drowsiness.
Drowsiness, Drowsiness Detection Technology, Facial Landmark, Command Line Interface Application.