Download

Vision-Based Drowsiness Detection System Using Convolutional Neural Networks

Various studies indicate that one of the main reasons of road accidents are related to fatigue and driver drowsiness. It is a global problem which not only effect human lives, in the form of death and disability, but also has great impact on economy. The driver drowsiness detection system is a tool to monitor driver consciousness. The system aims to develop an algorithm to detect drowsiness by going in three phases: In the first phase we conducted data training through CNN algorithm .In the second phase driver’s face is identified through MTCNN al- gorithm. In phase third Identifying facial features through 68 face landmark. The final stage decided whether the driver is drowsy or not. When a driver detected to be fall a sleep, there will be a warning sound to alert the driver to save his and other people lives. Experimental results show that our system has an accuracy of about 99.4%.

Information

  • Students: Thekra Alshubrmi - Fai Alothman - Mona Alkhudhair
  • Supervisor: Dr. Shabana Habib
  • Research Specialization: Automated systems
  • Upload Date: 11/12/2021