University of Twente Student Theses
The development of a real-time monitoring system for fatigue detection on truckers
Brugman, S.R.D. (2022) The development of a real-time monitoring system for fatigue detection on truckers.
PDF
27MB |
Abstract: | Fatigue is a major determinant in traffic accidents. Normalization and automation of modern vehicles only increase the urgency for the development of an advanced driver assistant system (ADAS) that reliably can detect the driver’s fatigue state. In this thesis, an ADAS is proposed based on the Viola-Jones Algorithm, and drowsiness metric. Haar-like feature-based cascade classifiers are combined with AdaBoost to locate the face and then extract relevant landmarks. The eye ratio aspect (EAR) is determined from the obtained landmarks. Contrary to conventional methods, a peak detection function is used on the EAR sequence for the identification of blinking patterns. Subsequently, the classification of the drivers uses the percentage of eyelid closure (PERCLOS), blink frequency, and entropy to classify the fatigue state. The proposed system is evaluated using a metric evaluation on the Eyeblink8 dataset, achieving a satisfactory level of precision (0.839%) and recall (0.893%). Additionally, the user evaluation demonstrated the state-of-the-art and real-time performance accomplished by the proposed system. |
Item Type: | Essay (Bachelor) |
Clients: | Techspread, Enschede, Netherlands |
Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
Subject: | 50 technical science in general, 54 computer science |
Programme: | Creative Technology BSc (50447) |
Link to this item: | https://purl.utwente.nl/essays/92146 |
Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Repository Staff Only: item control page