A Networked Yawn Detector
2003 (English)Report (Other academic)
The driver’s mental state could be estimated from visual clues. The typical driver’s fatigue event could be detected or predicted by the dynamic facial expression events such like yawn. This paper demonstrate a networked surveillance system, where the driver’s facial expression parameters are extracted from real time video of face in car and sent via wireless network to a surveillance center, where the parameters could be evaluated to find if the driver is under fatigue situation. The parameter extraction using the Model-based coding (MBC) technique. A Hidden Markov Model (HMM) is used for recognizing the yawn event which characterize a typical fatigue event.
A prototype of such a networked system was set up and subjected to user tests. Promising results from user tests and their subjective evaluations are reported.
Place, publisher, year, edition, pages
2003. , 14 p.
DML Technical Report, ISSN 1652-8441
Signalbehandling, Fatigue detection, Hidden Markov Model, Model-based coding (MBC).
IdentifiersURN: urn:nbn:se:umu:diva-426OAI: oai:DiVA.org:umu-426DiVA: diva2:143437