학술논문

An adaptive Driver Fatigue Identification Method Based on HMM
Document Type
Conference
Source
2021 5th CAA International Conference on Vehicular Control and Intelligence (CVCI) Vehicular Control and Intelligence (CVCI), 2021 5th CAA International Conference on. :1-6 Oct, 2021
Subject
Power, Energy and Industry Applications
Transportation
Adaptation models
Maximum likelihood estimation
Adaptive systems
Computational modeling
Hidden Markov models
Fatigue
Data models
driving fatigue
hidden markov model
adaptive model
driving characteristics
Language
Abstract
Driver's characteristics have an essential influence on the performance improvement of advanced driver assistance systems (ADAS), which is also a key issue that needs to be solved in the research and development of on-board systems. This paper proposed an adaptive driver fatigue identification method based on the Hidden Markov Model (HMM) to reduce the influence of different drivers on fatigue detection. The model herein was designed based on the fatigue dynamic generation characteristics and individual characteristics. Based on driver classification, the fatigue identification model database was developed, which provided a basis for the design of the adaptive fatigue detection model. Finally, an adaptive fatigue detection method based on maximum likelihood estimation is proposed. Test results show that the proposed method can effectively improve the adaptability of the detection system, and the fatigue recognition accuracy has more tremendous advantages.