摘要
针对驾驶员疲劳状态检测问题,本文提出了基于红外图像处理和生理特征—心率的全天候疲劳检测算法,采用模糊神经网络专家系统对驾驶员的疲劳状态识别。假设驾驶员驾驶汽车的初始阶段(前十分钟内)是清醒的,这样在前十分钟内,模糊神经网络处于学习阶段并记忆驾驶员的状态,在十分钟之后模糊神经网络处于离线自学习,在线对驾驶员状态实时识别。通过实验表明该检测方法克服了光线和气候的影响,该识别方法具有较强的自适应能力。
To solve driver fatigue detection problem, an all-weather fatigue detection method is proposed based on infrared image processing and physiological feature - heart rate, which uses fuzzy neural network expert system to discern driver fatigue status. The discerning method assumes that the driver is in clear status in initial ten minutes when the fuzzy neural network is in learning status, which remembers the driver status. After ten minutes, the fuzzy neural network begins to offline self-learn and online real-time discern the driver status. Experiments show that the detection method overcomes the effects of light and weather and the discerning method has good self-adaptive ability.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2009年第3期636-640,共5页
Chinese Journal of Scientific Instrument
基金
国家教育部博士点专项基金(200806141056)
电子科技大学青年科技基金(L08010701)X0763资助项目
关键词
红外图像处理
心率
疲劳检测
糊神经网络
infrared image processing
physiological feature
fatigue detection
fuzzy neural network
