一种对于疲劳驾驶的评估及预警方法

摘要

This paper proposes a fatigue driving safety level assessment and early warning system named KEEP-COOL, which is based on multi-sensor information fusion. This system can be mainly divided into three components:Safety degree appraising and early warning decision scheme;Fatigue assessment method based on the driver's facial features;Traffic condition assessment of vehicles based on lane modeling. By combining on the results of analysis, the system is able to react differently with various safety levels.%本文提出了一种基于多传感信息融合的KEEP-COOL疲劳驾驶安全等级评估及预警方法。主要分为以下几个部分：安全等级的划分方法及预警决策方案、基于驾驶员面部特征的驾驶员疲劳度评估方法、基于车道线数学模型分析的车辆行驶状态评估方法。结合各项检测结果，首先划分出驾驶安全等级，然后根据不同的安全等级做出相应的预警及控制反应。