基于Yolov8算法的夜间疲劳驾驶检测研究
Research on Night Fatigue Driving Detection Based on Yolov8 Algorithm

夜间疲劳驾驶是导致交通事故的主要原因之一，它不仅威胁到驾驶员的安全，也危及公共安全。在驾驶员刚发生疲劳驾驶时进行实时检测并发出预警提醒，对于减少交通事故、保护人民生命财产具有重要意义。运用Yolov8模型对驾驶员的疲劳状态进行实时监测并做出预警，是一种有效的解决方案。本文基于Yolov8n模型，通过优化损失函数、增加MPDIoU来提升边界框回归精度，并收集数据集进行训练、验证和测试。研究结果表明，Yolov8n模型在夜间疲劳驾驶检测中表现出色，证明了Yolov8算法在夜间疲劳驾驶检测中的应用潜力。
Fatigue driving at night stands as a primary cause of traffic accidents, jeopardizing both the driver's safety and public safety. Real-time detection and immediate early warnings when drivers begin to exhibit signs of fatigue are crucial for reducing traffic accidents and safeguarding lives and property. Utilizing the YOLOv8 model for real-time monitoring and issuing early warnings for driver fatigue is an effective solution. This paper leverages the YOLOv8n model, optimizing the loss function and incorporating MPDIoU to enhance bounding box regression accuracy. Comprehensive datasets were collected for training, validation, and testing. The results demonstrate that the YOLOv8n model excels in detecting night-time drowsy driving, affirming the potential application of the YOLOv8 algorithm in this critical area.