摘要
农忙时节,农机跨区作业时经常出现驾驶员长时间驾驶农机作业导致疲劳驾驶的现象,易引发伤人、损物事故,造成人员伤亡、农机损坏。为了提高农机驾驶员疲劳检测的效率和准确性,降低因疲劳驾驶引发的事故风险,本研究对已有文献进行了广泛回顾,主要内容包括生理指标监测、行为表现分析以及使用高科技辅助工具进行疲劳预测等。研究结果表明,农机驾驶员疲劳检测应逐渐向更精确的卷积神经网络方向发展,该方向无需对驾驶员进行任何身体接触,符合人体舒适度需求,同时也能提高疲劳检测的准确性。而如何在检测准确性和实时性之间进行合理权衡,将是未来需要探索的重要问题,需要努力寻找最优方案,以确保该疲劳检测技术能够在实际场景中高效运行,并为驾驶员提供准确可靠的疲劳预警。
During the busy farming season,it is common for drivers to drive agricultural machinery across different areas for extended periods of time,resulting in fatigue driving.This can easily lead to injury and damage accidents,resulting in casualties and damage to agricultural machinery.In order to improve the efficiency and accuracy of fatigue detection for agricultural machinery drivers and reduce the risk of accidents caused by fatigue driving,this study extensively reviewed existing literature,including physiological indicator monitoring,behavioral performance analysis,and the use of high-tech auxiliary tools for fatigue prediction.The research results indicate that fatigue detection for agricultural machinery drivers should gradually develop towards more accurate convolutional neural networks,which do not require any physical contact with the driver,meet the needs of human comfort,and also improve the accuracy of fatigue detection.How to strike a reasonable balance between detection accuracy and real-time performance will be an important issue that needs to be explored in the future.Efforts need to be made to find the optimal solution to ensure that the fatigue detection technology can operate efficiently in practical scenarios and provide accurate and reliable fatigue warning for drivers.
作者
卢奥玮
董增
李珊辉
Lu Aowei;Dong Zeng;Li Shanhui(School of Information Engineering,Tarim University,Xinjiang Alar 843300)
关键词
卷积神经网络
疲劳检测
深度学习
convolutional neural network
fatigue testing
deep learning
