通过统计过程控制进行车辆驾驶员监测

Vehicle Driver Monitoring through the Statistical Process Control.

作者信息

Assuncao Arthur N, Aquino Andre L L, Câmara de M Santos Ricardo C, Guimaraes Rodolfo L M, Oliveira Ricardo A R

机构信息

Instituto Federal de Educação, Ciência e Tecnologia do Sudeste de Minas Gerais, Santos Dumont, MG 36240-000, Brazil.

Departamento de Computação, Universidade Federal de Ouro Preto, Ouro Preto, MG 35400-000, Brazil.

出版信息

Sensors (Basel). 2019 Jul 11;19(14):3059. doi: 10.3390/s19143059.

DOI:10.3390/s19143059

PMID:31336711

原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC6678967/

Abstract

This paper proposes the use of the Statistical Process Control (SPC), more specifically, the Exponentially Weighted Moving Average method, for the monitoring of drivers using approaches based on the vehicle and the driver's behavior. Based on the SPC, we propose a method for the lane departure detection; a method for detecting sudden driver movements; and a method combined with computer vision to detect driver fatigue. All methods consider information from sensors scattered by the vehicle. The results showed the efficiency of the methods in the identification and detection of unwanted driver actions, such as sudden movements, lane departure, and driver fatigue. Lane departure detection obtained results of up to 76.92% (without constant speed) and 84.16% (speed maintained at ≈60). Furthermore, sudden movements detection obtained results of up to 91.66% (steering wheel) and 94.44% (brake). The driver fatigue has been detected in up to 94.46% situations.

摘要

本文提出使用统计过程控制（SPC），更具体地说是指数加权移动平均法，来监测基于车辆和驾驶员行为的方法中的驾驶员。基于SPC，我们提出了一种车道偏离检测方法；一种检测驾驶员突然动作的方法；以及一种结合计算机视觉检测驾驶员疲劳的方法。所有方法都考虑了车辆分散的传感器信息。结果表明这些方法在识别和检测不必要的驾驶员行为方面是有效的，如突然动作、车道偏离和驾驶员疲劳。车道偏离检测在不保持恒定速度时的准确率高达76.92%，在速度保持在约60时的准确率为84.16%。此外，突然动作检测在方向盘方面的准确率高达91.66%，在刹车方面的准确率为94.44%。在高达94.46%的情况下检测到了驾驶员疲劳。