Abstract
Due to complex background interference and weak space–time connection, traditional driver fatigue detection methods perform poorly for open-pit truck drivers. For these issues, this paper presents a driver fatigue detection method based on Libfacedetection and an LRCN. The method consists of three stages: (1) using a face detection module with a tracking method to quickly extract the ROI of the face; (2) extracting and coding the features; (3) combining the coding model to build a spatiotemporal classification network. The innovation of the method is to utilize the spatiotemporal features of the image sequence to build a spatiotemporal classification model suitable for this task. Meanwhile, a tracking method is added to the face detection stage to reduce time expenditure. As a result, the average speed with the tracking method for face detection on video is increased by 74% in comparison with the one without the tracking method. Our best model adopts a DHLSTM and feature-level frame aggregation, which achieves high accuracy of 99.30% on the self-built dataset.
Highlights
Introduction on Video Sequences in OpenPitRecently, several sizeable open-pit truck accidents have aroused people’s attention to driver fatigue detection
This paper proposes an open-pit truck driver fatigue detection method based on an open-source library for face detection (Libfacedetection) and an long-term recurrent convolutional network (LRCN)
This paper proposes a network structure by combining Resnet and a double-hiddenlong longshort-term short-term memory neural network network (DHLSTM) to deeply explore the spatiotemporal features of driver fatigue
Summary
Introduction on Video Sequences in OpenPitRecently, several sizeable open-pit truck accidents have aroused people’s attention to driver fatigue detection. Open-pit trucks are among the most critical transportation equipment in surface mines [1]. Because of their high cost and huge size, once an accident occurs, it makes mining enterprises bear huge economic costs. Traditional methods classify behaviors based on the single frame information They only analyze features from the image level, such as by using convolutional neural networks (CNN) [3], template matching or binarization [4] to obtain status information of the target and identifying the state of fatigue by calculating the percentage of eyelid closure over the pupil over time (PERCLOS) [5,6] or the frequency of the mouth (FOM). Burcu and Yaşar [7] applied a multitask CNN model to get face characteristics and calculate the PERCLOS and the FOM to determine driver fatigue
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
