TY - JOUR
T1 - Acceleration-Based Deep Learning Method for Vehicle Monitoring
AU - Zhu, Yanjie
AU - Sekiya, Hidehiko
AU - Okatani, Takayuki
AU - Yoshida, Ikumasa
AU - Hirano, Shuichi
N1 - Funding Information:
Manuscript received November 10, 2020; revised May 17, 2021; accepted May 17, 2021. Date of publication May 20, 2021; date of current version July 30, 2021. This work was supported in part by the Tokyo Metropolitan Expressway Company Ltd., in part by the Shutoko Engineering Company Ltd., in part by the Highway Technology Research Center, and in part by Seiko Epson. The associate editor coordinating the review of this article and approving it for publication was Dr. Yen Kheng Tan. (Corresponding author: Yanjie Zhu.) Yanjie Zhu and Ikumasa Yoshida are with the Department of Urban and Civil Engineering, Tokyo City University, Tokyo 158-8557, Japan (e-mail: zhuyanjie2018@163.com).
Publisher Copyright:
© 2001-2012 IEEE.
PY - 2021/8/1
Y1 - 2021/8/1
N2 - An automatic vehicle monitoring system can provide supports not only for intelligent transportation systems, but also for bridge weigh-in-motion (BWIM) systems, which use structural response to identify vehicle weights. In this paper, we provide a vehicle monitoring solution for acceleration-based BWIM system, utilizing deep learning and wavelet transform methods. The monitoring task is divided into three subtasks, including valid sequence detection, valid axle location, and driving lane identification. In first procedure, a shallow convolutional neural network is trained using time-frequency spectrograms to discover valuable time series. After that, an adaptive wavelet transform method is employed to locate axles from each valid sequence. Finally, the driving lane can be determined by cross-comparing vibration responses. Comparing with solutions based on all-in-one deep networks, the proposed method is computationally efficient and has improved generalization capability owing to the three-step division of the task. Evaluation is conducted on a multi-lane highway bridge located in Tokyo. Results show that 97% of vehicles can be identified correctly. For all recognized vehicles, the accuracy of driving lane detection is 100%.
AB - An automatic vehicle monitoring system can provide supports not only for intelligent transportation systems, but also for bridge weigh-in-motion (BWIM) systems, which use structural response to identify vehicle weights. In this paper, we provide a vehicle monitoring solution for acceleration-based BWIM system, utilizing deep learning and wavelet transform methods. The monitoring task is divided into three subtasks, including valid sequence detection, valid axle location, and driving lane identification. In first procedure, a shallow convolutional neural network is trained using time-frequency spectrograms to discover valuable time series. After that, an adaptive wavelet transform method is employed to locate axles from each valid sequence. Finally, the driving lane can be determined by cross-comparing vibration responses. Comparing with solutions based on all-in-one deep networks, the proposed method is computationally efficient and has improved generalization capability owing to the three-step division of the task. Evaluation is conducted on a multi-lane highway bridge located in Tokyo. Results show that 97% of vehicles can be identified correctly. For all recognized vehicles, the accuracy of driving lane detection is 100%.
KW - Bridge weigh-in-motion
KW - MEMS accelerometer
KW - convolutional neural network
KW - vehicle monitoring
KW - wavelet transform
UR - http://www.scopus.com/inward/record.url?scp=85107233893&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85107233893&partnerID=8YFLogxK
U2 - 10.1109/JSEN.2021.3082145
DO - 10.1109/JSEN.2021.3082145
M3 - Article
AN - SCOPUS:85107233893
VL - 21
SP - 17154
EP - 17161
JO - IEEE Sensors Journal
JF - IEEE Sensors Journal
SN - 1530-437X
IS - 15
M1 - 9437183
ER -