Deep Deconvolution for Traffic Analysis With Distributed Acoustic Sensing Data

نویسندگان

چکیده

Distributed Acoustic Sensing (DAS) is a novel vibration sensing technology that can be employed to detect vehicles and analyse traffic flows using existing telecommunication cables. DAS therefore has great potential in future “smart city” developments, such as real-time incident detection. Though previous studies have considered vehicle detection under relatively light conditions, order for feasible real-world scenarios, algorithms need also perform robustly wide range of conditions. In this study we investigate the roadside simultaneous characterisation velocity individual vehicles. To improve temporal resolution accuracy, propose self-supervised Deep Learning approach deconvolves characteristic car impulse response from data, which refer Deconvolution Auto-Encoder (DAE). We show deconvolution data with our DAE leads better performance than original (non-deconvolved) data. subsequently apply 24-hour cycle, demonstrating feasibility proposed method process large volumes potentially near-real time.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Multicomponent imaging with distributed acoustic sensing

The usage of Distributed Acoustic Sensing (DAS) in geophysics is rapidly gaining popularity due to its dense spatial sampling and low operation cost if the optical fiber is easily accessible. In the borehole environment, optical fibers for DAS are often readily available as a part of other sensing tools, such as for temperature and pressure. Although DAS provides single axial strain measurement...

متن کامل

Traffic Analysis with Deep Learning

Recent advances in learning Deep Neural Network (DNN) architectures have received a great deal of attention due to their ability to outperform state-of-the-art classifiers across a wide range of applications, with little or no feature engineering. In this paper, we broadly study the applicability of deep learning to website fingerprinting. We show that unsupervised DNNs can be used to extract l...

متن کامل

TiO2 based surface acoustic wave gas sensor with modified electrode dimensions for enhanced H2 sensing application

The design and optimization of nanostructure-based surface acoustic wave (SAW) gas sensor is analyzed based on TiO2 sensing layer and modified electrode dimensions. The sensitivity of the gas sensor depends upon the type of sensing layer used and active surface area obtained by varying the aspect ratio. The performance of the sensor is observed from 0.1ppm to 100ppm concentration of ...

متن کامل

3D Deconvolution with Deep Learning

3D deconvolution is a challenging problem faced by astronomers and micrsoscopists. The goal of deconvolution algorithms is to reverse the effects of convolution on the observed data. Convolution can be understood as a blurring operation on the ground truth that results in a relatively blurred/unclear observation. It is challenging to model the exact convolution operation because of various fact...

متن کامل

High-resolution multicomponent distributed acoustic sensing

Distributed acoustic sensing (DAS) uses an optical fiber together with an interrogator unit to perform strain measurements. The usage of DAS in geophysics is attractive due to its dense spatial sampling and low operation cost if the optical fiber is freely accessible. In the borehole environment, optical fibers for DAS are often readily available as a part of other sensing tools, such as for te...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: IEEE Transactions on Intelligent Transportation Systems

سال: 2023

ISSN: ['1558-0016', '1524-9050']

DOI: https://doi.org/10.1109/tits.2022.3223084