Deep Learning Based Vehicle Detection and Classification Methodology Using Strain Sensors under Bridge Deck

Ma, Rujin; Zhang, Zhen; Dong, Yiqing; Pan, Yue

doi:10.3390/s20185051

Cited by 35 publications

(11 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…with minimal capital consumption and time consumption before traveling [18]. It is found that the research on the intelligent recommendation of tourism information in academic circles began at the beginning of this century [19]. Figure 1 describes the basic flow of user personalization model creation.…”

Section: Personalized Recommendationsmentioning

confidence: 99%

Personalized Recommendation of Rural Tourism Based on Traffic Classification and User Data Analysis

Wang

Zhang

2022

Security and Communication Networks

View full text Add to dashboard Cite

Recently, With the development of global agricultural industrialization, modern agriculture has the functions of improving the quality of the ecological environment and providing people with the functions of sightseeing, leisure, and vacation. How to take the user as the center and combine the user’s personalized characteristics to offer the rural tourism products they need has become a research problem with real-world application value and challenge. In this context, this study develops an intelligent recommendation model by extensively analyzing the contents of rural tourism information platforms and product recommendation factors, as well as a rough set algorithm and traffic classification. To minimize the attributes of rural tourism product information and extract the core attribute, an attribute reduction approach based on a different matrix is implemented. Moreover, user interest similarity is computed and ranked to recommend rural tourist products. In addition, a personalized tourism attraction recommendation model is presented based on geographic area and period. The model achieved the highest average accuracy of 0.87%. The relevant experimental test results reveal that the system can provide accurate recommendations and services for rural tourism products.

show abstract

Section: Personalized Recommendationsmentioning

confidence: 99%

Personalized Recommendation of Rural Tourism Based on Traffic Classification and User Data Analysis

Wang

Zhang

2022

Security and Communication Networks

View full text Add to dashboard Cite

show abstract

“…In particular, the shadow is a common barrier that makes it difficult to correctly identify backgrounds and objects [5]. The presence of a shadow can lead to a *Corresponding Author Email: hfarsi@birjand.ac.ir (H. Farsi) recognition error, and the object can be identified as a larger object [1,6].…”

Section: Introductionmentioning

confidence: 99%

Shadow Removal in Vehicle Detection Using ResUNet-a

Dorrani

Farsi

Mohamadzadeh

2023

IJEE

View full text Add to dashboard Cite

In traffic monitoring for video analysis systems, vehicle shadows have a negative effect on their performance. Shadow detection and removal are essential steps in accurate vehicle detection. In this paper, a new method is proposed for shadow detection using a novel convolution neural network architecture. In the proposed method, the edges of the image are first extracted. Edge extraction reduces calculation, and accelerates the execution of the method. The background of the frame is then removed and the main features are extracted using the ResUNet-a architecture. This architecture consists of two parts: the encoder and the decoder, which detect the shadow at the decoder output and then remove it. Deep learning is used to detect shadows, which increases the accuracy of the analysis. The ResUNet-a architecture can learn complex, hierarchical, and appropriate features from the image for accurate feature detection and discarding the irrelevant shadow, thereby outperforming conventional filters.The results show that the proposed method provides better performance on NJDOT traffic video, highway-1, and highway-3 datasets than popular shadow removal methods. Also, the method improves the evaluation criteria such as F-measure and runtime. The F-measure is 94 and 93% for highway-1 and highway-3, respectively.

show abstract

“…Wi-Fi channel state information [14] has also been used to provide similar non-contact traffic monitoring. Strain measurements [15] and vibration sensing [16] have been shown to provide vehicle detection and classification information.…”

Section: Introductionmentioning

confidence: 99%

Detecting Vehicle Loading Events in Bridge Rotation Data Measured with Multi-Axial Accelerometers

Ferguson

Woods

Hester

2022

Sensors

View full text Add to dashboard Cite

Structural Health Monitoring (SHM) is critical in the observation and analysis of our national infrastructure of bridges. Due to the ease of measuring bridge rotation, bridge SHM using rotation measurements is becoming more popular, as even a single DC accelerometer placed at each end of span can accurately capture bridge deformations. Event detection methods for SHM typically entail additional instrumentation, such as strain gauges or continuously recording video cameras, and thus the additional cost limits their utility in resource-constrained environments and for wider deployment. Herein, we present a more cost-effective event detection method which exploits the existing bridge rotation instrumentation (tri-axial MEMS accelerometers) to also act as a trigger for subsequent stages of the SHM system and thus obviates the need for additional vehicle detection equipment. We show how the generalised variance over a short sliding window can be used to robustly discriminate individual vehicle loading events, both in time and magnitude, from raw acceleration data. Numerical simulation results examine the operation of the event detector under varying operating conditions, including vehicle types and sensor locations. The method’s application is demonstrated for two case studies involving in-service bridges experiencing live free-flow traffic. An initial implementation on a Raspberry Pi Zero 2 shows that the proposed functionality can be realised in less than 400 ARM A32 instructions with a latency of 47 microseconds.

show abstract

Deep Learning Based Vehicle Detection and Classification Methodology Using Strain Sensors under Bridge Deck

Cited by 35 publications

References 48 publications

Personalized Recommendation of Rural Tourism Based on Traffic Classification and User Data Analysis

Personalized Recommendation of Rural Tourism Based on Traffic Classification and User Data Analysis

Shadow Removal in Vehicle Detection Using ResUNet-a

Detecting Vehicle Loading Events in Bridge Rotation Data Measured with Multi-Axial Accelerometers

Contact Info

Product

Resources

About