Vision-Based Deep Learning Algorithm for Detecting Potholes

Publikation: Beiträge in ZeitschriftenKonferenzaufsätze in FachzeitschriftenForschungbegutachtet

Standard

Vision-Based Deep Learning Algorithm for Detecting Potholes. / Gajjar, Kanushka; Niekerk, T. Van; Wilm, Thomas et al.
in: Journal of Physics: Conference Series, Jahrgang 2162, 012019, 25.01.2022.

Publikation: Beiträge in ZeitschriftenKonferenzaufsätze in FachzeitschriftenForschungbegutachtet

Harvard

APA

Vancouver

Gajjar K, Niekerk TV, Wilm T, Mercorelli P. Vision-Based Deep Learning Algorithm for Detecting Potholes. Journal of Physics: Conference Series. 2022 Jan 25;2162:012019. doi: 10.1088/1742-6596/2162/1/012019

Bibtex

@article{d82b640921de4768bbf6625d19feab9a,
title = "Vision-Based Deep Learning Algorithm for Detecting Potholes",
abstract = "Potholes on roads pose a major threat to motorists. Driving over a pothole has the potential to cause serious damage to a vehicle, which in turn may result in fatal accidents. Currently, many pothole detection methods exist. However, these methods do not utilize deep learning techniques to detect a pothole in real-time, determine the location thereof and display its location on a map. The success of determining an effective pothole detection method, which includes the aforementioned deep learning techniques, is dependent on acquiring a large amount of data, including images of potholes. Once adequate data had been gathered, the images were processed and annotated. The next step was to determine which deep learning algorithms could be utilized. Three different models, including Faster R-CNN, SSD and YOLOv3 were trained on the custom dataset containing images of potholes to determine which network produces the best results for real-time detection. It was revealed that YOLOv3 produced the most accurate results and performed the best in real-time, with an average detection time of only 0.836 s per image. The final results revealed that a real-time pothole detection system, integrated with a cloud and maps service, can be created to allow drivers to avoid potholes. ",
keywords = "Engineering",
author = "Kanushka Gajjar and Niekerk, {T. Van} and Thomas Wilm and Paolo Mercorelli",
year = "2022",
month = jan,
day = "25",
doi = "10.1088/1742-6596/2162/1/012019",
language = "English",
volume = "2162",
journal = "Journal of Physics: Conference Series",
issn = "1742-6588",
publisher = "IOP Publishing Ltd",
note = "International Conference on Applied Physics, Simulation and Computing - APSAC 2021 ; Conference date: 03-09-2021 Through 05-09-2021",

}

RIS

TY - JOUR

T1 - Vision-Based Deep Learning Algorithm for Detecting Potholes

AU - Gajjar, Kanushka

AU - Niekerk, T. Van

AU - Wilm, Thomas

AU - Mercorelli, Paolo

N1 - Conference code: 5

PY - 2022/1/25

Y1 - 2022/1/25

N2 - Potholes on roads pose a major threat to motorists. Driving over a pothole has the potential to cause serious damage to a vehicle, which in turn may result in fatal accidents. Currently, many pothole detection methods exist. However, these methods do not utilize deep learning techniques to detect a pothole in real-time, determine the location thereof and display its location on a map. The success of determining an effective pothole detection method, which includes the aforementioned deep learning techniques, is dependent on acquiring a large amount of data, including images of potholes. Once adequate data had been gathered, the images were processed and annotated. The next step was to determine which deep learning algorithms could be utilized. Three different models, including Faster R-CNN, SSD and YOLOv3 were trained on the custom dataset containing images of potholes to determine which network produces the best results for real-time detection. It was revealed that YOLOv3 produced the most accurate results and performed the best in real-time, with an average detection time of only 0.836 s per image. The final results revealed that a real-time pothole detection system, integrated with a cloud and maps service, can be created to allow drivers to avoid potholes.

AB - Potholes on roads pose a major threat to motorists. Driving over a pothole has the potential to cause serious damage to a vehicle, which in turn may result in fatal accidents. Currently, many pothole detection methods exist. However, these methods do not utilize deep learning techniques to detect a pothole in real-time, determine the location thereof and display its location on a map. The success of determining an effective pothole detection method, which includes the aforementioned deep learning techniques, is dependent on acquiring a large amount of data, including images of potholes. Once adequate data had been gathered, the images were processed and annotated. The next step was to determine which deep learning algorithms could be utilized. Three different models, including Faster R-CNN, SSD and YOLOv3 were trained on the custom dataset containing images of potholes to determine which network produces the best results for real-time detection. It was revealed that YOLOv3 produced the most accurate results and performed the best in real-time, with an average detection time of only 0.836 s per image. The final results revealed that a real-time pothole detection system, integrated with a cloud and maps service, can be created to allow drivers to avoid potholes.

KW - Engineering

UR - http://www.scopus.com/inward/record.url?scp=85124945075&partnerID=8YFLogxK

U2 - 10.1088/1742-6596/2162/1/012019

DO - 10.1088/1742-6596/2162/1/012019

M3 - Conference article in journal

AN - SCOPUS:85124945075

VL - 2162

JO - Journal of Physics: Conference Series

JF - Journal of Physics: Conference Series

SN - 1742-6588

M1 - 012019

T2 - International Conference on Applied Physics, Simulation and Computing - APSAC 2021

Y2 - 3 September 2021 through 5 September 2021

ER -

Dokumente

DOI

Zuletzt angesehen

Publikationen

  1. Optimized neural networks for modeling of loudspeaker directivity diagrams
  2. Model-based logistic controlling of converging material flows
  3. Trajectory-based computational study of coherent behavior in flows
  4. A Matlab/Simulink toolbox for inversion of local linear model trees
  5. Dynamically changing sequencing rules with reinforcement learning in a job shop system with stochastic influences
  6. Lyapunov stability analysis to set up a PI controller for a mass flow system in case of a non-saturating input
  7. Influence of Process Parameters and Die Design on the Microstructure and Texture Development of Direct Extruded Magnesium Flat Products
  8. Use of Machine-Learning Algorithms Based on Text, Audio and Video Data in the Prediction of Anxiety and Post-Traumatic Stress in General and Clinical Populations
  9. Combining a PI Controller with an Adaptive Feedforward Control in PMSM
  10. “Ideation is Fine, but Execution is Key”
  11. A Multilevel Inverter Bridge Control Structure with Energy Storage Using Model Predictive Control for Flat Systems
  12. Sliding mode and model predictive control for inverse pendulum
  13. Model predictive control for switching gain adaptation in a sliding mode controller of a DC drive with nonlinear friction
  14. Restoring Causal Analysis to Structural Equation ModelingReview of Causality: Models, Reasoning, and Inference (2nd Edition), by Judea Pearl
  15. Effects of diversity versus segregation on automatic approach and avoidance behavior towards own and other ethnic groups
  16. Understanding Low-Code Evolution, Adoption and Ecosystem for Software Development
  17. An observer for sensorless variable valve control in camless internal combustion engines
  18. The Low-Code Phenomenon: Mapping the Intellectual Structure of Research
  19. Evaluating a Bayesian Student Model of Decimal Misconceptions
  20. Operational integration of EMIS and ERP systems
  21. The identification of up-And downstream industries using input-output tables and a firm-level application to minority shareholdings
  22. Concepts
  23. Moving Towards Measuring Multifunctionality in Ecosystems: FieldScreen – A Mobile Positioning System for Non-Invasive Measurement of Plant Traits in Field Experiments
  24. Document assignment in multi-site search engines
  25. German Utilities and distributed PV