Vision-Based Deep Learning Algorithm for Detecting Potholes

Research output: Journal contributionsConference article in journalResearchpeer-review

Authors

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.

Original languageEnglish
Article number012019
JournalJournal of Physics: Conference Series
Volume2162
Number of pages11
ISSN1742-6588
DOIs
Publication statusPublished - 25.01.2022
EventInternational Conference on Applied Physics, Simulation and Computing - APSAC 2021 - Salerno, Italy
Duration: 03.09.202105.09.2021
Conference number: 5

Documents

DOI

Recently viewed

Publications

  1. Kinematic self-calibration of non-contact five-axis measuring machine using improved genetic algorithm.
  2. Binary Random Nets I
  3. Using learning protocols for knowledge acquisition and problem solving with individual and group incentives
  4. Homogenization modeling of thin-layer-type microstructures
  5. Closed-loop control of product geometry by using an artificial neural network in incremental sheet forming with active medium
  6. Performance and Comfort when Using Motion-Controlled Tools in Complex Tasks
  7. A simple fuzzy controller for robot manipulators with bounded inputs
  8. Microstructural development of as-cast AM50 during Constrained Friction Processing: grain refinement and influence of process parameters
  9. Latent structure perceptron with feature induction for unrestricted coreference resolution
  10. A change of values is in the air
  11. Analysis of Complexity Reduction in Kalman Filters Through Decoupling Control With Chattered Inputs in PMSM
  12. Volume of Imbalance Container Prediction using Kalman Filter and Long Short-Term Memory
  13. Improving short-term academic performance in the flipped classroom using dynamic geometry software
  14. A two-stage Kalman estimator for motion control using model predictive strategy
  15. A reference architecture for the integration of EMIS and ERP-Systems
  16. How to support teachers to give feedback to modelling tasks effectively? Results from a teacher-training-study in the Co²CA project
  17. Interrogating the Phenomenon of Suffering and Smiling by Nigerians: A Mixed Methods Study
  18. In situ synchrotron diffraction of the solidification of Mg4Y3Nd
  19. Stakeholder involvement for Water Framework Directive implementation in Germany
  20. Front in the mouth, front in the word
  21. Old Wine in New Bottles? The Case of Self-compassion and Neuroticism
  22. Die Liebe der Soziologie
  23. Antigott
  24. The use of force against terrorists
  25. Geometrical Accuracy in Two-Stage Incremental Sheet Forming with Active Medium
  26. Komparative studier i børnelitterature
  27. (Un)sichtbare Botschaften