IRO Journals
Journal of Electrical Engineering and Automation

Vision-Based Railway Wheel Profile Defect Detection System

Open Access
Published: 25 May, 2026
Pages: 155-167
  • Mari Mugesh Babu B. , Mari Mugesh Babu B.
    Student, Department of Electrical and Electronics Engineering, VV College of Engineering, Thoothukudi, India
    Student, Department of Electrical and Electronics Engineering, VV College of Engineering, Thoothukudi, India
  • Esakki Vel Raj S. , Esakki Vel Raj S.
    Student, Department of Electrical and Electronics Engineering, VV College of Engineering, Thoothukudi, India
    Student, Department of Electrical and Electronics Engineering, VV College of Engineering, Thoothukudi, India
  • Cruz Easterlin Raja M. , Cruz Easterlin Raja M.
    Student, Department of Electrical and Electronics Engineering, VV College of Engineering, Thoothukudi, India
    Student, Department of Electrical and Electronics Engineering, VV College of Engineering, Thoothukudi, India
  • Jegatheesh B. , Jegatheesh B.
    Student, Department of Electrical and Electronics Engineering, VV College of Engineering, Thoothukudi, India
    Student, Department of Electrical and Electronics Engineering, VV College of Engineering, Thoothukudi, India
  • Jeyaseeli S. Jeyaseeli S.
    Assistant Professor, Department of Electrical and Electronics Engineering, VV College of Engineering, Thoothukudi, India
    Assistant Professor, Department of Electrical and Electronics Engineering, VV College of Engineering, Thoothukudi, India
view PDF
view PDF

How to Cite

B., Mari Mugesh Babu, Esakki Vel Raj S., Cruz Easterlin Raja M., Jegatheesh B., and Jeyaseeli S. 2026. “Vision-Based Railway Wheel Profile Defect Detection System”. Journal of Electrical Engineering and Automation 8 (2): 155-67. https://doi.org/10.36548/jeea.2026.2.006.
ACM ACS APA ABNT Chicago Harvard IEEE MLA Turabian Vancouver AMA

Download Citation

Endnote/Zotero/Mendeley (RIS) BibTeX

Keywords

Railway Wheel Inspection
Laser Triangulation Profilometry
Computer Vision
HSV Colour Thresholding
Intensity-Weighted Centroid
RMS Deviation
Multi-Class Defect Classification
Real-Time Processing
ESP32

Abstract

Railway transport plays a vital role in the transportation of passengers and cargo; its safe operation depends greatly on the condition of railway wheels. Such wheel problems as sharp flange, thin flange, deep flange, hollow tyre, flat tyre, thin tyre, and reduced radius at the root of the flange result from prolonged wheel operation under high load conditions. These defects can cause vibration, increase in repair expenses, and derailment, therefore, regular inspection of railway wheel geometry is required. The purpose of this paper is to propose an automated visual detection system for railway wheel profile defects based on laser projection and image acquisition by a camera. This system works through laser line projection on the wheel surface and recording the deformation of the laser line through a USB camera. Using computer vision methods implemented in Python, a geometric profile is calculated and transformed into a waveform that is further compared with the standard one. Notable advancements in the technology include adaptive HSV brightness gating to achieve laser isolation, intensity-based centroid detection, frame averaging using 30 frames to reduce noise, and a rule-based multi-class classifier capable of detecting eight types of defects within a speed range of 15 to 18 frames per second. The system serves as an affordable and contactless solution for the identification of flaws on wheels.

References

  1. Lewis, R., R. S. Dwyer-Joyce, U. Olofsson, and R. I. Hallam. "Wheel Material Wear Mechanisms and Transitions." 14th International Wheelset Congress 2004, Orlando, USA.
  2. PHILLIPS, Anique, Anriëtte BEKKER, and Kristiaan SCHREVE. "Development of a Laser-Based Wheel Measurement System for Live Wear Parameter Detection." The Broadmoor Colorado Springs Co. 2025, USA.
  3. OpenCV: Image Processing in OpenCV. (n.d.). https://docs.opencv.org/4.x/d2/d96/tutorial_py_table_of_contents_imgproc.html
  4. Espressif Systems. (2021). ESP32 Technical Reference Manual Version 5.7. https://documentation.espressif.com/esp32_technical_reference_manual_en.pdf
  5. Ni, Yi-Qing, and Qiu-Hu Zhang. "A Bayesian Machine Learning Approach for Online Detection of Railway Wheel Defects Using Track-Side Monitoring." Structural Health Monitoring 2021, vol. 20, no. 4: 1536-1550.
  6. Das, Sumit Kumar. "Wheel Defect Detection with Advanced Machine Learning." International Journal for Research in Applied Science & Engineering Technology 2023, vol. 11, no. 11: 500-504.
  7. Zschiesche, Kira, Jana Seiler, Claudia Baulig, Martin Dambacher, Bogdan Galuska, Lukas Jäger, Johannes Pelz, Jonas Rombach, Andreas Sutorius, and Alexander Reiterer. "Laser-Based Mobile Railway Measurement Systems: An Overview." Multimodal Sensing and Artificial Intelligence for Sustainable Future 2025, 13570: 247-254.
  8. P. Weston, C. Roberts, G. Yeo, and E. Stewart, “Perspectives on Railway Track Geometry Condition Monitoring from In-Service Railway Vehicles,” Vehicle System Dynamics 2015, vol. 53, no. 7, 1063–1091.
  9. M. Molodova, Z. Li, A. Nu´nez, and R. Dollevoet, “Auto-˜ Matic Detection of Squats in Railway Infrastructure,” IEEE Transactions on Intelligent Transportation Systems 2014, vol. 15, no. 5, 1980–1990.