STM Journals Peer-reviewed research from Consortium e-Learning Network

Journal of Polymer & Composites Original Research

Integrated DIPlib and OpenCV Framework for Precise Geometric Characterisation of Woven Fibre-Reinforced Polymer Composites

  1. Kanak Panchal Department of Electronics & Telecommunication, MIT Academy of Engineering, Pune
  2. Janhavi Pohnerkar Department of Electronics & Telecommunication, MIT Academy of Engineering, Pune
  3. Dipti Sakhare Department of Electronics & Telecommunication, MIT Academy of Engineering, Pune
  4. Trupti Bhalerao Department of Electronics & Telecommunication, MIT Academy of Engineering, Pune
  5. Usha Verma Department of Electronics & Telecommunication, MIT Academy of Engineering, Pune
  6. Vaishali Wangikar Department of Electronics & Telecommunication, MIT Academy of Engineering, Pune
  7. Abhijeet Deogirikar Department of Computer Engineering, MIT Academy of Engineering, Pune

Abstract

The mechanical properties of woven fibre-reinforced polymer (FRP) composites stem entirely from the geometrical regularity inherent in their reinforcement structure. Changes in the size of the unit cell, fibre tow separation, weave angle, and fibre tow spacing will have an immediate effect on the stiffness and shear modulus of the material. In this paper, a combined machine vision system that incorporates both the OpenCV and DIPlib libraries is proposed for the automatic detection and quantification of the diamond (or rhombus) unit cells of plain weave, twill weave, and biaxial woven FRP composites. Convex polygonal approximation with additional constraints on angles finds out rhombuses as the basic unit cell of the weave fabric, whereas the DIPlib approach for computing the Feret diameter measures major diagonal, minor diagonal, sides of the diamond, weave angle, and inter-tow spacing with sub-millimetre accuracy. QR-code-based calibration ensures traceability of pixel-to-millimetre conversion, along with full accounting of the uncertainty. The proposed framework is validated by processing high resolution image sequences of plain weave carbon fiber epoxy (CF/EP) and twill weave glass fiber epoxy (GF/EP) samples, which results in mean absolute deviations lower than 0.20 mm and repeatability standard deviations below 0.12 mm, meeting the required tolerances for manufacturing aerospace structures.

Keywords

References (21)

  1. Chavan P, Pujari S, Sakhare DY, Deogirikar A, Shewale D. Co-ordinate Measurement Machine Using Computer Vision. 2025 International Conference on Emerging Smart Computing and Informatics (ESCI). 2025:1-6. doi:10.1109/esci63694.2025.10988393
  2. Gonzalez RC, Woods RE. Digital Image Processing. 4th ed. Pearson; 2018.
  3. Ghosh S, Ahmed S. Machine learning method for defect detection and classification in industrial settings. Procedia Comput Sci. 2021;187:290–297. https://doi.org/10.1016/j.procs.2021.04.043
  4. Zhang W, et al. An overview of industrial image segmentation using deep learning. Int J Ind Inf. 2025;45(1):34–49. https://doi.org/10.1016/j.industrialinf.2025.1012
  5. Dutta N, Ray I. Machine learning-based automated image processing for quality control. IEEE Trans Ind Inf. 2019;15(10):5555–5564. https://doi.org/10.1109/TII.2019.2936573
  6. Luceno S, et al. DIPlib: A Library for Quantitative Image Analysis. https://diplib.org; 2023.
  7. DIPlib-OpenCV interface module documentation. https://diplib.org/diplib-docs/dip_opencv_interface.html; 2023.
  8. Luengo C. OpenCV is not designed for quantification. https://www.crisluengo.net/archives/1140/; 2023.
  9. Stack Overflow contributors. Measuring the distance between two lines using DipLib (PyDIP). https://stackoverflow.com/questions/70560162; 2022.
  10. Stack Overflow contributors. How to measure average thickness of segmented image using DipLib. https://stackoverflow.com/questions/73807919; 2024.
  11. OpenCV.org. Hybrid CV/DL pipelines with OpenCV 4.4 G-API. https://opencv.org/blog; 2020.
  12. Bradski G. The OpenCV Library. Dr Dobb’s Journal of Software Tools; 2000.
  13. Hendriks CLL, van Vliet LJ. DIPlib: A scientific image processing library. Delft University of Technology; 1999.
  14. Itseez. Open source computer vision library. https://github.com/opencv/opencv; 2015.
  15. Lomov SV, et al. Meso-FE modelling of textile composites: Road map, data flow and algorithms. Compos Sci Technol. 2007;67(9):1870–1891.
  16. Sutton MA, Orteu JJ, Schreier HW. Image Correlation for Shape, Motion and Deformation Measurements. Springer; 2009.
  17. Potluri P, Manan A. Mechanics of non-orthogonally interlaced textile composites. Compos Part A. 2007;38(4):1216–1226.
  18. Wang B, Zhong S, Lee TL, Fancey KS, Mi J. Non-destructive testing and evaluation of composite materials/structures: A state-of-the-art review. Advances in Mechanical Engineering. 2020;12(4):168781402091376. doi:10.1177/1687814020913761
  19. Zhang Z. A flexible new technique for camera calibration. IEEE Trans PAMI. 2000;22(11):1330–1334.
  20. Chavan P, et al. Coordinate Measurement Machine Using Computer Vision with ArUco/QR Fiducial Markers. ESCI 2025 Proceedings. IEEE; 2025.
  21. OpenCV.org. Open source computer vision library. https://github.com/opencv/opencv; 2015.
Support