Autors: Petrović E., Simonović M., Jovanović D.D., Dimitrov, L. V., Nikolić V.
Title: Deep learning-based instance segmentation for detection of tire tread area
Keywords: Instance segmentation, Mask R-CNN, Tire tread inspection, Transfer learning

Abstract: V.The condition of tires is crucial for ensuring optimal vehicle performance and safety, with tread depth being one of the most critical parameters. Manual contact measurements of tire tread depth can be subjective and inconvenient. Automating the tire tread inspection process could improve efficiency, accuracy, and consistency, which would help ensure timely tire replacements and reduce the risk of traffic accidents. This study investigates a non-contact, visual inspection method for tire tread, focusing primarily on the automatic detection and segmentation of the tire's tread area. A Mask R-CNN approach with the ResNet50 backbone was used for tire detection and segmentation. The model was trained with 247 images and tested on 62 images, achieving a mean average precision (mAP) of 0.6081, based on the MS COCO metric. Features of the segmented tire areas were extracted using the Histogram of Oriented Gradients (HOG) and classified based on tread conditions. The results suggest that this method could be an efficient and accurate approach for tire tread inspection, and it holds significant potential for applications in industries like automotive and transportation. Overall, this work lays a foundation for future research, which will further develop and refine automated tire inspection systems.

References

  1. S. Marjanović, D. RadivojevRoad and road environment as a factor of road safety. In: Proceedings of 5th International Conference Road Safety in Local Community Republic of Srpska, Banja Luka.2016.
  2. J.A. Cowley, S. Kim, M.S. WogalterPeople do not identify tire aging as a safety hazard. In: Proc. Hum. Factors Ergon. Soc. Annu. Meet., 50, 2006. pp. 860–864..
  3. Wang, X.B., Li, A.J., Ci, Q.P., Shi, M., Jing, T.L., Zhao, W.Z., The study on tire tread depth measurement method based on machine vision. Adv. Mech. Eng., 11(4), 2019.
  4. HORIBA MIRA, About HORIBA MIRA. 〈https://www.horiba-mira.com/about-horiba-mira/〉 (accessed Mar. 10, 2023).
  5. ROSPA, Tyre tread depth guide. 〈https://www.rospa.com/rospaweb/docs/advice-services/road-safety/vehicles/tyre-tread-depth.pdf〉 (accessed Mar. 10, 2023).
  6. Chen, P., Shubinsky, G.D., Jan, K.H., Chen, C.A., Sidla, O., Poelzleitner, W., Inspection of tire tread defects using image processing and pattern recognition techniques. Vision, Sensors, and Control for Automated Manufacturing Systems, 1993, SPIE, 14–21.
  7. Guo, Q., Zhang, C., Liu, H., Zhang, X., Defect detection in tire X-ray images using weighted texture dissimilarity. J. Sens., 2016, 2016, e4140175, 10.1155/2016/4140175.
  8. A.A. Raj, S. Kaviarasu, K. Saran, F. Mohammednatheem CNN based tyre life prediction and defect identification system. In: Proc. of 6th International Conference on Communication and Electronics Systems (ICCES), Coimbatre, India, (2021) 1-5, https://doi: 10.1109/ICCES51350.2021.9488961.
  9. S.M. Mynul Karim, Y. Rahman, M.A. Hai, R. MahfuzaTire wear detection for accident avoidance employing convolutional neural networks. In: Proc. of 8th NAFOSTED Conference on Information and Computer Science (NICS), Hanoi, Vietnam, 2021. pp. 364-368, https://doi: 10.1109/NICS54270.2021.9701504.
  10. Zheng, Z., Shen, J., Shao, Y., Zhang, J., Tian, C., Yu, B., Zhang, Y., Tire defect classification using a deep convolutional sparse-coding network. Meas. Sci. Technol., 32, 2021, 055, 10.1088/1361-6501/abddf3401.
  11. Kiruthikaa, R., R, P., Samyuktha, R., Swathi, P., Vinush, S., Tire quality inspection system based on deep learning. Int. J. Eng. Technol. Manag. Sci. 7 (2023), 143–148, 10.46647/ijetms.2023.v07i04.022.
  12. S.Y. Huang, Y.C. Chen, J.K. WangMeasurement of tire tread depth with image triangulation. In: Int. Symp. Comput. Consum. Control (IS3C), 2016. pp. 303–306.
  13. I. Jovančević, A. Arafat, J.J. Orteu, T. SentenacAirplane tire inspection by image processing techniques. In: Proceedings of the 5th Mediterr. Conf. Embed. Comput. (MECO), 2016. 17679.
  14. Lee, S., et al. Classification of tire tread wear using accelerometer signals through an artificial neural network. J. Korean Soc. Ind. Converg. 23:2 (2020), 163–171.
  15. Kim, H.J., Han, J.Y., Lee, S., Kwag, J.R., Kuk, M.G., Han, I.H., Kim, M.H., A road condition classification algorithm for a tire acceleration sensor using an artificial neural network. Electronics, 9, 2020, 404.
  16. Petrović, E., Ristić Durrant, D., Simonović, M., Ćojbašić, Ž., Nikolić, V., Vision-based inspection of tire tread depth. Trans. Famena. 45 (2021), 19–28.
  17. Hafiz, A.M., Bhat, G.M., A survey on instance segmentation: state of the art. Int. J. Multimed. Inf. Retr. 9 (2020), 171–189.
  18. M. Shu, Deep learning for image classification on very small datasets using transfer learning, PhD diss., Iowa State University, 2019.
  19. K. He, G. Gkioxari, P. Dollar, R. Girshick, R.-C.N.N. MaskProc. IEEE Int. Conf. Comput. Vis., 2017. pp. 2961–2969..
  20. Piyathilaka, L., Preethichandra, D.M.G., Izhar, U., Kahandawa, G., Real-time concrete crack detection and instance segmentation using deep transfer learning. Eng. Proc., 2, 2020, 91.
  21. P. Jaikumar, R. Vandaele, V. OjhaTransfer learning for instance segmentation of waste bottles using Mask R-CNN algorithm. In: Proceedings of the Int. Conf. Intell. Syst. Des. Appl., 2020. pp. 140–149.
  22. COCO Dataset. 〈https://cocodataset.org/#home〉 (accessed Mar. 10, 2023).
  23. Everingham, M., Van Gool, L., Williams, C.K.I., Winn, J., Zisserman, A., The Pascal visual object classes (VOC) challenge. Int. J. Comput. Vis. 88 (2010), 303–338.
  24. H. Rezatofighi, N. Tsoi, J. Gwak, A. Sadeghian, I. Reid, S. SavareseGeneralized intersection over union: a metric and a loss for bounding box regression. In: Proceedings of the IEEE/CVF Conf Comput Vis Pattern Recognit. 2019. pp. 658-666..
  25. Plainsight, Sense Data Annotation. 〈https://plainsight.ai/platform/sense-data-annotation/〉 (accessed Mar. 27, 2025).
  26. 4SmartMachines, Intelligent Automation Lab. 〈https://www.4smartmachines.com/ialab〉 (accessed Mar. 27, 2025).
  27. GitHub, COCO Annotator. 〈https://github.com/jsbroks/coco-annotator#readme〉 (accessed Mar. 27, 2025).
  28. GitHub, VIA Annotation Tool. 〈https://github.com/nearkyh/via-1.0.5〉 (accessed Mar. 27, 2025).

Issue

Progress in Engineering Science, vol. 2, 2025, Albania, https://doi.org/10.1016/j.pes.2025.100080

Цитирания (Citation/s):
1. Filho W.M.M., Yoshimura S.L., Marques A.S., Sousa J.Z., Siqueira G., Alves M.A.L., Ferreira W.G., Application of Lean Manufacturing Tools to Reduce Defect Incidence in Tire Production, 2025, SAE Technical Papers, issue 0, DOI 10.4271/2025-36-0021, issn 01487191, eissn 26883627 - 2025 - в издания, индексирани в Scopus

Вид: статия в списание, публикация в реферирано издание, индексирана в Scopus