A Comparative Study of Object Detection Systems: A Case Study on Detecting Surface Defects on Simulated Automotive Bodies after Spray Painting
Keywords:
Deep learning, YOLO, Image analysis, Defect detection, Body surface inspectionAbstract
This study evaluates the performance of the YOLOv12s model for detecting defects on automotive body surfaces after painting. The defects were based on real production process characteristics and simulated in the lab with 75 pieces, which were photographed and modified to create 150 images for training and testing the model. YOLOv12s achieved an average precision of 0.886. The study also compares YOLOv12s with YOLOv5s and commercial deep learning software like CiRA and Zebra Aurora Vision. The evaluation used precision, recall, accuracy, and F1-score as metrics. YOLOv12s had the highest recall rate of 95.2% and an F1-score of 88.9%, outperforming open-source alternatives and showing performance comparably to CiRA. Zebra Aurora Vision, however, demonstrated an interesting overall performance, with an F1-score of 95.8%. All models showed limitations in detecting small defects or those obscured by the coating or uneven paint. The study concludes that YOLOv12s has strong potential as a cost-effective, efficient alternative for body surface defect detection, with further development needed for real-time production line use.
References
A. Saberironaghi, J. Ren, and M. ElGindy, “Defect detection methods for industrial products using deep learning techniques: a review,” Algorithms, vol. 16, no. 2, p. 95, February. 2023.
W. Liqun, W. Jiansheng, and W. Dingjin, “Research on vehicle parts defect detection based on deep learning,” in 2 nd International Symposium on Big Data and Applied Statistics (ISBDAS2019), Dalian, 2019, pp. 1-11.
H.-I. Lin and F. S. Wibowo, “Image data assessment approach for deep learning-based metal surface defectdetection systems,” IEEE Access, vol. 9, pp. 47621-47638, March. 2021.
Y.-T. Chang, W. K. T. M. Gunarathne, and T. K. Shih, “Deep learning approaches for dynamic object understanding and defect detection,” Journal of Internet Technology, vol. 21, no. 3, pp. 783-790, May. 2020.
Z. Jiang, X. Hu, and S. Wang, “Image classification of car paint defect detection based on convolutional neural networks,” in the 2 nd International Conference on Robotics, Automation and Intelligent Control (ICRAIC 2022), Changsha, 2022, pp. 1-8.
J. Zhang, J. Xu, L. Zhu, K. Zhang, T. Liu, D. Wang, and X. Wang, “An improved MobileNet-SSD algorithm for automatic defect detection on vehicle body paint,”Multimedia Tools and Applications, vol. 79, pp. 23367-23385, June. 2020.
F. Chang, M. Liu, M. Dong, and Y. Duan, “A mobile vision inspection system for tiny defect detection on smooth carbody surfaces based on deep ensemble learning,” Measurement Science and Technology, vol. 3 0, no. 12, p. 125905, September. 2019.
F. Chang, M. Dong, M. Liu, L. Wang, and Y. Duan, “A lightweight appearance quality assessment system based on parallel deep learning for painted car body,” IEEE Transactions on Instrumentation and Measurement, vol. 69, no. 8, pp. 5298-5307, August. 2020.
I. Konovalenko, P. Maruschak, V. Brevus, and O. Prentkovskis, “Recognition of scratches and abrasions on metal surfaces using a classifier based on a convolutional neural network,” Metals, vol. 11, no. 4, p. 549, March. 2021.
H. Huang and K. Zhu, “Automotive parts defect detection based on YOLOv7,” Electronics, vol. 13, no. 10, p. 1817, May. 2024.
X. Tao, D. Zhang, W. Hou, W. Ma, and D. Xu, “Industrial weak scratches inspection based on multifeature fusion network,” IEEE Transactions on Instrumentation and Measurement, vol. 70, p. 5000514, September. 2020.
L. Yao, Y. Wan, M. Yao, and B. Xu, “Detecting fine scratches on smooth surfaces with multiscale wavelet representation,” Measurement Science and Technology, vol. 2 3 , no. 8 , p. 085603, June. 2012.
H. Marko, L. Jelečević, and G. Gledec, “A comparative study of YOLOv5 models performance for image localization and classification,” in Proceedings of the Central European Conference on Information and Intelligent Systems, Dubrovnik, 2022, pp. 349-356.
R. Usamentiaga, D. G. Lema, O. D. Pedrayes, and D. F. Garcia, “Automated surface defect detection in metals: a comparative review of object detection and semantic segmentation using deep learning,” IEEE Transactions on Industry Applications, vol. 5 8, no. 3 , pp. 4203-4213, May-June. 2022.
Y. Tian, Q. Ye, and D. Doermann, “YOLOv12: Attention-centric real-time object detectors,” 2 0 2 5 , arXiv: 2502.12524.
C C. K. Loo, S. Boonsang, T. Sasisaowapak, S. Chuwongin, T. Tongloy, S. Nahavandi, and K. W. Wong, “CiRA CORE: a low code platform that makes AI work for industry 4.0,” in 2024 IEEE International Conference on Systems, Man, and Cybernetics (SMC), Sarawak, 2024, pp. 1187-1188.
C. Chousangsuntorn, T. Tongloy, S. Chuwongin, and S. Boonsang, “Classification model of optical character recognition failures in unrecovered slider serial numbers in hard disk drive manufacturing and image capture processes,” in Thirteenth International Conference on Signal Processing Systems (ICSPS 2021), Shanghai, 2022.
R. Alonso, G. Noce, V. Cutrona, and D. R. Recupero, “Extending asset lifespan through data augmentation-assisted quality control,” IFAC-PapersOnLine, vol. 58, no. 8, pp. 73-78, September. 2024.
P. Prusaczyk, W. Kaczmarek, J. Panasiuk, and K. Besseghieur, “Integration of robotic arm and vision system with processing software using TCP/IP protocol in industrial sorting application,” in Proceedings of the 15th Conference on Computational Technologies in Engineering, Jora Wielka, 2019, pp. 020032-1-020032-8.
A. Boonrod, P. Piyaprapaphan, N. Kittipongphat, D. Theerakulpisut, and A. Boonrod, “Deep learning for osteoporosis screening using an anteroposterior hip radiograph image,” European Journal of Orthopaedic Surgery & Traumatology, vol. 3 4, pp. 3045-3051, August. 2024.
J. Nalepa, M. Czardybon, and M. Walczak, “Real-time lung segmentation from whole-body CT scans using Adaptive Vision Studio: a visual programming software suite,” in RealTime Image and Video Processing 2018, Orlando, 2018.
S. Jantawong, K. Saninjak, T. Tikapunya, and J. Jomfong, “Image detection of GT pesticide test kit results in organic agricultural products using the CiRACORE deep learning model,” International Journal of Science and Innovative Technology, vol. 7, no. 2, pp. 86-99, October. 2024.
K. Radlak, M. Frackiewicz, M. Szczepanski, M. Kawulok, and M. Czardybon, “Adaptive Vision Studio – Educational tool for image processing learning,” in 2015 IEEE Frontiers in Education Conference (FIE), El Paso, 2015, pp. 1-8.
P. Domingos, “A few useful things to know about machine learning,” Communications of the ACM, vol. 55, no. 10, pp. 78-87, October. 2012.
I. Goodfellow, B. Yoshua, and A. Courville, Deep Learning, Cambridge, MA, USA: MIT Press, 2016.
C. Sun, A. Shrivastava, S. Singh, and A. Gupta, “Revisiting unreasonable effectiveness of data in deep learning era,” in 2017 IEEE International Conference on Computer Vision (ICCV), Venice, 2017, pp. 843-852.
P. Bergmann, M. Fauser, D. Sattlegger, and C. Steger, “MVTec AD — A comprehensive real-world dataset for unsupervised anomaly detection,” in 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Long Beach, 2019, pp. 9584-9592.
D. Krstinić, M. Braović, L. Šerić and D. Božić-Štulić, “Multi-label classifier performance evaluation with confusion matrix,” in International Conference on Soft Computing, Artificial Intelligence and Machine Learning (SAIM 2020) , Copenhagen, 2020, pp. 1-14.
R. Islam, Z. H. Zamil, E. Rayed, M. Kabir, M. F. Mridha, S. Nishimura, and J. Shin, “Deep learning and computer vision techniques for enhanced quality control in manufacturing processes,” IEEE Access, vol. 1 2, pp. 121449-121479, September. 2024.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2025 Faculty of Industrial Technology, Suan Sunandha Rajabhat University
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
บทความที่ได้รับการตีพิมพ์เป็นลิขสิทธิ์ของคณะเทคโนโลยีอุตสาหกรรม มหาวิทยาลัยราชภัฎสวนสุนันทา
ข้อความที่ปรากฏในบทความแต่ละเรื่องในวารสารวิชาการเล่มนี้เป็นความคิดเห็นส่วนตัวของผู้เขียนแต่ละท่านไม่เกี่ยวข้องกับมหาวิทยาลัยราชภัฎสวนสุนันทา และคณาจารย์ท่านอื่นๆในมหาวิทยาลัยฯ แต่อย่างใด ความรับผิดชอบองค์ประกอบทั้งหมดของบทความแต่ละเรื่องเป็นของผู้เขียนแต่ละท่าน หากมีความผิดพลาดใดๆ ผู้เขียนแต่ละท่านจะรับผิดชอบบทความของตนเองแต่ผู้เดียว
