|
DOI: 10.14489/vkit.2024.03.pp.030-036
Васильев М. Е., Коськин А. В., Шалимов А. С.
АВТОМАТИЗАЦИЯ ОБНАРУЖЕНИЯ ДЕФЕКТОВ ПОВЕРХНОСТЕЙ ИЗДЕЛИЙ НА ОСНОВЕ СВЕРТОЧНЫХ НЕЙРОННЫХ СЕТЕЙ
(с. 30-36)
Аннотация. Рассмотрено применение сверточных нейронных сетей в целях автоматизации обнаружения дефектов поверхностей изделий. Основной фокус исследования сосредоточен на алгоритме обнаружения объектов сверточных нейронных сетей YOLOv5 и его улучшенной версии YOLOv5s-KEB. Эти алгоритмы предлагают эффективное решение для точного и быстрого распознавания дефектов, что позволяет производителям сократить время анализа изделий и повысить качество своей продукции.
Ключевые слова: автоматизация; обнаружение дефектов; поверхности изделий; алгоритм YOLOv5; сверточные нейронные сети.
Vasilev M. E., Koskin A. V., Shalimov A. S.
AUTOMATED DEFECT DETECTION ON PRODUCT SURFACES BASED ON CONVOLUTIONAL NEURAL NETWORKS
(pp. 30-36)
Abstract. This study is devoted to the development and analysis of models for detecting defects on product surfaces using convolutional neural networks, such as YOLOv5 and YOLOv5s-KEB. These models are based on the use of convolutional layers, recurrent layers and attention mechanisms, which allows them to effectively detect various types of defects such as cracks, gaps, pits and scratches. The study compared various models including YOLOv5s, YOLOv5s-KEB and their modifications such as Model A, Model B and Model C. The study found that using k-means generated anchors resulted in significant improvement in inference time and accuracy of defect detection. Other methods for improving models, such as the use of the BiFPN module and the use of the ECA mechanism, were also considered. The results of the study can be useful for developing more efficient and accurate models for detecting defects on product surfaces in the manufacturing industry. They can help improve product quality control and improve product safety and reliability. In addition, the results of the study can be used to create new methods for detecting defects on product surfaces, which can be more efficient and accurate than existing methods.
Keywords: Automation; Defect detection; Product surfaces; YOLOv5 algorithm; Convolutional neural networks.
М. Е. Васильев, А. В. Коськин, А. С. Шалимов (Орловский государственный университет имени И.С. Тургенева, г. Орел, Россия) E-mail:
Этот e-mail адрес защищен от спам-ботов, для его просмотра у Вас должен быть включен Javascript
M. E. Vasilev, A. V. Koskin, A. S. Shalimov (Oryol State University named after I. S. Turgenev, Orel, Russia) E-mail:
Этот e-mail адрес защищен от спам-ботов, для его просмотра у Вас должен быть включен Javascript
1. Girshick R., Donahue J., Darrell T., Malic J. Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation // Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2014. P. 580–587.
2. Girshick R. Fast R-CNN. ArXiv // ArXiv Preprint ArXiv:1504.08083. 2015.
3. Ren S., He K., Girshick R., Sun J. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks // Advances in Neural Information Processing Systems. 2015. V. 28.
4. Redmon J., Diwala S., Girshick R., Farhadi A. You only Look once: Unified, Real-Time Object Detection // Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2016. P. 779–788.
5. Ssd: Single Shot Multibox Detector / W. Liu, D. Anguelov, D. Erhan et al. // Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11 – 14, 2016, Proceedings, Part I 14. Springer International Publishing, 2016. P. 21–37.
6. Do Vision Transformers See Like Convolutional Neural Networks? / M. Raghu, T. Unterthiner, S. Kornblith et al. // Advances in Neural Information Processing Systems. 2021. V. 34. P. 12116–12128.
7. Kou X., Liu S., Cheng K., Qian Y. Development of a YOLO-V3-Based Model for Detecting Defects on Steel Strip Surface // Measurement. 2021. V. 182. P. 109454.
8. Intelligent System for Estimation of the Spatial Position of Apples Based on YOLOv3 and Real Sense Depth Camera D415 / N. Andriyanov, I. Khasanshin, C. Zhang et al. // Symmetry. 2022. V. 14. No. 1. P. 148.
9. Tulbure A. A., Tulbure A. A., Dulf E. H. A Review on Modern Defect Detection Models Using DCNNs–Deep Convolutional Neural Networks // Journal of Advanced Research. 2022. V. 35. P. 33–48.
10. Du X., Chen J., Zhang H., Wang J. Fault Detection of Aero-Engine Sensor Based on Inception-CNN // Aerospace. 2022. V. 9. No. 5. P. 236.
11. Li X., Wang C., Ju H., Li Z. Surface Defect Detection Model for Aero-Engine Components Based on Improved YOLOv5 // Applied Sciences. 2022. V. 12. No. 14. P. 7235.
12. Redmon J., Farhadi A. Yolo9000: Better, Faster, Stronger ArXiv Preprint // ArXiv Preprint ArXiv. 2017. V. 1612.
13. Redmon J., Farhadi A. Yolov3: An Incremental Improvement // ArXiv Preprint ArXiv:1804.02767. 2018.
14. Bochkovskiy A., Wang C. Y., Liao H. Y. M. Yolov4: Optimal Speed and Accuracy of Object Detection // ArXiv Preprint ArXiv:2004.10934. 2020.
15. Path Aggregation Network for Instance Segmentation / S. Liu, L. Qi, H. Qin et al. // Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018. P. 8759–8768.
16. Feature Pyramid Networks for Object Detection / T. Y. Lin, P. Dollar, R. Girshick et al. // Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2017. P. 2117–2125.
17. Ромащенко М. А., Васильченко Д. В., Пухов Д. А. Использование нейросетевых алгоритмов для визуального контроля топологии печатных плат // Вестник Воронежского государственного технического университета. 2022. Т. 18. № 3. С. 78–82.
18. Вахидова К. Л., Минцаева М. Ш., Исаева М. Р., Пашаев В. В. Алгоритм распознавания основных дефектов поверхностей колец подшипников с применением метода фрактальной размерности и нейронных сетей // Инженерный вестник Дона. 2022. № 3(87). С. 124–132.
1. Girshick R., Donahue J., Darrell T., Malic J. (2014). Rich feature hierarchies for accurate object detection and semantic segmentation. Proceedings of the IEEE conference on computer vision and pattern recognition, 580 – 587.
2. Girshick R. (2015). Fast r-cnn. arXiv. arXiv preprint arXiv:1504.08083.
3. Ren S., He K., Girshick R., Sun J. (2015). Faster r-cnn: Towards real-time object detection with region proposal networks. Advances in neural information processing systems, 28.
4. Redmon J., Diwala S., Girshick R., Farhadi A. (2016). You only look once: Unified, real-time object detection. Proceedings of the IEEE conference on computer vision and pattern recognition, 779 – 788.
5. Liu W., Anguelov D., Erhan D. (2016). Ssd: Single shot multibox detector. Computer Vision–ECCV 2016: 14th European Conference, Part I, 14, 21 – 37. Amsterdam, The Netherlands, October 11 – 14, 2016, Proceedings, Part I 14. Springer International Publishing, 2016. P. 21–37.
6. Raghu M., Unterthiner T., Kornblith S. (2021). Do vision transformers see like convolutional neural networks? Advances in Neural Information Processing Systems, 34, 12116 – 12128.
7. Kou X., Liu S., Cheng K., Qian Y. et al. (2021). Development of a YOLO-V3-based model for detecting defects on steel strip surface. Measurement, 182.
8. Andriyanov N., Khasanshin I., Zhang C. (2022). Intelligent system for estimation of the spatial position of apples based on YOLOv3 and real sense depth camera D415. Symmetry, 14(1).
9. Tulbure A. A., Tulbure A. A., Dulf E. H. (2022). A review on modern defect detection models using DCNNs–Deep convolutional neural networks. Journal of Advanced Research, 35, 33 – 48.
10. Du X., Chen J., Zhang H., Wang J. (2022). Fault detection of aero-engine sensor based on inception-CNN. Aerospace, 9(5).
11. Li X., Wang C., Ju H., Li Z. (2022). Surface defect detection model for aero-engine components based on improved YOLOv5. Applied Sciences, Vol. 12 14.
12. Redmon J., Farhadi A. (2017). Yolo9000: better, faster, stronger arXiv preprint, 1612.
13. Redmon J., Farhadi A. (2018). Yolov3: An incremental improvement. ArXiv preprint arXiv:1804.02767. 2018.
14. Bochkovskiy A., Wang C. Y., Liao H. Y. M. (2020). Yolov4: Optimal speed and accuracy of object detection. ArXiv preprint arXiv:2004.10934.
15. Liu S., Qi L., Qin H. (2018). Path aggregation network for instance segmentation. Proceedings of the IEEE conference on computer vision and pattern recognition, 8759 – 8768.
16. Lin T. Y., Dollar P., Girshick R. (2017). Feature pyramid networks for object detection. Proceedings of the IEEE conference on computer vision and pattern recognition, 2117 – 2125.
17. Romashchenko M. A., Vasil'chenko D. V., Puhov D. A. (2022). Using neural network algorithms for visual inspection of printed circuit board topology. Vestnik Voronezhskogo gosudarstvennogo tekhnicheskogo universiteta, 18(3), 78 – 82. [in Russian language]
18. Vahidova K. L., Mintsaeva M. Sh., Isaeva M. R., Pashaev V. V. (2022). Algorithm for recognizing the main defects in the surfaces of bearing rings using the fractal dimension method and neural networks. Inzhenerniy vestnik Dona, 87(3), 124 – 132. [in Russian language]
Статью можно приобрести в электронном виде (PDF формат).
Стоимость статьи 500 руб. (в том числе НДС 20%). После оформления заказа, в течение нескольких дней, на указанный вами e-mail придут счет и квитанция для оплаты в банке.
После поступления денег на счет издательства, вам будет выслан электронный вариант статьи.
Для заказа скопируйте doi статьи:
10.14489/vkit.2024.03.pp.030-036
и заполните форму
Отправляя форму вы даете согласие на обработку персональных данных.
.
This article is available in electronic format (PDF).
The cost of a single article is 500 rubles. (including VAT 20%). After you place an order within a few days, you will receive following documents to your specified e-mail: account on payment and receipt to pay in the bank.
After depositing your payment on our bank account we send you file of the article by e-mail.
To order articles please copy the article doi:
10.14489/vkit.2024.03.pp.030-036
and fill out the form
.
|
Разработка концепции и создание сайта - ООО «Издательский дом «СПЕКТР»