H-Beam Steel Detection from Images using YOLOV5 with Transfer Learning Technique
Article Sidebar
Main Article Content
Abstract
Effective warehouse management is very important for running a business. An important step is to check the stock in the warehouse compared to the stock numbers stored in the system. This process requires multiple counting staff to check the stock of all products in the warehouse. It takes a long time to work and is costly and may cause errors in the inspection. Therefore, our research objective is present a model for detecting H-section hot-rolled steel for stock checking. A pre-trained model, YOLOV5m, is chosen and the learned transfer is used which can detect H section steel with precision is 0.992, recall is 0.822 and F1 score is 0.899.
Article Details
References
Xiaohu, Liu & Jineng, Ouyang. (2018). Research on steel bar detection and counting method based on contours. 294-297. 10.1109/ELTECH.2018.8401470.
Ghazali, Muhammad Faiz & Wong, Lai Kuan & See, John. (2017). Automatic Detection and Counting of Circular and Rectangular Steel Bars. 10.1007/978-981-10-1721-6_22.
Fan, Z., Lu, J., Qiu, B., Jiang, T., An, K., Josephraj, A. and Wei, C., 2019. Automated Steel Bar Counting and Center Localization with Convolutional Neural Networks. [online] arXiv.org. Available at: https://arxiv.org/abs/1906.00891
Hernández-Ruiz, Ana & Martinez-Nieto, Alejandro & Buldain Pérez, Julio David. (2021). Steel Bar Counting from Images with Machine Learning. Electronics. 10. 10.3390/electronics10040402.
Shin, Y.; Heo, S.; Han, S.; Kim, J.; Na, S. An Image-Based Steel Rebar Size Estimation and Counting Method Using a Convolutional Neural Network Combined with Homography. Buildings 2021, 11, 463. https://doi.org/10.3390/buildings11100463
Yang, Han & Fu, Chenghua. (2019). Quantity Detection of Steel Bars Based on Deep Learning. OALib. 06. 1-9. 10.4236/oalib.1105784.
Y. Zhu, C. Tang, H. Liu and P. Huang, "End-Face Localization and Segmentation of Steel Bar Based on Convolution Neural Network," in IEEE Access, vol. 8, pp. 74679-74690, 2020, doi: 10.1109/ACCESS.2020.2989300.
Li, Yang & Lu, Yujie & Chen, Jun. (2021). A deep learning approach for real-time rebar counting on the construction site based on YOLOv3 detector. Automation in Construction. 124. 103602. 10.1016/j.autcon.2021.103602.
W. Zhang, C. Zhu, G. E. O. Widera (1996). On the use of the upper-bound method for load determination in H-beam rolling. Journal of Materials Processing Technology Volume 56, Issues 1–4, January 1996, Pages 820-833.
Ben Guan, Yong Zang, Diping Wu, and Qin Qin (2017). Stress-inheriting behavior of H-beam during roller straightening process. Journal of Materials Processing Technology Volume 244, June 2017, Pages 253-272.