Scrap Steel Classification Model Based on Intelligent Algorithms

DOI: https://doi.org/10.62517/jbdc.202601133

Author(s)

Yuan Shuo, Guo Zhiming, Su Zhenxiao, Zhang Jiayu

Affiliation(s)

School of Artificial Intelligence, North China University of Technology, Tangshan, Hebei, China

Abstract

As a key green resource in the steel industry, the classification and grading quality of scrap steel directly affect enterprise costs and product quality. To address issues such as low efficiency, strong subjectivity, and high risk of traditional manual inspection, this paper proposes an intelligent scrap steel classification and grading method based on deep learning. the SeaFormer model is used to segment the wagon area, suppressing interference from complex backgrounds; a coordinate attention mechanism is embedded in YOLOv11 to enhance positioning accuracy and robustness in occluded and stacked scenarios; additionally, a slice-assisted super-inference framework is combined to improve small object detection in high-resolution images. Experimental results show that the CA-YOLOv11 achieves an mAP and F1 score of 87.6% and 87.3%, respectively, representing improvements of 1.3% and 0.6% over the baseline model. It performs excellently in detection accuracy, generalization ability, and real-time performance, providing a feasible solution for the intelligent recycling of scrap steel.

Keywords

Scrap Steel Classification; Machine Vision; Coordinate Attention Mechanism

References

[1] Ma Feiyue. Analysis and Prospect of the Current Situation of Waste Steel Classification and Grading [J]. Environmental Engineering, 2023, 41(S2):1163-1166. [2] Xiao Pengcheng, Yin Jihu, Wang Chao, et al. Accurate Identification of Dense Small Scrap Material Types Based on YOLO-SNBP [J]. China Metallurgy, 2025, 35(07):216-227. [3] Zhang Junxia, Xu Lingxiao, Ye Jinping, et al. Design of a Scrap Steel Impurity Recognition System Based on Machine Vision [J]. Metallurgical Equipment, 2025, (05):37-40. [4] Zhang, C.; Zhu, L.; Yu, L. Review of Attention Mechanism in Convolutional Neural Networks. Comput. Eng. Appl. 2021, 57, 64–72. [5] Xu, W.; Xiao, P.; Zhu, L.; Zhang, Y.; Chang, J.; Zhu, R.; Xu, Y. Classification and rating of steel scrap using deep learning. Eng. Appl. Artif. Intell. 2023, 123, 106241. [6] Mei, Y. Research on Intelligent Classification of Scrap Based on Machine Vision and LIBS Technology. Ph. D. Thesis, University of Science and Technology Beijing, Beijing, China, 21 December 2020.