Detection and identification of network carbide defects in steel based on improved CAM-YOLOx-DeepSORT algorithm

doi:10.13251/j.issn.0254-6051.2024.06.044

Abstract

Abstract: In the detection and identification of network carbides defects in mold steels, algorithms based on image processing and machine learning are commonly used, which can achieve automation to a certain extent, while the identification accuracy still has limitations in cases where the number of network carbides is large and the morphology is complex. To address the limitations of low recognition rate mentioned above, an improved CAM-YOLOx-DeepSORT algorithm is proposed, which utilizes the CAM attention mechanism and combines with the YOLOx object detection algorithm and DeepSORT object tracking algorithm to realize automatic detection and identification of the network carbides in mold steels. The results show that the algorithm can efficiently and accurately detect network carbides,the detection accuracy reaches 99.1%, thus providing guidance for quality control of mold steels.

Key words: network carbides, mold steel, image processing, detection and identification, YOLOx

CLC Number:

TG142.1

Huo Jinliang, Cai Ying, Tong Haisheng. Detection and identification of network carbide defects in steel based on improved CAM-YOLOx-DeepSORT algorithm[J]. Heat Treatment of Metals, 2024, 49(6): 282-286.

References

[1]高志喆, 陈小艳, 王永金, 等. 热处理对复合铸造高铬高碳钢/碳钢耐磨材料微观组织及力学性能的影响[J]. 金属热处理, 2023, 47(8): 163-167.
Gao Zhizhe, Chen Xiaoyan, Wang Yongjin, et al. Effect of heat treatment on microstructure and mechanical properties of composite cast high chromium high carbon steel/carbon steel wear-resistant materials[J]. Heat Treatment of Metals, 2023, 47(8): 163-167.
[2]崔兰芳, 李洪, 周洪涛. 压铸铝合金变速箱壳体应力热裂分析[J]. 铸造, 2016, 65(8): 795-797.
Cui Lanfang, Li Hong, Zhou Hongtao. Failure analysis on stress hot tearing of casting aluminum alloy transmission box[J]. Foundry, 2016, 65(8): 795-797.
[3]廖霞. 82B盘条网状碳化物快速检验方法研究[J]. 世界有色金属, 2018, 513(21): 219-220.
Liao Xia. Study on rapid inspection method of 82B wire rod network carbide[J]. World Nonferrous Metals, 2018, 513(21): 219-220.
[4]李慧慧. 基于深度学习的金属表面瑕疵识别[D]. 桂林: 桂林电子科技大学, 2021.
Li Huihui. Recognition of metal surface defects based on deep learning[D]. Guilin: Guilin University of Electronic Science and Technology, 2021.
[5]邵先鑫. 基于机器视觉的金属表面缺陷检测的设计与研究[D]. 沈阳: 中国科学院大学(中国科学院沈阳计算技术研究所), 2022.
Shao Xianxin. Design and research of metal surface defect detection based on machine vision[D]. Shenyang: University of Chinese Academy of Sciences (Shenyang Institute of Computing Technology, Chinese Academy of Sciences), 2022.
[6]Jiang P, Ergu D, Liu F, et al. A review of Yolo algorithm developments[J]. Procedia Computer Science, 2022, 199: 1066-1073.
[7]Panboonyuen T, Thongbai S, Wongweeranimit W, et al. Object detection of road assets using transformer-based YOLOx with feature pyramid decoder on Thai highway panorama[J]. Information, 2021, 13(1): 5-10.
[8]Yi C, Xu B, Chen J, et al. An improved YOLOX model for detecting strip surface defects[J]. Steel Research International, 2022, 93(11): 2200505.
[9]Charuchinda P, Kasetkasem T, Kumazawa I, et al. On the use of class activation map for land cover mapping[C]//2019 16th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON). IEEE, 2019: 653-656.
[10]Lin L, He H, Xu Z, et al. Realtime vehicle tracking method based on YOLOv5+DeepSORT[J]. Computational Intelligence and Neuroscience, 2023, 10(6): 7974201.
[11]Zheng Z, Wang P, Liu W, et al. Distance-IoU loss: Faster and better learning for bounding box regression[C]//Proceedings of the AAAI Conference on Artificial Intelligence. 2020: 12993-13000.