An End-to-End Steel Surface Defect Detection Approach via Fusing Multiple Hierarchical Features
Open accessThe DOI matches its Crossref record; 43 of the 53 references it cites were found in a registry or an index.
Abstract
A complete defect detection task aims to achieve the specific class and precise location of each defect in an image, which makes it still challenging for applying this task in practice. The defect detection is a composite task of classification and location, leading to related methods is often hard to take into account the accuracy of both. The implementation of defect detection depends on a special detection data set that contains expensive manual annotations. In this paper, we proposed a novel defect detection system based on deep learning and focused on a practical industrial application: steel plate defect inspection. In order to achieve strong classification ability, this system employs a baseline convolution neural network (CNN) to generate feature maps at each stage, and then the proposed multilevel feature fusion network (MFN) combines multiple hierarchical features into one feature, which can include more location details of defects. Based on these multilevel features, a region proposal network (RPN) is adopted to generate regions of interest (ROIs). For each ROI, a detector, consisting of a classifier and a bounding box regressor, produces the final detection results. Finally, we set up a defect detection data set NEU-DET for training and evaluating our method. On the NEU-DET, our method achieves 74.8/82.3 mAP with baseline networks ResNet34/50 by using 300 proposals. In addition, by using only 50 proposals, our method can detect at 20 ft/s on a single GPU and reach…
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“There are some shortcomings in manual testing, such as low confidence and high labor intensity.”
From A New Steel Defect Detection Algorithm Based on Deep Learning (2021) · cites this paper as [5]
In the citing paper
“most classification tasks can only find the defects with the highest confidence level in the defect category, resulting in imprecise classification results [3].”
From Steel Surface Defect Recognition: A Survey (2022) · cites this paper as [3]
In the citing paper
“there are many types of steel surface defects and some of these defects may overlap”
From Steel Surface Defect Recognition: A Survey (2022) · cites this paper as [3]
In the citing paper
“Using manual inspection of steel surface defects, it is difficult to distinguish the type of defect and accurately detect the location of the defect [2].”
From A deep learning model for steel surface defect detection (2024) · cites this paper as [2]
In the citing paper
“There are many categories of steel surface defects and the size of each defect category varies widely. The average size of each category in the NEU-DET [34] dataset is shown in Table1.”
From A deep learning model for steel surface defect detection (2024) · cites this paper as [34]
1 more claim about this paper
Create a free account to see every claim other papers make about this one.
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474 · cited 1×
“The third type is ResNet using residual blocks to make networks deeper without overfitting [23].”
MethodVery Deep Convolutional Networks for Large-Scale Image Recognition
Simonyan, K., & Zisserman, A. (2014) · arXiv · doi:10.48550/arxiv.1409.1556 · Cited by 114,148
ImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155 · cited 1×
“cessful networks are based on the sequential pipeline architecture [25], which establish the basic structure of CNN and prove the importance of depth of networks.”
BackgroundLeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2016.2577031 · Cited by 73,472 · cited 6×
“Next, RPN [14] is employed to generate region proposals”
BackgroundMethodInfluentialFaster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2015) · arXiv · doi:10.48550/arxiv.1506.01497 · Cited by 70,812 · cited 6×
“Next, RPN [14] is employed to generate region proposals”
BackgroundMethodInfluentialImageNet: A large-scale hierarchical image database
Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., & Fei-Fei, L. (2009) · 2009 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2009.5206848 · Cited by 65,333 · cited 1×
“As current research in transfer learning [15], the key to drive large networks is pretraining on ImageNet [22].”
BackgroundGradient-based learning applied to document recognition
Lecun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998) · Proceedings of the IEEE · doi:10.1109/5.726791 · Cited by 58,805 · cited 1×
“combined? There are two essential conditions: nonadjacent, because adjacent layers have highly local correlation [32], and”
BackgroundYou Only Look Once: Unified, Real-Time Object Detection
Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.91 · Cited by 53,803 · cited 1×
“The representative direct regression methods are YOLO [29] and SSD [30].”
MethodFully convolutional networks for semantic segmentation
Long, J., Shelhamer, E., & Darrell, T. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298965 · Cited by 38,834 · cited 2×
“The most famous region-based detectors are the “R-CNN family” [27] [28] [14].”
BackgroundMethod
Cited by
Items 401–500 of 1,338
Duan, B., Wang, D., Ma, Y., Wang, G., & Liu, H. (2025) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2025.110730 · Cited by 14
Ding, Z., Song, W., & Zhan, S. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3158989 · Cited by 16
Hierarchical multi-scale network for cross-scale visual defect detection
Tang, R., Liu, Z., Song, Y., Duan, G., & Tan, J. (2023) · Journal of Intelligent Manufacturing · doi:10.1007/s10845-023-02097-1 · Cited by 15
“Based on the NEU defect detection dataset (He et al., 2019) and COCO mAP metrics (Lin et al., 2014), HMSNet improves the mAP by a large margin (3.2 mAP) while maintaining a fast speed compared with the baseline.”
InfluentialMeDERT: A Metal Surface Defect Detection Model
Wang, C., & Xie, H. (2023) · IEEE Access · doi:10.1109/access.2023.3262264 · Cited by 21
“…the ‘‘Aluminum profile surface defects’’ dataset [10], ‘‘Magnetic tile’’ dataset [11], and ‘‘North-eastern University surface defect database for defect detection task (NEU-DET)’’ dataset [12] are compared with the mainstream targets detection models such as Faster RCNN, YOLOv4, YOLOv5, and DERT.”
MethodA Self-Supervised CNN for Particle Inspection on Optical Element
Hou, W. S., Tao, X., & Xu, D. (2021) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2021.3077994 · Cited by 13
“Existing defect detection methods [14]–[17] are based on transfer learning of ImageNet weights for small training samples.”
BackgroundMethodAn Online Pipeline Structural Health Monitoring Method Based on the Spatial Deformation Fitting
Yu, H., Chen, X., Ren, M., et al. (2021) · IEEE Transactions on Industrial Electronics · doi:10.1109/tie.2021.3101003 · Cited by 14
TLU-Net: A Deep Learning Approach for Automatic Steel Surface Defect DetectionPreprint
Damacharla, P., V., A. R. M., Ringenberg, J., et al. (2022) · doi:10.31224/2679 · Cited by 14
ODverse33: Is the New YOLO Version Always Better? A Multi Domain benchmark from YOLO v5 to v11Preprint
Jiang, T., & Zhong, Y. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5538699 · Cited by 28
Liu, H., Liu, S., Liu, Z., Niu, B., Xie, J., Luo, C., & Shi, Z. (2025) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2025.111645 · Cited by 19
Zheng, X., Liu, W., & Huang, Y. (2025) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2025.112179 · Cited by 18
An Explainable Laser Welding Defect Recognition Method Based on Multi-Scale Class Activation Mapping
Liu, T., Zheng, H., Bao, J., et al. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3148739 · Cited by 13
QCNet: query context network for salient object detection of automatic surface inspection
Sun, J., Yan, S., & Song, X. (2022) · The Visual Computer · doi:10.1007/s00371-022-02597-w · Cited by 15
“Besides, there are some methods using Superpixel [15] or CRF [16–18] as post-processing to preserve the object boundaries, and EGNet [55] introduced extra supervision specially designed for edge refinement.”
MethodAutomated Scrap Steel Grading via a Hierarchical Learning-Based Framework
Tu, Q., Li, D., Xie, Q., Dai, L., & Wang, J. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3206816 · Cited by 17
“In particular, the DL techniques have been applied to several industrial scenes [3], [4], [5], [6], [7], which provide practical feasibility for its application in scrap steel scenarios.”
MethodSteel Sheet Defect Detection using Feature Pyramid Network and RESNET
Uraon, P. K., Verma, A., & Badholia, A. (2022) · International Conference on Edge Computing and Applications (ICECAA) · doi:10.1109/icecaa55415.2022.9936318 · Cited by 15
“A manual examination system of defects inspection of steel surfaces paired with conventional vision systems [5]-[6] completes the typical steel surface flaw identification [7]-[8].”
MethodFeng, Q., Li, F., Li, H., Liu, X., Fei, J., Xu, S., Lu, C. B., & Yang, Q. (2023) · The Visual Computer · doi:10.1007/s00371-023-03056-w · Cited by 14
“He, et al. [24] designed an end-to-end steel surface defect detection network (DDN) that uses a pretrained ResNet network as the backbone to generate feature maps at different scales, and the region of interest (ROI) is obtained using a region proposal network (RPN), which then generates the final…”
Steel strip surface defect detection based on multiscale feature sensing and adaptive feature fusion
Mi, Z., Gao, Y., Xu, X., & Tang, J. (2024) · AIP Advances · doi:10.1063/5.0196580 · Cited by 20
MSSIF-Net: an efficient CNN automatic detection method for freight train images
Zhang, L., Hu, Y., Chen, J., Li, C., & Li, K. (2022) · Neural Computing and Applications · doi:10.1007/s00521-022-08035-1 · Cited by 14
Image Restoration for Blast Furnace Burden Surface Based on Dust Multiscattering Model
Li, Y., Jiang, Z., Yi, Z., Pan, D., Yang, B., & Gui, W. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3284022 · Cited by 13
“recognition [2], defect detection [3], [4], and fault diagno-”
BackgroundParallel Dilated CNN for Detecting and Classifying Defects in Surface Steel Strips in Real-Time
Ragab, K. (2021) · Lecture notes in networks and systems · doi:10.1007/978-3-030-82193-7_11 · Cited by 13
Zhu, L., Baolin, D., Xiaomeng, Z., Fang, S., Zhen, C., Zhou, J., & Shumin, C. (2022) · Scientific Programming · doi:10.1155/2022/4481495 · Cited by 16
Gao, S., Jiang, Y., Xia, T., Li, Y., Zhu, Y., & Xi, L. (2024) · Advanced Engineering Informatics · doi:10.1016/j.aei.2024.103080 · Cited by 14
Yu, J., Sun, Y., Latinović, N., Kong, C., Han, B., & Zhang, X. (2025) · Journal of Food Composition and Analysis · doi:10.1016/j.jfca.2025.107744 · Cited by 12
YOLO-LSDI: An Enhanced Algorithm for Steel Surface Defect Detection Using a YOLOv11 Network
Wang, F., Jiang, X., Han, Y., & Wu, L. (2025) · Electronics · doi:10.3390/electronics14132576 · Cited by 19
Lema, D. G., Sánchez-González, L., Usamentiaga, R., & delaCalle, F. J. (2025) · Journal of Intelligent Manufacturing · doi:10.1007/s10845-025-02672-8 · Cited by 25
“In Song et al. (2020), it is addressed this by adapting the NEU dataset for binary semantic segmentation, emphasizing defect contours through a saliency-based approach.”
Wu, S., Yang, H., Liao, L., Song, C., Fang, Y., & Yang, Y. (2025) · PLoS ONE · doi:10.1371/journal.pone.0334048 · Cited by 18
“…Feature Fusion Network (MFN) facilitates the incorporation of features from multiple layers, encompasses detailed semantic information, and emphasizes areas of interest via the Regional Suggestion Network (RPN), thereby propelling research forward in the area of steel defect detection [21].”
EDDNet: An Efficient and Accurate Defect Detection Network for the Industrial Edge Environment
Qin, R., Chen, N., & Huang, Y. (2022) · IEEE International Conference on Software Quality, Reliability and Security (QRS) · doi:10.1109/qrs57517.2022.00090 · Cited by 18
“To evaluate the performance of the designed model, we select three surface defect datasets: NEU-DET [26], PCB-Defect [27], and DAGM2007 [28].”
MethodEvolutionary neural architecture search combining multi-branch ConvNet and improved transformer
Yang, X., & Ma, Y. (2023) · Scientific Reports · doi:10.1038/s41598-023-42931-3 · Cited by 14
Luo, J., Huang, P., He, P., Wei, B., Guo, X., Xiao, H., Sun, Y., Tian, S., Zhou, M., & Feng, P. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3328088 · Cited by 22
“Researchers have carried out a lot of meaningful works on pixel-level local feature extraction [3], [4], [5], [6], which can be summarized into three branches in general.”
BackgroundDing, M., Zhou, W., Xu, Y., & Xu, Y. (2023) · IEEE Transactions on Aerospace and Electronic Systems · doi:10.1109/taes.2023.3342797 · Cited by 16
“The goal of instance segmentation is to categorize and localize each target in images using a pixel-level mask [14], [15].”
MethodAnomalySeg: Deep Learning-Based Fast Anomaly Segmentation Approach for Surface Defect Detection
Song, Y., Xia, W., Li, Y., Li, H., Yuan, M., & Zhang, Q. (2024) · Electronics · doi:10.3390/electronics13020284 · Cited by 17
A Highly Efficient and Lightweight Detection Method for Steel Surface Defect
Xu, C., Li, J., & Li, X. (2024) · Journal of Nondestructive Evaluation · doi:10.1007/s10921-024-01084-7 · Cited by 16
“NEU-DET [39] is a publicly available dataset in the field of steel surface defect detection and consists of images of six typical steel surface defects collected by Northeastern University.”
Yedukondalu, J., Karaddi, S. H., Bindu, C. H., Sharma, D., Sarkar, A. K., & Sharma, L. D. (2024) · Arabian Journal for Science and Engineering · doi:10.1007/s13369-024-09230-z · Cited by 15
“Conventional machine vision [3, 4] and human visual examination [5, 6] of steel surfaces together form the standard for flaw detection.”
Han, F., Hua, H., Zhang, R., Zou, Y., Xue, L., & Wang, C. (2024) · Physica Scripta · doi:10.1088/1402-4896/ad65bd · Cited by 12
Luo, A., Wen, G., Cheng, Y., Mei, S., Dong, H., & Liu, X. (2024) · Neurocomputing · doi:10.1016/j.neucom.2024.128622 · Cited by 13
Comparison and Optimization of Rail Defect Detection Methods Based on Object Detection Model
Zhang, H., Cui, X., Yin, Y., Tang, C., Ding, H., Zhao, X., & Zhong, J. (2025) · Tribology Transactions · doi:10.1080/10402004.2024.2449503 · Cited by 13
Smoothing Complete Feature Pyramid Networks for Roll Mark Detection of Steel Strips
Luo, Q., Jiang, W., Su, J., Ai, J., & Yang, C. (2021) · Sensors · doi:10.3390/s21217264 · Cited by 13
“For making a fair comparison, in Table 3, we selected four more methods which have been verified on the NEU database, RAF-SSD [14], SCN [27], DIN [28], and DDN [17].”
BackgroundMethodAkhyar, F., Furqon, E. N., & Lin, C. (2022) · Sensors · doi:10.3390/s22114257 · Cited by 14
A Deep Regression Framework Toward Laboratory Accuracy in the Shop Floor of Microelectronics
Evangelidis, A., Dimitriou, N., Leontaris, L., Ioannidis, D., Tinker, G., & Tzovaras, D. K. (2022) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2022.3182343 · Cited by 15
“A joint detection and classifi-cation scheme is proposed in [13] for steel plate defect inspection through the fusion of multilevel features.”
MethodMorales‐Forero, A., Bassetto, S., & Coatanéa, É. (2022) · AI & Society · doi:10.1007/s00146-022-01591-z · Cited by 23
Chen, C., Shen, F., & Dai, C. (2023) · Measurement · doi:10.1016/j.measurement.2023.112953 · Cited by 13
Xu, Y. W., Chen, J., Liang, Y., Zhai, Y., Ying, Z., Zhou, W., Genovese, A., Piuri, V., & Scotti, F. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3249221 · Cited by 15
“Surface defect inspection is mainly processed through automated visual inspection (AVI) instruments [1], [2], [3].”
MethodWang, Y., Zhang, Y., Jiang, Z., Li, Z., Chen, J., & Lu, J. (2023) · IEEE Sensors Journal · doi:10.1109/jsen.2023.3336009 · Cited by 13
Inspector gadget: a data programming-based labeling system for industrial images
Heo, G., Roh, Y., Hwang, S., Lee, D., & Whang, S. E. (2020) · Proceedings of the VLDB Endowment · doi:10.14778/3421424.3421429 · Cited by 15
“(e) NEU [16] 20 30 40 50 60 70 80 Size of Dev.”
BackgroundMethodInfluentialComputer vision in branding: A conceptual framework and future research agenda
Li, Y., Lee, H. H. M., & Blasco‐Arcas, L. (2025) · Journal of Business Research · doi:10.1016/j.jbusres.2025.115329 · Cited by 16
Li, Q., Ding, C., Wang, B., Jiao, J., Huang, W., & Zhu, Z. (2025) · Advanced Engineering Informatics · doi:10.1016/j.aei.2025.103514 · Cited by 11
Li, D., Hua, S., Li, Z., Gong, X., & Wang, J. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3180423 · Cited by 11
A multiple feature-maps interaction pyramid network for defect detection of steel surface
Zhao, X., Zhao, J., & He, Z. (2023) · Measurement Science and Technology · doi:10.1088/1361-6501/acb073 · Cited by 12
Lim, W.-H., Bonab, M. B., & Chua, K. H. (2023) · IEEE Access · doi:10.1109/access.2023.3266737 · Cited by 14
“In manufacturing, CNN have been extensively used for the purpose of visual inspection including applications for civil works [17], [18], railroad tracks [19], [20], electronic components [21], [22] and steel surfaces [23], [24], [25], [26], [27], [28], [29].”
MethodZhang, X., Zhang, J., Chen, J., Guo, R., & Wu, J. (2023) · The Visual Computer · doi:10.1007/s00371-023-02980-1 · Cited by 12
“…dataset (AIRSKIN_DET), German texture background defects dataset (DAGM2007 [29]), Northeast-ern University steel surface defect dataset (NEU_DET [12]), Peking University PCB surface defect dataset (PCB [30]), and Tianchi_Fabric defect intelligent detection dataset (TianChi_Fabric [31]), are…”
Wang, H., Zhang, R., Feng, M., Liu, Y., & Yang, G. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3309374 · Cited by 19
“classification [17], [18], [19], and fine-grained pixel-level”
BackgroundMicro LED defect detection with self-attention mechanism-based neural network
Zhong, Z., Li, C., Chen, M., Wu, H. Y., & Kiyoshi, T. (2024) · Digital Signal Processing · doi:10.1016/j.dsp.2024.104474 · Cited by 11
Space to depth convolution bundled with coordinate attention for detecting surface defects
Wan, W., Wang, L., Wang, B., Yu, H., Shi, K., & Liu, G. (2024) · Signal Image and Video Processing · doi:10.1007/s11760-024-03122-3 · Cited by 13
“NEU-DET [34] is a typical dataset of steel surface defects published by Northeastern University.”
Liu, B., Zhao, Y., Chen, B.-L., Yu, C., & Chang, K. (2024) · Measurement Science and Technology · doi:10.1088/1361-6501/ad4fb6 · Cited by 13
“The Northeastern University (NEU) surface defect dataset [32] is primarily employed for evaluating and comparing the performance of surface defect detection algorithms.”
Luo, Q., Xu, W., Su, J., Yang, C., Gui, W., Sílven, O., & Liu, L. (2024) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2024.3413144 · Cited by 18
Li, J., Wu, J., & Shao, Y. (2024) · Applied Sciences · doi:10.3390/app14177913 · Cited by 20
“The accuracy and timeliness [2] of data detection directly affect the improvement level of industrial product packaging line monitoring and governance, and even the entire industrial production and logistics quality control system [3].”
BackgroundSDMS-YOLOv10: improved Yolov10-based algorithm for identifying steel surface flaws
Xie, H., Zhou, H., Chen, R., & Wang, B. (2025) · Nondestructive Testing And Evaluation · doi:10.1080/10589759.2025.2474103 · Cited by 13
“Next, we apply the methodologies outlined in the literature [13] to select semantically pertinent key-value pairs by focusing on the top-k routing region k ¼ 1 ð Þ .”
InfluentialPang, M., Lian, Z., Dong, H., Xu, J., Zhang, Y., Zhou, A., & Dou, J. (2025) · Measurement · doi:10.1016/j.measurement.2025.118061 · Cited by 15
Luo, Q., Liu, K., Su, J., Yang, C., Gui, W., Liu, L., & Sílven, O. (2021) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2021.3098825 · Cited by 13
“utomated visual inspection (AVI) instrument plays an essential role for hot-rolled steel mills to ensure product quality [1] [2] [3].”
BackgroundCopper Strip Surface Defect Detection Model Based on Deep Convolutional Neural Network
Xu, Y., Wang, D., Duan, B., Yu, H., & Liu, H. (2021) · Applied Sciences · doi:10.3390/app11198945 · Cited by 11
Multilayer Feature Extraction of AGCN on Surface Defect Detection of Steel Plates
Zhang, C., Cui, J., & Liu, W. (2022) · Computational Intelligence and Neuroscience · doi:10.1155/2022/2549683 · Cited by 18
“e use of steel is essential in many industries, such as petrochemical, vehicle engineering, aerospace, military defense, and other elds, and its excellent performances have an invaluable role [1].”
BackgroundXie, Q., Zhou, W., Ma, L., Chen, Z., Wu, W., & Wang, X. (2022) · Signal Image and Video Processing · doi:10.1007/s11760-022-02375-0 · Cited by 12
Cost-Sensitive YOLOv5 for Detecting Surface Defects of Industrial Products
Liu, B., Gao, F., & Li, Y. (2023) · Sensors · doi:10.3390/s23052610 · Cited by 13
“On th other hand, the two-stage detector Faster-RCNN [4] is also utilized in defect detection [46], which especially perfroms better in denser detection.”
MethodAn Adaptive Shunt Model for Steel Defect Detection based on YOLOX
Liu, S., & Jia, M. (2023) · IEEE Information Technology, Networking, Electronic and Automation Control Conference (ITNEC) · doi:10.1109/itnec56291.2023.10082274 · Cited by 11
“B. Implementation Details For the NEU-DET dataset, the input size of the model is set to 224×224.”
BackgroundInfluentialFu, H., & Cheng, G. (2023) · Energy Sources Part A Recovery Utilization and Environmental Effects · doi:10.1080/15567036.2023.2230935 · Cited by 12
“In this process, “None” signifies that no auxiliary dataset was employed, “VOC1800” denotes the use of 1,800 images from the PASCAL VOC2012 dataset, “VOC2012” represents the application of the entire PASCAL VOC2012 dataset, and “NEU-DET” refers to a similar dataset.”
MethodInfluentialMa, Q., Zhang, E., Chen, Y., Duan, J., & Shao, L. (2023) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2023.107131 · Cited by 15
A New Graph-Based Method for Class Imbalance in Surface Defect Recognition
Wang, Y., Gao, L., Li, X., Gao, Y., & Xie, X. (2021) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2021.3063755 · Cited by 15
“3) ACB-GCN is evaluated in four famous benchmarks, NEU-CLS-64 [18], wood data set [19], textile data set [20] and GC10 (GC10-DET-Metallic-Surface-Defect-Data Sets) [21], to fully show that the method can address the coupling between class imbalance and ISIV, so it can extract discriminative…”
MethodYolov4 High-Speed Train Wheelset Tread Defect Detection System Based on Multiscale Feature Fusion
Zhang, C., Hu, X., He, J., & Hou, N. (2022) · Journal of Advanced Transportation · doi:10.1155/2022/1172654 · Cited by 14
“He et al. [17] proposed an end-to-end two-step defect detection method for steel rolling defects and explored the trade-off between detection speed and accuracy for different numbers of regions.”
A weighted multi‐source domain adaptation approach for surface defect detection
Hu, B., & Wang, J. (2022) · IET Image Processing · doi:10.1049/ipr2.12484 · Cited by 12
Enhanced You Only Look Once X for surface defect detection of strip steel
Wu, R., Zhou, F., Li, N., Liu, H., Guo, N., & Wang, R. (2022) · Frontiers in Neurorobotics · doi:10.3389/fnbot.2022.1042780 · Cited by 12
“The commonly used one is the NEU surface defect database (Song and Yan, 2013; He et al., 2019; Bao et al., 2021) collected and collated by Mr. Song Kechen from Northeastern University, which contains six types of typical strip steel surface defects: Crazing (Cr), Inclusion (In), Patches (Pa),…”
MethodEnhanced Non-Maximum Suppression for the Detection of Steel Surface Defects
Kang, S., Palakonda, V., Kim, I., Kang, J., & Yun, S. (2023) · Mathematics · doi:10.3390/math11183898 · Cited by 11
Deng, Y., & Song, Y. (2023) · ISIJ International · doi:10.2355/isijinternational.isijint-2023-118 · Cited by 13
Xu, S., Zheng, H., Tao, S., Chai, Y., He, Q., & Chen, H. (2024) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2024.3374313 · Cited by 10
“In recent years, with the continuous advancement of deep-learning technology in the computer vision field [6], [7], [8], several studies have explored the potential of deep-learning techniques for pig face recognition.”
BackgroundLi, Z., Gao, L., Li, X., & Gao, Y. (2024) · Advanced Engineering Informatics · doi:10.1016/j.aei.2024.102527 · Cited by 11
Boudiaf, A., Benlahmidi, S., Dahane, A., & Bouguettaya, A. (2024) · Journal of Failure Analysis and Prevention · doi:10.1007/s11668-024-01927-5 · Cited by 14
“In this work, we have used a public dataset of surface defects in hot-rolled steel strips, which were obtained from the laboratories of Northeastern University (NEU) [2, 40, 41].”
Shi, Z., Fang, Y., & Song, H. (2024) · Sensors · doi:10.3390/s24144660 · Cited by 13
A three-stage framework for accurate detection of high-speed train body paint film defects
Wang, S., Xu, Z., Wang, Y., Tan, Z., & Zhu, D. (2024) · Advanced Engineering Informatics · doi:10.1016/j.aei.2024.102838 · Cited by 10
Discrete Wavelet Transform-Based Detection Transformer for Battery Weld Defect Detection
Zhang, K., Liao, L., & Wang, Y. (2024) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2024.3502839 · Cited by 12
DF-YOLOv7: steel surface defect detection based on focal module and deformable convolution
Zhang, W., Huang, T., Xu, J. Y., Yu, Q., He, Y., Lai, S., & Xu, Y. (2024) · Signal Image and Video Processing · doi:10.1007/s11760-024-03679-z · Cited by 12
“As shown in Table 3, a comparison of various YOLO detectors—including SSD, YOLOv3, YOLOv4, The proposed algorithm is compared with some SOTA detection methods [15, 29–32] to further demonstrate the superiority of DF-YOLOv7 for surface defect detection.”
Self-citationFederated Learning-Based Equipment Fault-Detection Algorithm
Han, J., Zhang, X., Xie, Z., et al. (2024) · Electronics · doi:10.3390/electronics14010092 · Cited by 10
Latent normal images-based zero-negative sample rail surface defect segmentation method
Yan, B., Yang, F., Qiu, S., Wang, J., Xu, L., Wang, W., & Peng, J. (2025) · Automation in Construction · doi:10.1016/j.autcon.2025.106097 · Cited by 11
Huang, J., Wu, Y. Z., Zhou, X., et al. (2025) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2025.3550227 · Cited by 10
“T HE advancements in steel surface defect detection technologies [1], [2], [3] have revolutionized industrial quality control through the integration of deep learning and computer vision approaches.”
Sun, W., Na, M., Chen, L., Yang, S., Li, Y., & Tian, S. (2025) · Machines · doi:10.3390/machines13040301 · Cited by 13
Wu, P., Wen, W., Han, L., et al. (2025) · The Innovation · doi:10.1016/j.xinn.2025.101084 · Cited by 10
Zhao, Z., & Wu, T. (2022) · Scientific Programming · doi:10.1155/2022/4385565 · Cited by 13
“proposed a point pattern matching positioning algorithm based on azimuth environment characteristics to realize the detection of casting pin defects [2].”
BackgroundApplication of Improved YOLOv5 Model for Strip Surface Defect Detection
Nuo, Y., & Guo, W. J. (2022) · Global Reliability and Prognostics and Health Management (PHM) · doi:10.1109/phm-yantai55411.2022.9942194 · Cited by 10
Toward TR-PCB Bubble Detection via an Efficient Attention Segmentation Network and Dynamic Threshold
Zhou, J., Zhu, Q. X., Wang, Y., Zhou, X., Feng, M., Liu, X., & Mo, Y. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3244806 · Cited by 10
“He et al. [15] proposed an end-to-end defect detection pipeline DDN that integrates residual network (ResNet) and an region proposal network (RPN) for precise defect classification and localization.”
BackgroundGao, Y., Li, X., & Gao, L. (2023) · International Journal of Computer Integrated Manufacturing · doi:10.1080/0951192x.2023.2229270 · Cited by 10
“However, these common methods require an explicit feature extraction, if the extracted features are improper, the recognition methods will not work as expected (He et al. 2020).”
BackgroundCoRe: Contrastive and Restorative Self-Supervised Learning for Surface Defect Inspection
Wu, H., Li, B., Tian, L., Sun, Z., Dong, C., & Liao, W. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3291776 · Cited by 14
“These datasets include NEU-DET [4], MT Defect [25], KolektorSDD2 dataset [26], RSDDs dataset [27], and Severstal Steel dataset [28], which encompass various label formats (i.e., instance-level label for defect detection task and pixel-level label for defect segmentation task).”
BackgroundMethodInfluentialAn RGB-D-Based Thickness Feature Descriptor and Its Application on Scrap Steel Grading
Gao, Z., Lu, H., Lei, J., et al. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3328089 · Cited by 10
Zeng, W., Wang, J., Chen, P., Zhong, Z., Hu, G., & Bao, W. (2024) · Signal Image and Video Processing · doi:10.1007/s11760-024-03270-6 · Cited by 14
“We conducted tests on the NEU-DET [39] (hot-rolled strip steel surface defect detection dataset) using MobileNetv1, MobileNet-Pro with 3 × 3 convolution kernels, and MobileNet-Pro with 7 × 7 convolution kernels.”
Feature fusion–based preprocessing for steel plate surface defect recognition
Tian, Y., Zhang, T., Zhang, Q., Li, Y., & Wang, Z. (2020) · Mathematical Biosciences & Engineering · doi:10.3934/mbe.2020305 · Cited by 11
Vehicle Tire Text Reader: Text Spotting and Rectifying for Small, Curved, and Rotated Characters
Gao, F., Ge, Y., Lu, S., et al. (2021) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2021.3111973 · Cited by 10
Low-Cost Real-Time Automated Optical Inspection Using Deep Learning and Attention Map
Shih, Y., Kuo, C.-C., & Lee, C. (2022) · Intelligent Automation & Soft Computing · doi:10.32604/iasc.2023.027659 · Cited by 13
Pei, Y., Hou, M., Han, Q., Weng, T., Tian, Y., Chen, G., Liu, J., & Chen Xi, W. U. (2022) · Computational Intelligence and Neuroscience · doi:10.1155/2022/7539857 · Cited by 10
“To solve these problems, the deep learning method is applied to steel surface defect detection, which includes steel surface defect recognition, detection [9], segmentation, and other tasks.”
MethodAn improved YOLOX method for surface defect detection of steel strips
Wang, X., & Zhuang, K. (2023) · IEEE International Conference on Power, Electronics and Computer Applications (ICPECA) · doi:10.1109/icpeca56706.2023.10075827 · Cited by 10
“[6] designed a multilevel feature fusion network to obtain rich location information for detecting defects of steels based on Faster RCNN.”
MethodA dataset for surface defect detection on complex structured parts based on photometric stereo
Wu, L., Yu, R., Li, Y., Liu, Y., Song, Y., & Han, D. (2025) · Scientific Data · doi:10.1038/s41597-025-04454-6 · Cited by 17
TRIDENet: A noise-robust and boundary-aware detection framework for industrial surface defects
Pan, W., Wang, G., Lv, C., Ma, Z., & Wang, X. (2026) · Optics & Laser Technology · doi:10.1016/j.optlastec.2026.114843 · Cited by 19
Photovoltaic glass edge defect detection based on improved SqueezeNet
Xiong, J., He, Z., Zhou, Q., & Yang, R. (2024) · Signal Image and Video Processing · doi:10.1007/s11760-023-02954-9 · Cited by 10
“…and artificial intelligence, has also developed rapidly in recent years, and defect detection on the surface of objects based on machine vision technology has a large number of applications in various fields such as semiconductor, photovoltaic, automotive, agriculture, etc. [14, 15, 29, 30, 34].”
Weiher, K., Rieck, S., Pankrath, H., Beuß, F., Geist, M., Sender, J., & Fluegge, W. (2023) · Procedia CIRP · doi:10.1016/j.procir.2023.09.088 · Cited by 12
A Deep-Learning Method Based on the Multistage Fusion of Radar and Camera in UAV Obstacle Avoidance
Wang, X., Wang, X., Zhou, Z., & Song, Y. (2024) · IEEE Transactions on Aerospace and Electronic Systems · doi:10.1109/taes.2024.3406495 · Cited by 13
“ResNet presents several advantages as follows [43].”
Metadata
- Authors
- He, Y., Song, K., Meng, Q., & Yan, Y.
- Affiliations
- Northeastern University (CN)
- Loughborough University (GB)
- Venue
- IEEE Transactions on Instrumentation and Measurement
- Published
- 2020-04
- Volume
- 69
- Issue
- 4
- Pages
- 1493-1504
- ISSN
- 0018-9456 (ISSN-L)
- 1557-9662
- OpenAlex
- W2944303778
- Semantic Scholar
- 46d4a2b7ccf38b80b09574cf17544ff2297a8fbf
- Cited by
- 1,341
- References
- 53
- Citations by year
- 2026: 188 · 2025: 304 · 2024: 269 · 2023: 238 · 2022: 186 · 2021: 111 · 2020: 34
- FWCI
- 65.18
- Type
- journal article
- Language
- en
- Open access
- green
- figshare.com
- Licence
- Locations
- IEEE Transactions on Instrumentation and Measurement (publishedVersion)
- Figshare (submittedVersion, cc-by)
- Funders
- National Natural Science Foundation of China (51805078, 51374063)
- National Basic Research Program of China (2017YFB0304200)
- Fundamental Research Funds for the Central Universities (N170304014, N150308001)
- China Scholarship Council (201806085007)
- Topics
- Physical Sciences › Engineering › Industrial and Manufacturing Engineering › Industrial Vision Systems and Defect Detection
- Physical Sciences › Engineering › Civil and Structural Engineering › Infrastructure Maintenance and Monitoring
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Keywords
- multi-level feature fusion, hierarchical features, region proposal network, bounding box regression, NEU-DET dataset, real-time defect detection, industrial visual inspection, object detection
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.