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 301–400 of 1,338
Deep Residual Neural Network-Based Defect Detection on Complex Backgrounds
Ho, C.-C., Hernandez, M. A. B., Chen, Y., Lin, C., & Chen, C. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3144224 · Cited by 23
A Scratch Detection Method Based on Deep Learning and Image Segmentation
Yang, L., Zhou, F., & Wang, L. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3186054 · Cited by 30
Zhao, Z., Hu, B.-T., Feng, Y.-X., Zhao, B., Yang, C., Hong, Z.-X., & Tan, J.-R. (2022) · International Journal of Production Research · doi:10.1080/00207543.2022.2138613 · Cited by 22
“Meanwhile, other best-performing defect detection methods (Chu et al. 2017; He et al. 2020; Yang et al. 2020; Tao et al. 2021) proposed recently are reimplemented to make a comparison.”
MethodInfluentialSwin Transformer Combined with Convolution Neural Network for Surface Defect Detection
Li, Y., Xiang, Y., Guo, H., Liu, P., & Liu, C. (2022) · Machines · doi:10.3390/machines10111083 · Cited by 25
Wang, H., Wang, W., Yan, J., Fu, C., & Liu, W. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3225035 · Cited by 21
“Among them, the neural network method with new network models proposed, such as convolutional neural networks (CNNs) [22], recurrent neural networks (RNN) [23], and generative adversarial networks (GAN) [24], has surpassed most other methods in performance.”
MethodCANet: Context-aware Aggregation Network for Salient Object Detection of Surface Defects
Wan, B., Zhou, X., Zhu, B. B., Xiao, M., Sun, Y., Zheng, B., Zhang, J., & Yan, C. C. (2023) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2023.103820 · Cited by 21
Multiscale Local and Global Feature Fusion for the Detection of Steel Surface Defects
Zhang, L., Fu, Z., Guo, H., Sun, Y., Li, X., & Xu, M. (2023) · Electronics · doi:10.3390/electronics12143090 · Cited by 22
“For example, He et al. [4] designed a multilevel feature-fusion network that combines multiple hierarchical features from CNNs so as to have more location details.”
BackgroundMethodAutomatic Solder Defect Detection in Electronic Components Using Transformer Architecture
Liu, Y., & Wu, H. (2023) · IEEE Transactions on Components Packaging and Manufacturing Technology · doi:10.1109/tcpmt.2023.3344096 · Cited by 20
“He et al. [19] utilized a CNN to extract multiscale image features, which were then fused with a mul-tifeature network for feature fusion, followed by region-based analysis.”
MethodDetecting Teeth Defects on Automotive Gears Using Deep Learning
Allam, A., Moussa, M. A., Tarry, C., & Veres, M. (2021) · Sensors · doi:10.3390/s21248480 · Cited by 26
“[16] added a multilevel feature fusion network (MFN) to the feature maps extracted from a CNN trained with the NEU-DET defect classification dataset [17].”
MethodEnhanced multiscale attentional feature fusion model for defect detection on steel surfaces
Xia, Y., Lu, Y., Jiang, X., & Xu, M. (2024) · Pattern Recognition Letters · doi:10.1016/j.patrec.2024.11.024 · Cited by 21
Detecting Machining Defects inside Engine Piston Chamber with Computer Vision and Machine Learning
Abagiu, M. M., Cojocaru, D., Manta, L. F., & Mariniuc, A. M. (2023) · Sensors · doi:10.3390/s23020785 · Cited by 22
Hussain, M., & Hill, R. C. (2023) · IEEE Access · doi:10.1109/access.2023.3283596 · Cited by 19
“[28] claim a novel defect detection system, focused on industrial applications using steel surface defect as a use case.”
Hussain, M. (2023) · Big Data and Cognitive Computing · doi:10.3390/bdcc7020120 · Cited by 19
“He et al. [16] proposed a novel approach for defect detection in industrial settings, explicitly focusing on defect identification in steel surfaces.”
BackgroundExplainable efficient and optimized feature fusion network for surface defect detection
Sundarrajan, K., & Rajendran, B. K. (2023) · The International Journal of Advanced Manufacturing Technology · doi:10.1007/s00170-023-11789-0 · Cited by 19
Du, Y., Chen, J., Zhou, H., Yang, X., Wang, Z., Zhang, J., Shi, Y., Chen, X., & Zheng, X. (2023) · Optics Communications · doi:10.1016/j.optcom.2023.129736 · Cited by 22
Yang, B., Liu, Z., Duan, G., & Tan, J. (2023) · Pattern Recognition · doi:10.1016/j.patcog.2023.110073 · Cited by 22
Wang, B., Zhou, W., Yan, W., Jiang, Q., & Cong, R. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3330222 · Cited by 22
“This method has already been successfully applied in detecting surface defects in various materials such as steel strip [1], [2], glass [3], [4], and additional steel strips [5], [6].”
MethodMPFANet: a multipath feature aggregation network for steel surface defect detection
Li, Z., Tai, Y., Huang, Z., Peng, T., & Zhang, Z. (2024) · Measurement Science and Technology · doi:10.1088/1361-6501/ad1daf · Cited by 19
“He et al [20] introduced a multi-level feature fusion module based on CNN, where the fused feature contains more detailed information about defect locations.”
Resilient machine learning for steel surface defect detection based on lightweight convolution
Liu, L., Zhang, Y., & Karimi, H. R. (2024) · The International Journal of Advanced Manufacturing Technology · doi:10.1007/s00170-024-14403-z · Cited by 23
“Two publicly available datasets for detecting defects on steel surfaces are employed, namely NEU-DET [29] and GC10-DET [30].”
Meng, Y., Xu, H., Ma, Z., Zhou, J., & Hui, D. (2022) · Applied Intelligence · doi:10.1007/s10489-022-03671-5 · Cited by 19
“Nowadays, deep learning methods have been used in industry, especially surface defect detection [31, 32].”
MethodChen, X., Zhang, X., Shi, Y., & Pang, J. (2025) · Sensors · doi:10.3390/s25051333 · Cited by 27
“The NEU-DET dataset [18] is a collection focused on steel material defect detection, covering six different types of surface defects, namely, crazing, patches, inclusion, pitted surface, rolled-in scale, and scratches, with example images shown in Figure 6.”
Deep Learning-Based Robot Vision: High-End Tools for Smart Manufacturing
Zhang, H., Liu, L. Z., Xie, H., Jiang, Y., Zhou, J., & Wang, Y. (2022) · IEEE Instrumentation & Measurement Magazine · doi:10.1109/mim.2022.9756392 · Cited by 23
Physics Informed Synthetic Image Generation for Deep Learning-Based Detection of Wrinkles and Folds
Manyar, O. M., Cheng, J., Levine, R., Krishnan, V., Barbič, J., & Gupta, S. K. (2022) · Journal of Computing and Information Science in Engineering · doi:10.1115/1.4056295 · Cited by 23
An End-to-End Steel Surface Classification Approach Based on EDCGAN and MobileNet V2
Jin, G., Liu, Y., Qin, P., Hong, R., Xu, T., & Lu, G. (2023) · Sensors · doi:10.3390/s23041953 · Cited by 23
“propose a novel defect detection system and focus on steel plate defect detection [15].”
BackgroundBi, L. H., & Gao, Q. (2023) · Sensors · doi:10.3390/s23073761 · Cited by 24
“[27] proposed a deep learning-based defect detection method for steel plates.”
BackgroundYu, Y., Chan, S., Tang, T., Zhou, X., Yao, Y., & Zhang, H. (2023) · Electronics · doi:10.3390/electronics12081856 · Cited by 20
Deformable residual attention network for defect detection of train wheelset tread
Zhang, C., Xu, Y., Sheng, Z., He, J., & Yin, L. (2023) · The Visual Computer · doi:10.1007/s00371-023-02885-z · Cited by 20
Optimization Algorithm for Steel Surface Defect Detection Based on PP-YOLOE
Qu, Y., Wan, B., Wang, C., Ju, H., Yu, J., Kong, Y., & Chen, X. (2023) · Electronics · doi:10.3390/electronics12194161 · Cited by 21
A Novel ST-YOLO Network for Steel-Surface-Defect Detection
Ma, H., Zhang, Z., & Zhao, J. (2023) · Sensors · doi:10.3390/s23229152 · Cited by 25
“Dynamic NEU-DET [36] is a dataset of steel surfaces consisting of 1800 photos of hot-rolled steel strips.”
BackgroundMobile Phone Surface Defect Detection Based on Improved Faster R-CNN
Wang, T., Zhang, C., Ding, R., et al. (2021) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr48806.2021.9412119 · Cited by 18
“[14] proposed a regression and classification based framework for industrial surface defect detection, and finally achieved the state-ofthe-art performance.”
BackgroundImpact of Image Compression on the Performance of Steel Surface Defect Classification with a CNN
Benbarrad, T., Eloutouate, L., Arioua, M., Elouaai, F., & Laanaoui, M. D. (2021) · Journal of Sensor and Actuator Networks · doi:10.3390/jsan10040073 · Cited by 22
“Each row shows one example image from each of 300 samples of a class [21,29], Figure 6 is Adapted from [21], with permission from © 2013 Elsevier.”
BackgroundMethodLightweight edge-attention network for surface-defect detection of rubber seal rings
Huang, Z., Hu, H., Shen, Z., Zhang, Y., & Zhang, X. (2022) · Measurement Science and Technology · doi:10.1088/1361-6501/ac6663 · Cited by 18
“These networks have achieved many successful applications in different defect detection tasks, such as concrete crack detection [23, 24], loosened bolt detection [25, 26] and strip steel surface defect detection [27].”
A Model for Surface Defect Detection of Industrial Products Based on Attention Augmentation
Li, G., Shao, R., Wan, H., Zhou, M., & Li, M. (2022) · Computational Intelligence and Neuroscience · doi:10.1155/2022/9577096 · Cited by 21
Ye, Q., Dong, Y., Zhang, X., Zhang, D., & Wang, S. (2024) · Optics and Lasers in Engineering · doi:10.1016/j.optlaseng.2024.108035 · Cited by 21
A triple semantic-aware knowledge distillation network for industrial defect detection
Wen, Z., Liu, J., Zhao, H., & Wang, Q. (2025) · Computers in Industry · doi:10.1016/j.compind.2025.104252 · Cited by 27
TAF2-Net: Triple-Branch Attentive Feature Fusion Network for Ultrasonic Flaw Detection
Li, W., Qi, J., & Sun, H. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3150592 · Cited by 25
“He et al. [15] used baseline CNN to generate feature maps at each stage and then fused the above features through a multilevel feature fusion network for defect detection of steel surface image.”
MethodDeep Adversarial Data Augmentation for Fabric Defect Classification With Scarce Defect Data
Lu, B., Zhang, M., & Huang, B. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3185609 · Cited by 21
“Fabric defect classification task is generally completed by extracting features and then feeding the features into a classifier to get category labels [22].”
MethodDefect Detection Method Based on Knowledge Distillation
Zhou, Q., Wang, H., Tang, Y., & Wang, Y. (2023) · IEEE Access · doi:10.1109/access.2023.3252910 · Cited by 18
“INTRODUCTION Industrial defect detection, a key step in ensuring the quality of industrial production, is one of the most important technologies to ensure product quality and maintain production stability [1].”
BackgroundFu, M., Wu, J., Wang, Q., Sun, L., Ma, Z., Zhang, C., Guan, W., Li, W., Chen, N., Wang, D., et al. (2023) · IET Signal Processing · doi:10.1049/sil2.12208 · Cited by 18
Graph-Based Active Learning With Uncertainty and Representativeness for Industrial Anomaly Detection
Xiao, K., Cao, J., Zeng, Z., & Ling, B. W. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3279422 · Cited by 20
“We conducted experiments on the proposed method using both the THT and the publicly available NEU datasets [41], and the experimental results demonstrate the effectiveness of the proposed method.”
MethodZhu, J., He, G., & Zhou, P. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3284050 · Cited by 22
Yang, H., Fang, Y., Liu, L., Ju, H. X., & Kang, K. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3284021 · Cited by 24
“[7] proposed a new defect detection algorithm based on ResNet and region proposal network (RPN), and designed”
MethodLema, D. G., Usamentiaga, R., & Garcia, D. F. (2024) · Measurement Science and Technology · doi:10.1088/1361-6501/ad48a4 · Cited by 18
“A deep learning-based defect identification method for examining steel plates is presented in [10].”
Li, X., Xu, C., Li, J., Zhou, X., & Li, Y. (2024) · Nondestructive Testing And Evaluation · doi:10.1080/10589759.2024.2408441 · Cited by 19
“He et al. [31] proposed a deep neural network (DNN) for steel defect detection based on Faster R-CNN, which integrates multi-level features using a proposed multi-layer feature fusion network (MFN) and generates regions of interest (ROIs) via the region proposal networks (RPNs) to produce final…”
GPR Closed-Loop Denoising Based on Bandpass Filtering Constraints
Liu, X., Liu, S., Jia, Z., Vogt, D., Tian, S., Liu, X., & Lu, Q. (2024) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2024.3498868 · Cited by 17
Wavelet Texture Descriptor for Steel Surface Defect Classification
Belila, D., Khaldi, B., & Aiadi, O. (2024) · Materials · doi:10.3390/ma17235873 · Cited by 19
“Method Accuracy Precision Recall F1-Score NEU-CLS BRISK + SVM (2022) [43] 95.00 AECLPBP (2013) [11] 98.93 GCLBP (2018) [13] 99.11 Decaf (2019) [20] 99.27 ResNet34 (2019) [20] 99.33 VGG19+ (2023) [44] 97.20 RPNet (2024) [19] 98.73 VGG16 + CNN (2024) [45] 99.44 LWT + AE + SVM (2024) [32] 99.”
InfluentialWang, Y., Zhang, W., Chen, D., Zhang, G., Gong, T., Liang, Z., Yin, A., Zhang, Y., & Ding, W. (2025) · Optics & Laser Technology · doi:10.1016/j.optlastec.2025.112802 · Cited by 16
A real–time detection framework for surface defects in ceramic tableware based on deep learning
Sun, P., Hua, C., Ding, W., & Hua, C. (2025) · Expert Systems with Applications · doi:10.1016/j.eswa.2025.128101 · Cited by 17
Jeddi, A. B., Shafieezadeh, A., & Nateghi, R. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3236321 · Cited by 19
“Separate fully convolutional subnetworks are ubiquitously employed in the literature as detection heads for classification and regression tasks (e.g., [32]).”
MethodCollaborative Filter Pruning for Efficient Automatic Surface Defect Detection
Wang, H., Fan, X. N., Ling, P., Wang, B., Chen, H., & Jin, Y. (2023) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2023.3288208 · Cited by 17
“Following recent studies [8], [25], [26], [27], the image sizes of the ELPV-DC dataset, NEU-CLS dataset, GC10-DET dataset, and X-SDD dataset are set to 288 × 288, 200 × 200, 224 × 224, and 128 × 128, respectively.”
MethodChen, X., Jiang, Z., Cheng, H., Zheng, H., & Du, Y. (2023) · Measurement Science and Technology · doi:10.1088/1361-6501/ad0690 · Cited by 19
“…the good performance and widespread applicability of the lightweight LF-YOLOv4 model for enhancing image feature fusion improved in this article, we also conducted experimental analysis on the open-source data-set of steel surface defects, namely NEU-DET, provided by Northeastern University [47].”
YOLO-DD: Improved YOLOv5 for Defect Detection
Wang, J., Wang, W., Zhang, Z., Lin, X., Zhao, J., Chen, M., & Luo, L. (2023) · Computers, materials & continua/Computers, materials & continua (Print) · doi:10.32604/cmc.2023.041600 · Cited by 26
Camouflaged object detection via cross-level refinement and interaction network
Ge, Y., Ren, J., Zhang, Q., He, M., Bi, H., & Zhang, C. (2024) · Image and Vision Computing · doi:10.1016/j.imavis.2024.104973 · Cited by 19
Ye, W., Ren, J., Li, C., Liu, W., Zhang, Z., & Lu, C. (2024) · Structural Control and Health Monitoring · doi:10.1155/2024/2967927 · Cited by 19
“Te NEU-DETdataset contains images of 6 types of surface defects on hot-rolled steel strips, totaling 1800 images [39].”
BackgroundMethodPei, J., Li, S., & Li, Y. (2024) · Digital Signal Processing · doi:10.1016/j.dsp.2024.104582 · Cited by 20
Research on Surface Defect Detection Algorithm of Strip Steel Based on Improved YOLOV3
Zhang, N., & Mi, Z. (2021) · Journal of Physics Conference Series · doi:10.1088/1742-6596/1907/1/012015 · Cited by 17
“Experimental results show that the mAP of the improved YOLOV3 algorithm on NEU-DET[8] data sets reaches 89.”
MethodInfluentialDeep learning model for imbalanced multi-label surface defect classification
Liu, Y., Yuan, Y., & Liu, J. (2021) · Measurement Science and Technology · doi:10.1088/1361-6501/ac41a6 · Cited by 18
“However, most of the datasets are balanced, like [19, 22, 24], which means that the numbers of different classes are always close to each other.”
Wang, Y., Khan, W., & Chung, S.-H. (2024) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2024.3431069 · Cited by 18
“Unlikemostindustrialproductsdiscussedinexistingliter-ature,whichtypicallyhavebroadandflatsurfaces(suchas steel[6]andsolarcells[7]),medicalcathetersarecylindrical.”
BackgroundMethodInfluentialLiao, Y., Lv, M., MingYong, H., Qu, M., Zou, K., Chen, L., & Feng, L. (2024) · Drones · doi:10.3390/drones8090436 · Cited by 18
STE-YOLO: A Surface Defect Detection Algorithm for Steel Strips
Li, D., Wang, E., Li, Z., Yin, Y., Zhang, L., & Zhao, C. (2024) · Electronics · doi:10.3390/electronics14010054 · Cited by 18
DSP-YOLO: an improved YOLO11-based method for steel surface defect detection
Li, X., Fan, Z., Liu, Q., & Wan, X. (2025) · Measurement Science and Technology · doi:10.1088/1361-6501/adfc88 · Cited by 24
“We chose two widely utilized datasets for steel surface defects: NEU-DET [58], constructed by Northeastern University, and GC10-DET [59], released by Tianjin University.”
Metro Anomaly Detection Based on Light Strip Inductive Key Frame Extraction and MAGAN Network
Liu, R., Liu, W., Li, H., Wang, H., Geng, Q., & Dai, Y. (2021) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2021.3128961 · Cited by 21
Attention Deeplabv3 model and its application into gear pitting measurement
Xi, D., Qin, Y., & Wang, Z. (2021) · Journal of Intelligent & Fuzzy Systems · doi:10.3233/jifs-210810 · Cited by 15
Nath, V., Chattopadhyay, C., & Desai, K. A. (2022) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2022.105553 · Cited by 18
An inspection method of rail head surface defect via bimodal structured light sensors
Zheng, J., Wang, L., Liu, J., Wang, S., Wang, S., Wang, L., & Zhang, J. (2022) · International Journal of Machine Learning and Cybernetics · doi:10.1007/s13042-022-01736-y · Cited by 16
“Defect Detection Network(DDN) in [21] proposes a defect detection method in steel surface defect detection field.”
BackgroundMethodResultInfluentialBaiqiao, 郝. H., Yugang, 范. F., & Zhihuan, 宋. S. (2023) · Acta Optica Sinica · doi:10.3788/aos221532 · Cited by 16
A One‐Stage Deep Learning Model for Industrial Defect Detection
Li, Z., Wei, X., Hassaballah, M., & Jiang, X. (2023) · Advanced Theory and Simulations · doi:10.1002/adts.202200853 · Cited by 15
Yang, H., Zhu, Z., Lin, C., Hui, W., Wang, S., & Zhao, Y. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3273681 · Cited by 17
“Instead of traditional manual detection with low efficiency, industrial defect detection aims to detect abnormal areas that deviate from normal efficiently and accurately, which is one of the key technologies to ensure product quality and maintain production stability [3], [4], [5].”
BackgroundSurface Defect Detection With Channel-Spatial Attention Modules and Bi-Directional Feature Pyramid
Hai-tao, X., & Kai, Z. (2023) · IEEE Access · doi:10.1109/access.2023.3303897 · Cited by 17
“research institute of Song Kechen’s team [22] at Northeastern University.”
MethodImplementing Deep Learning-Based Intelligent Inspection for Investment Castings
Yousef, N., & Sata, A. V. (2023) · Arabian Journal for Science and Engineering · doi:10.1007/s13369-023-08240-7 · Cited by 18
Chen, F., Deng, M., Gao, H., Yang, X., & Zhang, D. (2023) · Cluster Computing · doi:10.1007/s10586-023-04181-w · Cited by 15
“The proposed method is primarily evaluated on the NEU-DET [40] dataset.”
Zhang, R. C., Tang, J., Xia, H., Pan, X., Yu, W., & Qiao, J. (2024) · Neural Computing and Applications · doi:10.1007/s00521-023-09329-8 · Cited by 17
“The semantics of deep features are more robust [28].”
Liu, G., Yan, Y., & Meng, J. (2024) · Measurement Science and Technology · doi:10.1088/1361-6501/ad25df · Cited by 19
“The BM-YOLOv8 model was first pre-trained in NEU-DET [29] before learning the sample defects on the inner-wall outer surface of the ABS brake master cylinder.”
Self-citationAdvancing Ultrasonic Defect Detection in High-Speed Wheels via UT-YOLO
Zhang, Q., Peng, J., Tian, K., Wang, A., Li, J., & Gao, X. (2024) · Sensors · doi:10.3390/s24051555 · Cited by 17
“He et al. [22] improved detection accuracy using multi-scale feature diffusion.”
MethodAn Optical Lens Defect Detection Method for Micro Vision Based on WGSO-YOLO
Lin, Q. C., Takamasu, K., & Chen, M. (2024) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2024.3384267 · Cited by 20
“For instance, He et al. [11] proposed an end-to-end defect detection approach via fusing multiple hierarchical features, which get 0.823 mAP on NEU-DET dataset.”
MethodInfluentialA light-weight defect detection model for capacitor appearance based on the Yolov5
Xu, L., Xu, X., Xia, Q., Yao, Y., & Jiang, Z. (2024) · Measurement · doi:10.1016/j.measurement.2024.114717 · Cited by 16
Li, G., Peng, F., Wu, Z., Wang, S., & Xu, R. Y. D. (2024) · IEEE Sensors Journal · doi:10.1109/jsen.2024.3388714 · Cited by 17
“Three real industrial defect datasets, Pipe Weld Defect [24], Hot Rolled Strip Defect [25], and Aluminum Alloy Defect , are selected for experiments.”
BackgroundTian, S., Ma, H., Huang, P., Wang, X., Li, T., & Huang, R. (2024) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2024.108887 · Cited by 21
AEDN-YOLO: an efficient one-stage detection network for strip steel surface defects
Wei, M., Chen, B., Liu, J., Yuan, N., Liu, J., & Ji, Z. (2024) · Engineering Research Express · doi:10.1088/2631-8695/ad681d · Cited by 15
Li, L., Zhang, R., Xie, T., He, Y., Zhou, H., & Zhang, Y. (2024) · Electronics · doi:10.3390/electronics13183783 · Cited by 17
Self-citationSteel Sheet Defect Detection Based on Deep Learning Method
Zeng, W., You, Z., Huang, M., Kong, Z., Yu, Y., & Le, X. (2019) · International Conference on Intelligent Control and Information Processing · doi:10.1109/icicip47338.2019.9012199 · Cited by 19
“In [19], ResNet is used as baseline architecture to extract features on hierarchical levels, followed by region proposal network to extract region proposals.”
MethodEnvironment-Adaptable Printed-Circuit Board Positioning Using Deep Reinforcement Learning
Solórzano, C. A. P., & Tsai, D. (2022) · IEEE Transactions on Components Packaging and Manufacturing Technology · doi:10.1109/tcpmt.2022.3142033 · Cited by 16
Boudiaf, A., Harrar, K., Benlahmidi, S., Zaghdoudi, R., Ziani, S., & Taleb, S. (2022) · International Conference on Image and Signal Processing and their Applications (ISPA) · doi:10.1109/ispa54004.2022.9786361 · Cited by 18
Xuezhen, L. I., Jing, J., Bao, J., Lu, P., Xie, Y., & An, Y. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3280508 · Cited by 18
“Although supervised methods can effectively detect surface defects [2], [3], [4], they cannot deal well with the problem of the limited availability of marked negative samples, and the application of anomaly detection and location algorithms based on deep learning is becoming increasingly popular…”
BackgroundWei, C., Bao, Y., Zheng, C., & Ji, Z. (2025) · Journal of Intelligent Manufacturing · doi:10.1007/s10845-025-02613-5 · Cited by 19
“For example, He et al. (2019) proposed a multilevel-feature fusion network (MFN), integrating features from multiple CNN layers to improve localization accuracy.”
Dong, L., Zhu, H., Ren, H., Lin, T., & Lin, K.-P. (2025) · Expert Systems with Applications · doi:10.1016/j.eswa.2025.128247 · Cited by 15
Mende, H., Peters, A., Ibrahim, F., & Schmitt, R. H. (2022) · Procedia CIRP · doi:10.1016/j.procir.2022.05.254 · Cited by 20
Lin, L., Wang, Y., Zhao, S., Liu, J., Zhang, S., & Zhang, G. (2022) · Journal of Electronic Imaging · doi:10.1117/1.jei.31.6.063056 · Cited by 15
Wu, H., Li, B., Tian, L., Feng, J., & Dong, C. (2023) · The Visual Computer · doi:10.1007/s00371-023-02809-x · Cited by 14
“In contrast, computer vision-based detection methods have been proven a cost-effective solution in some defect inspection scenarios, such as the defect inspection of wind turbine blades[3–6]andthesteelsurfacedefectdetection[7,8].”
InfluentialAn Autoencoder with Convolutional Neural Network for Surface Defect Detection on Cast Components
Chamberland, O., Reckzin, M., & Hashim, H. A. (2023) · Journal of Failure Analysis and Prevention · doi:10.1007/s11668-023-01695-8 · Cited by 14
“[29] proposed the use of a CNNmodel for real-time strip steel surface defect detection.”
MethodWang, T., Li, Z., Xu, Y. W., Chen, J., Genovese, A., Piuri, V., & Scotti, F. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3315404 · Cited by 22
ADDet: An Efficient Multiscale Perceptual Enhancement Network for Aluminum Defect Detection
Zhu, J., Pang, Q., Li, S., Tian, S., Li, J., & Li, Y. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3343789 · Cited by 19
“Inspired by the idea of DDN using multilayer features for region proposal [11], MSEM adopts adaptive pooling to capture global context information.”
BackgroundMethodCombined Anomaly Aware Weakly Supervised Lightweight Model for Surface Defect Inspection
Zhang, Z., Shang, Z., Wang, X., & Ma, J. (2024) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2023.3348835 · Cited by 17
“He et al. [23] proposed a novel defect detection system focusing on steel plate defect inspection.”
BackgroundAutomotive Parts Defect Detection Based on YOLOv7
Huang, H., & Zhu, K. (2024) · Electronics · doi:10.3390/electronics13101817 · Cited by 18
FEB-YOLOv8: A Steel Surface Defect Detection Algorithm Based on Improved YOLOv8s
Yuan, J., Wan, F., Lei, G., et al. (2023) · International Conference on Computer Science, Electronic Information Engineering and Intelligent Control Technology (CEI) · doi:10.1109/cei60616.2023.10527815 · Cited by 14
“He et al [6] proposed a multi-level feature fusion network for steel defect detection algorithm, although the defect detection of this system can complete the task of steel surface defect detection, but it is slightly insufficient in the detection speed, it is difficult to meet the demand of…”
BackgroundTSEDNet:Task-specific encoder–decoder network for surface defects of strip steel
Guo, Y., Wei, J., & Feng, X. (2024) · Measurement · doi:10.1016/j.measurement.2024.115438 · Cited by 22
YOLOdrive: A Lightweight Autonomous Driving Single-Stage Target Detection Approach
Wang, L., Hua, S., Zhang, C., Yang, G., Ren, J., & Li, J. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3439863 · Cited by 14
“In [9], [10], and [11], convolutional neural networks are utilized to automatically learn the feature representation of an image and combined with methods, such as region proposal networks or anchor frames, to generate candidate regions, which are then classified and regressed to obtain the final…”
MethodYOLO-AFK: Advanced Fine-Grained Object Detection for Complex Solder Joints Defect
Wang, X., Xuan, Y., Huang, X., & Yan, Q. (2024) · IEEE Access · doi:10.1109/access.2024.3495540 · Cited by 16
DFP-YOLO: a lightweight machine tool workpiece defect detection algorithm based on computer vision
Shi, T., Ding, Y., Zhu, K., & Su, Y.-J. (2024) · The Visual Computer · doi:10.1007/s00371-024-03705-8 · Cited by 22
Multi-scale and dynamic snake convolution-based YOLOv9 for steel surface defect detection
Chen, J., Jin, W., Liu, Y., Huang, X., & Zhang, Y. (2025) · The Journal of Supercomputing · doi:10.1007/s11227-025-07036-w · Cited by 18
“For instance, He et al. [7] proposed an end-to-end defect detection network integrating multilevel features onto a single feature map with mean average precision (mAP) of 82.3% for steel surface defect detection on the NEU-DET data-set, which has a sizeable intra-class variation and a high…”
Influential
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.