Article
Mathematics
Xinglong Feng, Xianwen Gao, Ling Luo
Summary: This paper proposes a deep learning model for strip steel defect classification, achieving a high classification accuracy by introducing FcaNet and CBAM technologies. Additionally, through ensemble learning optimization, the recognition rate of oxide scale defects and overall defect classification accuracy were significantly improved.
Article
Materials Science, Multidisciplinary
Yichuan Shao, Shuo Fan, Haijing Sun, Zhenyu Tan, Ying Cai, Can Zhang, Le Zhang
Summary: Defect classification is crucial in steel surface defect detection. Traditional methods using convolutional neural networks (CNNs) improve accuracy by increasing network depth and parameter count. However, this approach overlooks the memory overhead and diminishing accuracy gains. To address these issues, a multi-scale lightweight neural network model (MM) is proposed, which uses a fusion encoding module and Gaussian difference pyramid. Experimental results show that MM network achieves 98.06% accuracy in defect classification, surpassing other networks in both parameter reduction and accuracy.
Article
Engineering, Multidisciplinary
Wenli Zhao, Kechen Song, Yanyan Wang, Shubo Liang, Yunhui Yan
Summary: This paper proposes a feature-aware network (FaNet) for few shot defect classification, which can effectively distinguish new classes with a small number of labeled samples. In FaNet, ResNet12 is used as the baseline, and the feature-attention convolution module (FAC) is applied to extract comprehensive feature information from the base classes. An online feature-enhance integration module (FEI) is adopted during the test phase to average the noise from defect images, further enhancing image features among different tasks. In addition, a large-scale strip steel surface defects few shot classification dataset (FSC-20) with 20 different types is constructed. Experimental results show that the proposed method achieves the best performance compared to state-of-the-art methods for 5-way 1-shot and 5-way 5-shot tasks. The dataset and code are available at: https://github.com/VDT-2048/FSC-20.
Article
Computer Science, Information Systems
Shunfeng Li, Chunxue Wu, Naixue Xiong
Summary: Surface defects in strip steel are common and pose hidden risks. Accurate classification is crucial, and the proposed hybrid network architecture (CNN-T) combining CNN and Transformer encoder achieves improved performance compared to pure Transformer networks and CNNs.
Article
Metallurgy & Metallurgical Engineering
Mao-xiang Chu, Yao Feng, Yong-hui Yang, Xin Deng
Summary: The study proposed an anti-noise multi-class classification method for steel surface defects classification. The method involved constructing ASVHs classifier that is insensitive to feature and label noise, and pruning defect samples to enhance efficiency and accuracy. Experimental results demonstrated high efficiency and accuracy for corrupted defect samples in steel surface.
JOURNAL OF IRON AND STEEL RESEARCH INTERNATIONAL
(2021)
Article
Computer Science, Artificial Intelligence
Jiang Chang, Shengqi Guan
Summary: This paper proposes an image generation model called CH-GANs to solve the problem of dataset expansion in deep learning tasks. The model highlights image categories and generates true-to-life images with clear categories. A novel discriminator that can judge both the authenticity and the image category is designed. The lightweight image classification network GhostNet is improved and trained on the dataset expanded by CH-GAN for accurate classification of strip steel defects. Experimental results demonstrate the effectiveness and superiority of the proposed methods in image generation and classification.
INTERNATIONAL JOURNAL OF PATTERN RECOGNITION AND ARTIFICIAL INTELLIGENCE
(2022)
Article
Chemistry, Analytical
Cancan Yi, Qirui Chen, Biao Xu, Tao Huang
Summary: Due to the shortage of defect samples and high cost of labelling, it is difficult to obtain diverse defect data in hot-rolled strip production, affecting the accuracy of defect identification on steel surface. This paper proposes the SDE-ConSinGAN model, which uses GAN to train a single-image model and implements image-feature cutting and splicing. The model adjusts the number of iterations dynamically to reduce training time and highlights detailed defect features of training samples. Experimental results show that the generated defect images have higher quality and more diversity compared to current methods.
Article
Engineering, Electrical & Electronic
Wenqi Cui, Kechen Song, Hu Feng, Xiujian Jia, Shaoning Liu, Yunhui Yan
Summary: Researchers propose a novel autocorrelation-aware aggregation network (A3Net) for salient object detection of strip steel surface defects. The use of attention mechanism and scale interaction module contributes to the superior performance of the proposed method on both public and newly built datasets.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2023)
Article
Engineering, Electrical & Electronic
Xinglong Feng, Ling Luo, Xianwen Gao
Summary: The detection of surface defects in hot-rolled strip steel is crucial for ensuring final quality. Current research has primarily focused on modifying models, but the proposed SDDA data augmentation method improves model accuracy without increasing inference time.
JOURNAL OF ELECTRONIC IMAGING
(2022)
Article
Chemistry, Multidisciplinary
Huaping Guo, Shanggui Zhan, Li Zhang, Wenbo Zhu, Yange Sun, Jing Wang
Summary: Strip steel is a crucial material for industries like aerospace, shipbuilding, and pipelines, and any defects in the strip steel can lead to significant economic losses. Detecting these defects is challenging due to the complex variations in strip steel. This paper proposes a novel method based on a U-shaped residual network, utilizing attention mechanisms in the encoder to extract multi-scale defect features and a decoder to capture contextual data. Experimental results demonstrate that this method effectively segments surface defect objects with clear boundaries compared to other advanced techniques.
APPLIED SCIENCES-BASEL
(2022)
Article
Engineering, Electrical & Electronic
Xinglong Feng, Xianwen Gao, Ling Luo
Summary: This paper proposes a new dataset XLData-CLS for detecting surface defects in strip steel, and achieves high classification accuracy by improving the feature extraction layer of the model for high-resolution images.
JOURNAL OF ELECTRONIC IMAGING
(2022)
Article
Chemistry, Multidisciplinary
Shiqing Wu, Shiyu Zhao, Qianqian Zhang, Long Chen, Chenrui Wu
Summary: In this paper, a method combining feature extraction, feature transformation, and nearest neighbors is proposed to classify steel surface defects, achieving significant progress in addressing the degradation problem caused by network deepening.
APPLIED SCIENCES-BASEL
(2021)
Article
Multidisciplinary Sciences
Xinglong Feng, Xianwen Gao, Ling Luo
Summary: This study proposed a new hot rolled steel strip defect dataset X-SDD for the actual detection of defects on the surface of hot rolled steel strip. Various algorithms were tested on X-SDD, with the results showing that the proposed algorithm achieved high accuracy and outperformed other comparable algorithms.
Article
Engineering, Multidisciplinary
Rongqiang Liu, Min Huang, Zheming Gao, Zhenyuan Cao, Peng Cao
Summary: In this paper, a defect detection module called MSC-DNet is proposed to localize and classify surface defects. The MSC-DNet utilizes a parallel structure of dilated convolution to capture the multi-scale context information of defects. A feature enhancement and selection module is also introduced to reduce confusing information. Experimental results show that the proposed MSC-DNet achieves high accuracy on benchmark datasets, meeting the quasi-real-time requirement.
Article
Metallurgy & Metallurgical Engineering
Wenyan Wang, Ziheng Wu, Kun Lu, Hongming Long, Dan Li, Jun Zhang, Peng Chen, Bing Wang
Summary: This study proposes an improved deep learning model for high accuracy surface defect classification of hot-rolled steel strip with only a few labeled samples. Through a transductive learning algorithm and feature fusion technique, the model can adapt to the needs of unknown samples and extract more sample information.
ISIJ INTERNATIONAL
(2022)
Article
Automation & Control Systems
Maoxiang Chu, Xiaoping Liu, Rongfen Gong, Liming Liu
CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS
(2018)
Article
Automation & Control Systems
Rongfen Gong, Chengdong Wu, Maoxiang Chu
CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS
(2018)
Article
Metallurgy & Metallurgical Engineering
Mao-xiang Chu, Xiao-ping Liu, Rong-fen Gong, Jie Zhao
JOURNAL OF IRON AND STEEL RESEARCH INTERNATIONAL
(2018)
Article
Mathematics, Interdisciplinary Applications
Chuang Gao, Minggang Shen, Xiaoping Liu, Lidong Wang, Maoxiang Chu
Article
Multidisciplinary Sciences
Maoxiang Chu, Xiaoping Liu, Rongfen Gong, Jie Zhao
Article
Automation & Control Systems
Rongfen Gong, Maoxiang Chu, Yonghui Yang, Yao Feng
CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS
(2019)
Article
Metallurgy & Metallurgical Engineering
Chuang Gao, Ming-gang Shen, Xiao-ping Liu, Nan-nan Zhao, Mao-xiang Chu
JOURNAL OF IRON AND STEEL RESEARCH INTERNATIONAL
(2020)
Article
Computer Science, Artificial Intelligence
Liming Liu, Maoxiang Chu, Rongfen Gong, Xinyu Qi
JOURNAL OF INTELLIGENT & FUZZY SYSTEMS
(2020)
Article
Computer Science, Artificial Intelligence
Liming Liu, Maoxiang Chu, Yonghui Yang, Rongfen Gong
INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS
(2020)
Article
Computer Science, Artificial Intelligence
Liming Liu, Maoxiang Chu, Rongfen Gong, Yongcheng Peng
PATTERN RECOGNITION
(2020)
Article
Computer Science, Artificial Intelligence
Liming Liu, Maoxiang Chu, Rongfen Gong, Li Zhang
Summary: The improved nonparallel support vector machine (INPSVM) proposed in this article inherits the advantages of nonparallel support vector machine (NPSVM) while also offering incomparable benefits over twin support vector machine (TSVM). INPSVM effectively eliminates noise effects and achieves higher classification accuracy for both linear and nonlinear datasets compared to other algorithms. Experimental results demonstrate the superior efficiency, accuracy, and robustness of INPSVM.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2021)
Article
Metallurgy & Metallurgical Engineering
Mao-xiang Chu, Yao Feng, Yong-hui Yang, Xin Deng
Summary: The study proposed an anti-noise multi-class classification method for steel surface defects classification. The method involved constructing ASVHs classifier that is insensitive to feature and label noise, and pruning defect samples to enhance efficiency and accuracy. Experimental results demonstrated high efficiency and accuracy for corrupted defect samples in steel surface.
JOURNAL OF IRON AND STEEL RESEARCH INTERNATIONAL
(2021)
Article
Metallurgy & Metallurgical Engineering
Li-ming Liu, Mao-xiang Chu, Rong-fen Gong, Xin-yu Qi
JOURNAL OF IRON AND STEEL RESEARCH INTERNATIONAL
(2020)
Article
Computer Science, Information Systems
Maoxiang Chu, Liming Liu, Yonghui Yang, Rongfen Gong
Proceedings Paper
Automation & Control Systems
Maoxiang Chu, Jie Zhao, Rongfen Gong, Liming Liu
2017 29TH CHINESE CONTROL AND DECISION CONFERENCE (CCDC)
(2017)