Attention mechanism and texture contextual information for steel plate defects detection

被引:7
|
作者
Zhang, Chi [1 ]
Cui, Jian [1 ]
Wu, Jianguo [1 ]
Zhang, Xi [1 ]
机构
[1] Peking Univ, Dept Ind Engn & Management, Beijing, Peoples R China
关键词
Statistical texture feature fusion; Contextual information mining; Attention mechanism; Steel surface defect detection; CLASSIFICATION;
D O I
10.1007/s10845-023-02149-6
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
In order to achieve rapid inference and generalization results, the majority of Convolutional Neural Network (CNN) based semantic segmentation models strive to mine high-level features that contain rich contextual semantic information. However, at steel plate defects detection scenario, some background textures' noises are similar to the foreground leading to hard distinguishment, which will significantly interfere with feature extraction. Texture features themselves often hold the most plentiful contextual information. Despite this, semantic segmentation tasks rarely take texture features into account when identifying surface defects on steel plates. In that case, the essential details, such as the edge texture and other intuitive low-level features, will generally cannot be included into the final feature map. To address the problems of inefficient accuracy and slow speed of existing detection, this study proposed a steel plate surface defect detection method using contextual information and attention mechanism, and utilizes a multi-layer feature extraction method and fusion framework based on low-level statistical textures. Through the identification of pixel-level spatial and correlation relationships, characteristics of low-level defects are extracted. Furthermore, to effectively incorporate statistical texture in CNN, a novel quantization technique has been developed. This quantization method allows for the conversion of continuous texture into various levels of intensity. The network parameters were iterated in a gradient direction, facilitating the defects division. Empirical results have demonstrated the feasibility of applying the proposed approach to practical steel plate testing. Additionally, ablation experiments have demonstrated that the method is capable of effectively enhancing surface defect detection for steel plates, resulting in industry-leading performance.
引用
收藏
页码:2193 / 2214
页数:22
相关论文
共 50 条
  • [41] Quantifying Contextual Information for Object Detection
    Zheng, Wei-Shi
    Gong, Shaogang
    Xiang, Tao
    2009 IEEE 12TH INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2009, : 932 - 939
  • [42] Visual relationship detection with contextual information
    Li Y.
    Wang Y.
    Chen Z.
    Zhu Y.
    Computers, Materials and Continua, 2020, 63 (03): : 1575 - 1589
  • [43] Saliency detection for strip steel surface defects using multiple constraints and improved texture features
    Song, Guorong
    Song, Kechen
    Yan, Yunhui
    OPTICS AND LASERS IN ENGINEERING, 2020, 128
  • [44] Towards the steel plate defect detection: Multidimensional feature information extraction and fusion
    Hao, Zhiqiang
    Wang, Zhigang
    Bai, Dongxu
    Zhou, Shiyang
    CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE, 2021, 33 (21):
  • [45] YOLO-SLD: An Attention Mechanism-Improved YOLO for License Plate Detection
    Chung, Ming-An
    Lin, Yu-Jou
    Lin, Chia-Wei
    IEEE ACCESS, 2024, 12 : 89035 - 89045
  • [46] On the real-time information processing in the detection of surface defects in steel strips
    Lu, Yimin
    Weng, Hong
    Luo, Zhiyong
    Huazhong Ligong Daxue Xuebao/Journal Huazhong (Central China) University of Science and Technology, 1994, 22 (11):
  • [47] Typical defects in plate and long steel products
    Papaefthymiou, Spyros
    Tzevelekou, Theofani
    Antonopoulos, Alexandros
    Gypakis, Antonios
    INTERNATIONAL JOURNAL OF STRUCTURAL INTEGRITY, 2016, 7 (05) : 645 - 655
  • [48] Research on defect detection algorithm of complex texture ceramic tiles based on visual attention mechanism
    Ouyang Z.
    Zhang H.
    Tang Z.
    Peng L.
    Yu S.
    Xibei Gongye Daxue Xuebao/Journal of Northwestern Polytechnical University, 2022, 40 (02): : 414 - 421
  • [49] Detection of Weft Knitting Fabric Defects Based on Windowed Texture Information And Threshold Segmentation by CNN
    Sun Yao
    Long Hai-ru
    ICDIP 2009: INTERNATIONAL CONFERENCE ON DIGITAL IMAGE PROCESSING, PROCEEDINGS, 2009, : 292 - 296
  • [50] A fundamental study on detection of defects in the web gap region of steel plate girder bridges by the plate wave ultrasonic technique
    Shirata, H
    Greimann, L
    Wipf, T
    Pharcs, B
    Nakagawa, N
    REVIEW OF PROGRESS IN QUANTITATIVE NONDESTRUCTIVE EVALUATION, VOLS 23A AND 23B, 2004, 23 : 1050 - 1057