Defect Detection Based on Attention Mechanism and Atrous Pyramid Pooling

WEI Jinyang; YUAN Mingzhe; CAO Feidao; BAI Haijun; WANG Wenhong

doi:10.13976/j.cnki.xk.2024.3148

Volume 53 Issue 5

Oct. 2024

Turn off MathJax

Article Contents

Abstract

References

INFORMATION AND CONTROL > 2024 > 53(5): 662-672. > DOI: 10.13976/j.cnki.xk.2024.3148 CSTR: 32166.14.xk.2024.3148

WEI Jinyang, YUAN Mingzhe, CAO Feidao, BAI Haijun, WANG Wenhong. Defect Detection Based on Attention Mechanism and Atrous Pyramid Pooling[J]. INFORMATION AND CONTROL, 2024, 53(5): 662-672. DOI: 10.13976/j.cnki.xk.2024.3148

Citation:

PDF (2348 KB)

Defect Detection Based on Attention Mechanism and Atrous Pyramid Pooling

1.
College of Information Engineering, Shenyang University of Chemical Technology, Shenyang 110142, China
2.
Guangzhou Industrial Intelligence Research Institute, Guangzhou 511458, China
3.
Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang 110016, China

More Information

Received Date: May 03, 2023
Revised Date: July 04, 2023
Accepted Date: June 02, 2023

Graphical Abstract

Abstract

Abstract

To solve the problem of low feature reconstruction accuracy in industrial product surface defect detection, which leads to high false positive rates at the image, pixel, and region levels, we propose an unsupervised defect detection method with improved deep feature reconstruction (DFR). First, we introduce jump connections into the feature reconstruction process to improve the feature reconstruction accuracy and the ability of the model to reconstruct positive sample features. Second, we introduce an attention mechanism to improve the attention of the algorithm to defect regions and explore the effect of spatial attention on defect detection for different targets. Third, we introduce atrous pyramid pooling into the feature reconstruction module to capture context at multiple scales without increasing the number of parameters to improve the ability of the model to detect defects at different scales. Finally, we use the L₂-SSIM loss function to constrain feature reconstruction to preserve the feature structure while maintaining pixel similarity. The proposed algorithm achieves 97.5%, 97.2%, and 93.1% detection accuracy at the image, pixel, and region levels, respectively, outperforming the comparison algorithm.
- image processing,
- defect detection,
- unsupervised learning,
- attention mechanism,
- atrous pyramid pooling

FullText(HTML)

References (26)

References

[1]	KUANG H L, DING Y R, LI R F, et al. Defect detection of bamboo strips based on LBP and GLCM features by using SVM classifier[C]//Chinese Control and Decision Conference. Piscataway, USA: IEEE, 2018: 3341-3345.
[2]	包从望, 胡才梦, 张彩红, 等. 基于改进Canny算子的齿轮缺陷检测边缘检测算法[J]. 组合机床与自动化加工技术, 2023, 587(1): 83-86, 91. BAO C W, HU C M, ZHANG C H, et al. Edge detection algorithm of gear defect detection based on improved Canny operator[J]. Modular Machine Tool & Automatic Manufacturing Technique, 2023, 587(1): 83-86, 91.
[3]	黄梦涛, 连一鑫. 基于改进Canny算子的锂电池极片表面缺陷检测[J]. 仪器仪表学报, 2021, 42(10): 199-209. HUANG M T, LIAN Y X. Lithium battery electrode plate surface defect detection based on improved Canny operator[J]. Chinese Journal of Scientific Instrument, 2021, 42(10): 199-209.
[4]	GE Q C, FANG M, XU J. Defect detection of industrial products based on improved Hough transform[C]//IEEE International Conference on Mechatronics and Automation. Piscataway, USA: IEEE, 2018: 832-836.
[5]	ZHOU Q Q, SITU Z X, TENG S, et al. Automatic sewer defect detection and severity quantification based on pixel-level semantic segmentation[J/OL]. Tunnelling and Underground Space Technology, 2022, 123[2023-05-01]. https://www.sciencedirect.com/science/article/pii/S0886779822000438. DOI: 10.1016/j.tust.2022.104403.
[6]	RONNEBERGER O, FISCHER P, BROX T. U-Net: Convolutional networks for biomedical image segmentation[C]//International Conference on Medical Image Computing and Computer-assisted Intervention. Berlin, Germany: Springer, 2015: 234-241.
[7]	HE K M, GKIOXARI G, DOLLáR P, et al. Mask R-CNN[C]//IEEE International Conference on Computer Vision. Piscataway, USA: IEEE, 2017: 2980-2988.
[8]	LI C L, SOHN K, YOON J, et al. CutPaste: Self-supervised learning for anomaly detection and localization[C]//IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway, USA: IEEE, 2021: 9664-9674.
[9]	YANG H, CHEN Y, SONG K, et al. Multiscale feature-clustering-based fully convolutional autoencoder for fast accurate visual inspection of texture surface defects[J]. IEEE Transactions on Automation Science and Engineering, 2019, 16(3): 1450-1467. doi: 10.1109/TASE.2018.2886031
[10]	MEI H Y Z. An unsupervised-learning-based approach for automated defect inspection on textured surfaces[J]. IEEE Transactions on Instrumentation and Measurement, 2018, 67(6): 1266-1277. doi: 10.1109/TIM.2018.2795178
[11]	GONG D, LIU L, LE V, et al. Memorizing normality to detect anomaly: Memory-augmented deep autoencoder for unsupervised anomaly detection[C/OL]//IEEE/CVF International Conference on Computer Vision. Piscataway, USA: IEEE, 2019[2023-04-26]. https://ieeexplore.ieee.org/document/9010977. DOI: 10.1109/ICCV.2019.00179.
[12]	AKAY S, ATAPOUR-ABARGHOUEI A, BRECKON T P. Skip-GANomaly: Skip connected and adversarially trained encoder-decoder anomaly detection[C/OL]//International Joint Conference on Neural Networks. Piscataway, USA: IEEE, 2019[2023-04-20]. https://ieeexplore.ieee.org/document/8851808. DOI: 10.1109/IJCNN.2019.8851808.
[13]	BAUR C, WIESTLER B, Albarqouni S, et al. Deep autoencoding models for unsupervised anomaly segmentation in brain MR images[M]//Lecture Notes in Computer Science: vol. 11383. Berlin, Germany: Springer-Verlag, 2019: 161-169.
[14]	YANG J, SHI Y, QI Z Q. Unsupervised anomaly segmentation via deep feature reconstruction[J]. Neurocomputing, 2021, 424(1): 9-22.
[15]	SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale image recognition[EB/OL]. (2015-04-10)[2023-04-15]. https://arxiv.org/abs/1409.1556. DOI: 10.48550/arXiv.1409.1556
[16]	蒋羽. 面向工业应用场景的深度学习缺陷检测方法研究[D]. 杭州: 浙江大学, 2022. JIANG Y. Research on deep learning defect detection methods for industrial application scenarios[D]. Hangzhou: Zhejiang University, 2022.
[17]	杨胜雄, 陈莹. 基于残差池化自编码机的无监督异常分割[J/OL]. 计算机集成制造系统[2023-04-10]. http://kns.cnki.net/kcms/detail/11.5946.TP.20220328.1632.011.html. YANG S H, CHEN Y. Unsupervised anomaly segmentation based on residual max-pooling autoencoder[J/OL]. Computer Integrated Manufacturing Systems[2023-04-10]. http://kns.cnki.net/kcms/detail/11.5946.TP.20220328.1632.011.html.
[18]	LIN T Y, DOLLAR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]//IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, USA: IEEE, 2017[2023-03-29]. https://ieeexplore.ieee.org/document/8099589. DOI: 10.1109/CVPR.2017.106.
[19]	CHEN L C, MURPHY K, KOKKINOS I, et al. DeepLab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully Connected CRFs[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 40(4): 834-848.
[20]	WOO S, PARK J, LEE J Y, et al. CBAM: Convolutional block attention module[C]//European Conference on Computer Vision. Berlin, Germany: Springer, 2018: 3-19.
[21]	JIE H, LI S, GANG S. Squeeze-and-excitation networks[C/OL]//IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway, USA: IEEE, 2018[2023-04-15]. https://ieeexplore.ieee.org/document/8578843. DOI: 10.1109/CVPR.2018.00745.
[22]	BERGMANN P, FAUSER M, SATTLEGGER D, et al. MVTec AD — A comprehensive real-world dataset for unsupervised anomaly detection[C/OL]//IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway, USA: IEEE, 2020[2023-04-08]. https://ieeexplore.ieee.org/document/8954181. DOI: 10.1109/CVPR.2019.00982.
[23]	BERGMANN P, FAUSER M, SATTLEGGER D, et al. Uninformed students: Student-teacher anomaly detection with discriminative latent embeddings[C]//IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway, USA: IEEE, 2020: 4183-4192
[24]	COHEN N, HOSHEN Y. Sub-Image anomaly detection with deep pyramid correspondences[EB/OL]. (2021-02-03)[2023-03-27]. https://arxiv.org/abs/2005.02357. DOI: 10.48550/arXiv.2005.02357.
[25]	YI J, YOON S. Patch SVDD: Patch-level SVDD for anomaly detection and segmentation[C]//Asian Conference on Computer Vision. Berlin, Germany: Springer, 2021: 375-390.
[26]	MVTec. The MVTec anomaly detection dataset[DS/OL]. [2023-03-05]. https://www.mvtec.com/company/research/datasets/mvtec-ad.

[1]	QIU Yunfei, WANG Yueyang. Automatic Liver Tumor Segmentation Method Integrating Multi-branch Texture Feature Extraction and Attention Mechanism[J]. INFORMATION AND CONTROL, 2024, 53(5): 673-688. DOI: 10.13976/j.cnki.xk.2024.3203
[2]	CHEN Xiang, LIU Qinming, HU Jiarui. Bearing Fault Diagnosis and Life Prediction Based on Attention Mechanism and Long Short-term Memory Network under Multi-source Sensor Data[J]. INFORMATION AND CONTROL, 2024, 53(2): 211-225. DOI: 10.13976/j.cnki.xk.2023.3056
[3]	HUANG Zhaoji, GAO Junli, TANG Zhaonian, SONG Haitao, GUO Jing. Slip Detection for Robot Grasping Based on Attention Mechanism and Visuo-Tactile Fusion[J]. INFORMATION AND CONTROL, 2024, 53(2): 191-198. DOI: 10.13976/j.cnki.xk.2023.2598
[4]	QIU Jier, TAO Hongfeng, CHENG Long, SHENG Linzhi. Bearing Fault Diagnosis Based on Self-attention Mechanism ACGAN[J]. INFORMATION AND CONTROL, 2022, 51(6): 753-762. DOI: 10.13976/j.cnki.xk.2022.2002
[5]	FU Shuangjie, CHEN Wei, YIN Zhong. Semantic Segmentation Algorithm Combining Self-attention and Feature Adaptive Fusion[J]. INFORMATION AND CONTROL, 2022, 51(6): 680-687, 698. DOI: 10.13976/j.cnki.xk.2022.1584
[6]	JIA Lin, LI Qi, LIANG Dong. Bridge Crack Location Algorithm Based on Improved Fully Convolutional One-stage Detector[J]. INFORMATION AND CONTROL, 2022, 51(3): 369-376. DOI: 10.13976/j.cnki.xk.2022.1221
[7]	LI Panpan, WANG Chaoli, SUN Zhanquan. Face Liveness Detection Based on Multi-feature Fusion with an Attention Mechanism[J]. INFORMATION AND CONTROL, 2021, 50(5): 631-640. DOI: 10.13976/j.cnki.xk.2021.0581
[8]	XIA Limin, LIU Tao, XIAO Yongliang. High-speed Railway Track Damage Detection Based on the Model of Visual Attention[J]. INFORMATION AND CONTROL, 2015, 44(3): 353-358. DOI: 10.13976/j.cnki.xk.2015.0353
[9]	JIANG Shu-juan, LIU Wei-jun. Study on Real-time Measurement for Laser Molten Pool Temperature Field by Images Colorimetric Method[J]. INFORMATION AND CONTROL, 2008, 37(6): 747-750,756.
[10]	ZHOU Ruixia, . A NEW KIND OF PYRAMID LINKING AND ITS APPLICATION IN IMAGE SEGMENTATION[J]. INFORMATION AND CONTROL, 1989, 18(5): 26-29.