[1]孙博言,王洪元,刘乾,等.基于多尺度和注意力机制的混合监督金属表面缺陷检测[J].智能系统学报,2023,18(4):886-893.[doi:10.11992/tis.202205042]
SUN Boyan,WANG Hongyuan,LIU Qian,et al.Hybrid supervised metal surface defect detection based on multi-scale and attention[J].CAAI Transactions on Intelligent Systems,2023,18(4):886-893.[doi:10.11992/tis.202205042]
点击复制
SUN Boyan,WANG Hongyuan,LIU Qian,et al.Hybrid supervised metal surface defect detection based on multi-scale and attention[J].CAAI Transactions on Intelligent Systems,2023,18(4):886-893.[doi:10.11992/tis.202205042]
基于多尺度和注意力机制的混合监督金属表面缺陷检测
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
18
期数:
2023年第4期
页码:
886-893
栏目:
吴文俊人工智能科学技术奖论坛
出版日期:
2023-07-15
- Title:
- Hybrid supervised metal surface defect detection based on multi-scale and attention
- Keywords:
- defect; detector; feature extraction; learnin galgorithm; learning system; image processing; metal; quality of product; deep learning
- 分类号:
- TP391
- 摘要:
- 针对缺陷检测中被检测样品中因缺陷目标形状各异引起的无法提取有效特征的问题，本文提出基于深度学习的缺陷检测模型。该模型使用改进后的多尺度特征融合模块，在控制计算量的基础上解决识别不同大小缺陷的问题。通过引入非局部注意力机制模块，模型对缺陷特征的提取能力得到加强；在训练中使用混合监督训练，探索模型所需要的标注量和检测准确度之间的关系。本文方法在KSDD、KSDD2和STEEL 3个数据集上都获得了比先进方法更好的精确度，对于不同类型的缺陷都能提取到有判别力的特征。与先进的完全监督方法和无监督方法相比，在数据集上精确度平均提高0.8%和11%。
- Abstract:
- Aiming at the problem in defect detection that effective features cannot be extracted due to different shapes of defect targets in the detected samples, this paper presents a defect detection model based on deep learning, which uses an improved multi-scale feature fusion module to solve the problem of identifying defects of different sizes on the basis of controlling the amount of calculation. By introducing a non-local attention mechanism module, the model’s ability of extracting defect features is enhanced. Furthermore, mixed-supervised training is used in training to explore the relationship between the amount of annotations required by the model and the detection accuracy. This method achieves better accuracy than the state-of-the-art methods on KSDD, KSDD2, and STEEL datasets, and can extract discriminative features for different types of defects. Compared with the state-of-the-art fully supervised and unsupervised methods, the average accuracy improvement on the dataset is 0.8% and 11%.
- 参考文献/References:
-
[1] 陶显, 侯伟, 徐德. 基于深度学习的表面缺陷检测方法综述[J]. 自动化学报, 2021, 47(5): 1017–1034
TAO Xian, HOU Wei, XU De. A survey of surface defect detection methods based on deep learning[J]. Acta automatica sinica, 2021, 47(5): 1017–1034
[2] TABERNIK D, ?ELA S, SKVAR? J, et al. Segmentation-based deep-learning approach for surface-defect detection[J]. Journal of intelligent manufacturing, 2020, 31(3): 759–776.
[3] KUMAR A. Computer-vision-based fabric defect detection: a survey[J]. IEEE transactions on industrial electronics, 2008, 55(1): 348–363.
[4] ABDI H, WILLIAMS L J. Principal component analysis[J]. Wiley interdisciplinary reviews:computational statistics, 2010, 2(4): 433–459.
[5] 虞祖耀, 王洪元, 张继. 基于机器视觉的织布瑕疵在线检测[J]. 计算机工程与设计, 2016, 37(10): 2851–2856
YU Zuyao, WANG Hongyuan, ZHANG Ji. On-line detection of weaving defects based on machine vision[J]. Computer engineering and design, 2016, 37(10): 2851–2856
[6] FUKUSHIMA K, MIYAKE S. Neocognitron: a self-organizing neural network model for a mechanism of visual pattern recognition[M]. Berlin: Springer Berlin Heidelberg, 1982: 267?285.
[7] PARK J K, KWON B K, PARK J H, et al. Machine learning-based imaging system for surface defect inspection[J]. International journal of precision engineering and manufacturing-green technology, 2016, 3(3): 303–310.
[8] KYEONG K, KIM H. Classification of mixed-type defect patterns in wafer Bin maps using convolutional neural networks[J]. IEEE transactions on semiconductor manufacturing, 2018, 31(3): 395–402.
[9] HE Yu, SONG Kechen, MENG Qinggang, et al. An end-to-end steel surface defect detection approach via fusing multiple hierarchical features[J]. IEEE transactions on instrumentation and measurement, 2020, 69(4): 1493–1504.
[10] LI Feng, XI Qinggang. DefectNet: toward fast and effective defect detection[J]. IEEE transactions on instrumentation and measurement, 2021, 70: 1–9.
[11] GE Ce, WANG Jing, WANG Jingyu, et al. Towards automatic visual inspection: a weakly supervised learning method for industrial applicable object detection[J]. Computers in industry, 2020, 121: 103232.
[12] LI Qizhu, ARNAB A, TORR P H S. Weakly- and semi-supervised panoptic segmentation[M]. Computer Vision - ECCV 2018. Cham: Springer International Publishing, 2018: 106?124.
[13] SALEH F, ALIAKBARIAN M S, SALZMANN M, et al. Built-in foreground/background prior for weakly-supervised semantic segmentation[M]. Computer Vision-ECCV 2016. Cham: Springer International Publishing, 2016: 413?432.
[14] MINHAS M S, ZELEK J. Semi-supervised anomaly detection using autoencoders[J]. Journal of computational vision and imaging systems, 2019: 03674.
[15] BERGMANN P, L?WE S, FAUSER M, et al. Improving unsupervised defect segmentation by applying structural similarity to autoencoders[EB/OL]. (2018-07-05)[2022-05-25]. https://arxiv.org/abs/1807.02011.
[16] 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: IEEE, 2020: 4182?4191.
[17] XIN Zihao, WANG Hongyuan, QI Pengyu, et al. Printed surface defect detection model based on positive samples[J]. Computers, materials & continua, 2022, 72(3): 5925–5938.
[18] SCHLEGL T, SEEBOCK P, WALDSTEIN S M, et al. F-AnoGAN: fast unsupervised anomaly detection with generative adversarial networks[J]. Medical image analysis, 2019, 54: 30–44.
[19] BOZIC J, TABERNIK D, SKOCAI D. Mixed supervision for surface-defect detection: from weakly to fully supervised learning[J]. Computers in industry, 2021, 129: 103459.
[20] MNIH V, HEESS N, GRAVES A, et al. Recurrent models of visual attention[EB/OL]. (2014-06-24)[2022-05-25]. https://arxiv.org/abs/1406.6247.
[21] XU K, BA J L, KIROS R, et al. Show, attend and tell: neural image caption generation with visual attention[C]//Proceedings of the 32nd International Conference on International Conference on Machine Learning-Volume 37. New York: ACM, 2015: 2048?2057.
[22] WANG Xiaolong, GIRSHICK R, GUPTA A, et al. Non-local neural networks[C]// IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2018: 7794?7803.
[23] DONG Hongwen, SONG Kechen, HE Yu, et al. PGA-net: pyramid feature fusion and global context attention network for automated surface defect detection[J]. IEEE transactions on industrial informatics, 2020, 16(12): 7448–7458.
- 相似文献/References:
-
[1]魏文戈,谭晓阳.密集堆叠下的高相似度木块横截面检测[J].智能系统学报,2019,14(4):642.[doi:10.11992/tis.201806001]
WEI Wenge,TAN Xiaoyang.Highly similar wood blocks detection under dense stacking[J].CAAI Transactions on Intelligent Systems,2019,14():642.[doi:10.11992/tis.201806001]
备注/Memo
收稿日期:2022-05-25。
基金项目:国家自然科学基金项目(61976028)；2022年江苏省研究生科研创新计划（KYCX22_3066）.
作者简介:孙博言，硕士研究生，主要研究方向为计算机视觉、缺陷检测;王洪元，教授，主要研究方向为计算机视觉、模式识别和智能系统，江苏省人工智能专委会常务理事，常州市计算机学会副理事长，CCF常州副主席。主持国家自然科学基金项目3项，获吴文俊人工智能科技进步三等奖。发表学术论文150余篇，出版专 著1部、教材3部;刘乾，硕士研究生，主要研究方向为计算机视觉
通讯作者:王洪元.E-mail:hywang@cczu.edu.cn
更新日期/Last Update:
1900-01-01