{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,26]],"date-time":"2026-06-26T16:57:43Z","timestamp":1782493063867,"version":"3.54.5"},"reference-count":58,"publisher":"MDPI AG","issue":"7","license":[{"start":{"date-parts":[[2020,4,1]],"date-time":"2020-04-01T00:00:00Z","timestamp":1585699200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"National Natural Science Founda- ion of China","award":["61421004"],"award-info":[{"award-number":["61421004"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Sensors"],"abstract":"<jats:p>The advent of convolutional neural networks (CNNs) has accelerated the progress of computer vision from many aspects. However, the majority of the existing CNNs heavily rely on expensive GPUs (graphics processing units). to support large computations. Therefore, CNNs have not been widely used to inspect surface defects in the manufacturing field yet. In this paper, we develop a compact CNN-based model that not only achieves high performance on tiny defect inspection but can be run on low-frequency CPUs (central processing units). Our model consists of a light-weight (LW) bottleneck and a decoder. By a pyramid of lightweight kernels, the LW bottleneck provides rich features with less computational cost. The decoder is also built in a lightweight way, which consists of an atrous spatial pyramid pooling (ASPP) and depthwise separable convolution layers. These lightweight designs reduce the redundant weights and computation greatly. We train our models on groups of surface datasets. The model can successfully classify\/segment surface defects with an Intel i3-4010U CPU within 30 ms. Our model obtains similar accuracy with MobileNetV2 while only has less than its 1\/3 FLOPs (floating-point operations per second) and 1\/8 weights. Our experiments indicate CNNs can be compact and hardware-friendly for future applications in the automated surface inspection (ASI).<\/jats:p>","DOI":"10.3390\/s20071974","type":"journal-article","created":{"date-parts":[[2020,4,2]],"date-time":"2020-04-02T11:57:14Z","timestamp":1585828634000},"page":"1974","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":69,"title":["A Compact Convolutional Neural Network for Surface Defect Inspection"],"prefix":"10.3390","volume":"20","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-8517-1068","authenticated-orcid":false,"given":"Yibin","family":"Huang","sequence":"first","affiliation":[{"name":"Institute of Automation, Chinese Academy of Sciences, University of Chinese Academy of Sciences, Beijing 100190, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Congying","family":"Qiu","sequence":"additional","affiliation":[{"name":"Civil Engineering &amp; Engineering Mechanics Department, Columbia University, New York, NY 10024, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiaonan","family":"Wang","sequence":"additional","affiliation":[{"name":"Institute of Automation, Chinese Academy of Sciences, University of Chinese Academy of Sciences, Beijing 100190, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shijun","family":"Wang","sequence":"additional","affiliation":[{"name":"Institute of Automation, Chinese Academy of Sciences, University of Chinese Academy of Sciences, Beijing 100190, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kui","family":"Yuan","sequence":"additional","affiliation":[{"name":"Institute of Automation, Chinese Academy of Sciences, University of Chinese Academy of Sciences, Beijing 100190, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2020,4,1]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"858","DOI":"10.1016\/j.apsusc.2013.09.002","article-title":"A noise robust method based on completed local binary patterns for hot-rolled steel strip surface defects","volume":"285","author":"Song","year":"2013","journal-title":"Appl. Surf. Sci."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"1600068","DOI":"10.1002\/srin.201600068","article-title":"An End-to-End Steel Strip Surface Defects Recognition System Based on Convolutional Neural Networks","volume":"88","author":"Yi","year":"2017","journal-title":"Steel Res. Int."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1080\/15599612.2018.1444829","article-title":"High-resolution optical inspection system for fast detection and classification of surface defects","volume":"12","author":"Ye","year":"2018","journal-title":"Int. J. Optomech."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"4141","DOI":"10.1007\/s11042-015-3041-3","article-title":"Fabric defect inspection using prior knowledge guided least squares regression","volume":"76","author":"Cao","year":"2017","journal-title":"Multimed. Tools Appl."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"275","DOI":"10.1007\/s00138-002-0084-z","article-title":"Wood inspection with non-supervised clustering","volume":"13","author":"Niskanen","year":"2003","journal-title":"Mach. Vis. Appl."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"1789","DOI":"10.1007\/s11676-017-0572-7","article-title":"A novel image segmentation approach for wood plate surface defect classification through convex optimization","volume":"29","author":"Chang","year":"2018","journal-title":"J. For. Res."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"1087","DOI":"10.1007\/s00170-011-3352-0","article-title":"Automated visual classification of frequent defects in flat steel coils","volume":"57","author":"Medina","year":"2011","journal-title":"Int. J. Adv. Manuf. Technol."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"1956","DOI":"10.2355\/isijinternational.ISIJINT-2015-201","article-title":"Invariant Feature Extraction Method Based on Smoothed Local Binary Pattern for Strip Steel Surface Defect","volume":"55","author":"Chu","year":"2015","journal-title":"Trans. Iron Steel Inst. Jpn."},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"6469","DOI":"10.1016\/j.ijleo.2013.05.004","article-title":"Fabric defect detection based on GLCM and Gabor filter: A comparison","volume":"124","author":"Raheja","year":"2013","journal-title":"Optik"},{"key":"ref_10","unstructured":"Ding, S., Liu, Z., and Li, C. (2011, January 26\u201328). AdaBoost learning for fabric defect detection based on HOG and SVM. Proceedings of the International Conference on Multimedia Technology, Hangzhou, China."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"965","DOI":"10.1007\/s12541-015-0125-y","article-title":"Fast defect detection for various types of surfaces using random forest with VOV features","volume":"16","author":"Kwon","year":"2015","journal-title":"Int. J. Precis. Eng. Manuf."},{"key":"ref_12","doi-asserted-by":"crossref","unstructured":"Sarkar, A., and Padmavathi, S. (2018). Image Pyramid for Automatic Segmentation of Fabric Defects, Springer.","DOI":"10.1007\/978-3-319-71767-8_49"},{"key":"ref_13","unstructured":"Weimer, D., Benggolo, A.Y., and Freitag, M. (December, January 30). Context-aware Deep Convolutional Neural Networks for Industrial Inspection. Proceedings of the Australasian Conference on Artificial Intelligence, Canberra, Australia."},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"417","DOI":"10.1016\/j.cirp.2016.04.072","article-title":"Design of deep convolutional neural network architectures for automated feature extraction in industrial inspection","volume":"65","author":"Weimer","year":"2016","journal-title":"CIRP Ann."},{"key":"ref_15","doi-asserted-by":"crossref","unstructured":"Kim, S., Kim, W., Noh, Y.K., and Park, F.C. (2017, January 14\u201319). Transfer learning for automated optical inspection. Proceedings of the International Joint Conference on Neural Networks, Anchorage, AK, USA.","DOI":"10.1109\/IJCNN.2017.7966162"},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Bian, X., Lim, S.N., and Zhou, N. (2016, January 7\u201310). Multiscale fully convolutional network with application to industrial inspection. Proceedings of the 2016 IEEE Winter Conference on Applications of Computer Vision (WACV), Lake Placid, NY, USA.","DOI":"10.1109\/WACV.2016.7477595"},{"key":"ref_17","doi-asserted-by":"crossref","unstructured":"Deng, J., Dong, W., Socher, R., Li, L.J., Li, K., and Fei-Fei, L. (2009, January 20\u201325). Imagenet: A large-scale hierarchical image database. Proceedings of the CVPR 2009, IEEE Conference on Ieee Computer Vision and Pattern Recognition, Miami, FL, USA.","DOI":"10.1109\/CVPR.2009.5206848"},{"key":"ref_18","unstructured":"Abadi, M., Barham, P., Chen, J., Chen, Z., Davis, A., Dean, J., Devin, M., Ghemawat, S., Irving, G., and Isard, M. (2016, January 2\u20134). Tensorflow: A System for Large-Scale Machine Learning. Proceedings of the 12th USENIX Symposium on Operating Systems Design and Implementation (OSDI \u201916), Savannah, GA, USA."},{"key":"ref_19","doi-asserted-by":"crossref","unstructured":"Szegedy, C., Ioffe, S., Vanhoucke, V., and Alemi, A. (2016, January 12\u201317). Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning. Proceedings of the Thirty-first AAAI Conference on Artificial Intelligence, Phoenix, AZ, USA.","DOI":"10.1609\/aaai.v31i1.11231"},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"He, K., Zhang, X., Ren, S., and Sun, J. (2016, January 8\u201316). Identity mappings in deep residual networks. Proceedings of the European Conference on Computer Vision, Amsterdam, The Netherlands.","DOI":"10.1007\/978-3-319-46493-0_38"},{"key":"ref_21","unstructured":"Iandola, F.N., Han, S., Moskewicz, M.W., Ashraf, K., Dally, W.J., and Keutzer, K. (2016). Squeezenet: Alexnet-level accuracy with 50x fewer parameters and <0.5 mb model size. arXiv."},{"key":"ref_22","doi-asserted-by":"crossref","unstructured":"Wu, C., Wen, W., Afzal, T., Zhang, Y., Chen, Y., and Li, H. (2017, January 21\u201326). A Compact DNN: Approaching GoogLeNet-Level Accuracy of Classification and Domain Adaptation. Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA.","DOI":"10.1109\/CVPR.2017.88"},{"key":"ref_23","unstructured":"Howard, A.G., Zhu, M., Chen, B., Kalenichenko, D., Wang, W., Weyand, T., Andreetto, M., and Adam, H. (2017). Mobilenets: Efficient convolutional neural networks for mobile vision applications. arXiv."},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"Zhang, X., Zhou, X., Lin, M., and Sun, J. (2018, January 18\u201322). ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA.","DOI":"10.1109\/CVPR.2018.00716"},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., and Chen, L.C. (2018, January 18\u201322). MobileNetV2: Inverted Residuals and Linear Bottlenecks. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA.","DOI":"10.1109\/CVPR.2018.00474"},{"key":"ref_26","doi-asserted-by":"crossref","unstructured":"Mullick, K., and Namboodiri, A.M. (2017, January 17\u201320). Learning deep and compact models for gesture recognition. Proceedings of the 2017 IEEE International Conference on Image Processing (ICIP), Beijing, China.","DOI":"10.1109\/ICIP.2017.8297033"},{"key":"ref_27","doi-asserted-by":"crossref","unstructured":"Chen, K., Tian, L., Ding, H., Cai, M., Sun, L., Liang, S., and Huo, Q. (2017, January 9\u201312). A Compact CNN-DBLSTM Based Character Model for Online Handwritten Chinese Text Recognition. Proceedings of the Iapr International Conference on Document Analysis and Recognition, Kyoto, Japan.","DOI":"10.1109\/ICDAR.2017.177"},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"59803","DOI":"10.1109\/ACCESS.2018.2873948","article-title":"MicronNet: A Highly Compact Deep Convolutional Neural Network Architecture for Real-Time Embedded Traffic Sign Classification","volume":"6","author":"Wong","year":"2018","journal-title":"IEEE Access"},{"key":"ref_29","doi-asserted-by":"crossref","unstructured":"Racki, D., Tomazevic, D., and Skocaj, D. (2018, January 12\u201315). A compact convolutional neural network for textured surface anomaly detection. Proceedings of the 2018 IEEE Winter Conference on Applications of Computer Vision (WACV), Lake Tahoe, NV, USA.","DOI":"10.1109\/WACV.2018.00150"},{"key":"ref_30","doi-asserted-by":"crossref","unstructured":"Kornblith, S., Shlens, J., and Le, Q.V. (2018). Do Better ImageNet Models Transfer Better?. arXiv.","DOI":"10.1109\/CVPR.2019.00277"},{"key":"ref_31","doi-asserted-by":"crossref","unstructured":"Chen, L.C., Zhu, Y., Papandreou, G., Schroff, F., and Adam, H. (2018). Encoder-decoder with atrous separable convolution for semantic image segmentation. arXiv.","DOI":"10.1007\/978-3-030-01234-2_49"},{"key":"ref_32","doi-asserted-by":"crossref","unstructured":"Sun, C., Shrivastava, A., Singh, S., and Gupta, A. (2017, January 22\u201329). Revisiting Unreasonable Effectiveness of Data in Deep Learning Era. Proceedings of the IEEE International Conference on Computer Vision, Venice, Italy.","DOI":"10.1109\/ICCV.2017.97"},{"key":"ref_33","doi-asserted-by":"crossref","unstructured":"Chollet, F. (2017, January 21\u201326). Xception: Deep learning with depthwise separable convolutions. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, HI, USA.","DOI":"10.1109\/CVPR.2017.195"},{"key":"ref_34","unstructured":"Han, S., Mao, H., and Dally, W.J. (2015). Deep compression: Compressing deep neural networks with pruning, trained quantization and huffman coding. arXiv."},{"key":"ref_35","unstructured":"Molchanov, P., Tyree, S., Karras, T., Aila, T., and Kautz, J. (2016). Pruning convolutional neural networks for resource efficient inference. arXiv."},{"key":"ref_36","unstructured":"Gupta, S., Agrawal, A., Gopalakrishnan, K., and Narayanan, P. (2015, January 6\u201311). Deep learning with limited numerical precision. Proceedings of the International Conference on Machine Learning, Lille, France."},{"key":"ref_37","doi-asserted-by":"crossref","unstructured":"Xu, T.B., Yang, P., Zhang, X.Y., and Liu, C.L. (2017, January 13\u201315). Margin-aware binarized weight networks for image classification. Proceedings of the International Conference on Image and Graphics, Shanghai, China.","DOI":"10.1007\/978-3-319-71607-7_52"},{"key":"ref_38","unstructured":"Tai, C., Xiao, T., Zhang, Y., and Wang, X. (2015). Convolutional neural networks with low-rank regularization. arXiv."},{"key":"ref_39","unstructured":"Hinton, G., Vinyals, O., and Dean, J. (2015). Distilling the knowledge in a neural network. arXiv."},{"key":"ref_40","doi-asserted-by":"crossref","first-page":"126","DOI":"10.1109\/MSP.2017.2765695","article-title":"Model compression and acceleration for deep neural networks: The principles, progress, and challenges","volume":"35","author":"Cheng","year":"2018","journal-title":"IEEE Signal Process. Mag."},{"key":"ref_41","unstructured":"Krizhevsky, A., Sutskever, I., and Hinton, G.E. (2012, January 12\u201315). ImageNet classification with deep convolutional neural networks. Proceedings of the International Conference on Neural Information Processing Systems, Sydney, Australia."},{"key":"ref_42","unstructured":"Simonyan, K., and Zisserman, A. (2014). Very Deep Convolutional Networks for Large-Scale Image Recognition. arXiv."},{"key":"ref_43","doi-asserted-by":"crossref","unstructured":"He, K., Zhang, X., Ren, S., and Sun, J. (2016, January 27\u201330). Deep residual learning for image recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA.","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref_44","doi-asserted-by":"crossref","unstructured":"Peng, C., Zhang, X., Yu, G., Luo, G., and Sun, J. (2017, January 21\u201326). Large kernel matters\u2014improve semantic segmentation by global convolutional network. Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA.","DOI":"10.1109\/CVPR.2017.189"},{"key":"ref_45","doi-asserted-by":"crossref","first-page":"834","DOI":"10.1109\/TPAMI.2017.2699184","article-title":"Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected crfs","volume":"40","author":"Chen","year":"2018","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"ref_46","first-page":"1","article-title":"Convolutional deep belief networks on cifar-10","volume":"40","author":"Krizhevsky","year":"2010","journal-title":"Unpubl. Manuscr."},{"key":"ref_47","unstructured":"Springenberg, J.T., Dosovitskiy, A., Brox, T., and Riedmiller, M. (2014). Striving for simplicity: The all convolutional net. arXiv."},{"key":"ref_48","doi-asserted-by":"crossref","unstructured":"Caputo, B., Hayman, E., and Mallikarjuna, P. (2005, January 17\u201321). Class-Specific Material Categorisation. Proceedings of the Tenth IEEE International Conference on Computer Vision, Beijing, China.","DOI":"10.1109\/ICCV.2005.54"},{"key":"ref_49","unstructured":"Kylberg, G. (2011). The Kylberg Texture Dataset v. 1.0. External Report (Blue Series) 35, Centre for Image Analysis, Swedish University of Agricultural Sciences and Uppsala University."},{"key":"ref_50","doi-asserted-by":"crossref","first-page":"1265","DOI":"10.1109\/TPAMI.2005.151","article-title":"A sparse texture representation using local affine regions","volume":"27","author":"Lazebnik","year":"2005","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"ref_51","unstructured":"DAGM (2020, March 26). Weakly Supervised Learning for Industrial Optical Inspection. DAGM Dataset. Available online: https:\/\/hci.iwr.uni-heidelberg.de\/node\/3616."},{"key":"ref_52","unstructured":"Jung, A.B., Wada, K., Crall, J., Tanaka, S., Graving, J., Yadav, S., Banerjee, J., Vecsei, G., Kraft, A., and Borovec, J. (2019, September 25). Imgaug. Available online: https:\/\/github.com\/aleju\/imgaug."},{"key":"ref_53","doi-asserted-by":"crossref","unstructured":"Nilsback, M.E., and Zisserman, A. (2008, January 16\u201319). Automated Flower Classification over a Large Number of Classes. Proceedings of the 2018 Sixth Indian Conference on Computer Vision, Graphics & Image Processing, Bhubaneswar, India.","DOI":"10.1109\/ICVGIP.2008.47"},{"key":"ref_54","first-page":"1","article-title":"A Generic Deep-Learning-Based Approach for Automated Surface Inspection","volume":"5","author":"Ren","year":"2017","journal-title":"IEEE Trans. Cybern."},{"key":"ref_55","unstructured":"Donahue, J., Jia, Y., Vinyals, O., Hoffman, J., Zhang, N., Tzeng, E., and Darrell, T. (2014, January 21\u201326). DeCAF: A deep convolutional activation feature for generic visual recognition. Proceedings of the 31st International Conference on Machine Learning, Beijing, China."},{"key":"ref_56","doi-asserted-by":"crossref","first-page":"1069","DOI":"10.1016\/j.procir.2018.03.264","article-title":"A New Ensemble Approach based on Deep Convolutional Neural Networks for Steel Surface Defect classification","volume":"72","author":"Chen","year":"2018","journal-title":"Procedia CIRP"},{"key":"ref_57","doi-asserted-by":"crossref","unstructured":"Zagoruyko, S., and Komodakis, N. (2016). Wide residual networks. arXiv.","DOI":"10.5244\/C.30.87"},{"key":"ref_58","doi-asserted-by":"crossref","unstructured":"Yu, Z., Wu, X., and Gu, X. (2017, January 10\u201313). Fully convolutional networks for surface defect inspection in industrial environment. Proceedings of the International Conference on Computer Vision Systems, Shenzhen, China.","DOI":"10.1007\/978-3-319-68345-4_37"}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/20\/7\/1974\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T09:14:20Z","timestamp":1760174060000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/20\/7\/1974"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2020,4,1]]},"references-count":58,"journal-issue":{"issue":"7","published-online":{"date-parts":[[2020,4]]}},"alternative-id":["s20071974"],"URL":"https:\/\/doi.org\/10.3390\/s20071974","relation":{},"ISSN":["1424-8220"],"issn-type":[{"value":"1424-8220","type":"electronic"}],"subject":[],"published":{"date-parts":[[2020,4,1]]}}}