Automatic Metallic Surface Defect Detection using ShuffleDefectNet

Avlokulov Anvar; Young Im Cho (조영임)

doi:10.9708/jksci.2020.25.03.019

Automatic Metallic Surface Defect Detection using ShuffleDefectNet

Journal of The Korea Society of Computer and Information
Abbr : JKSCI
2020, 25(3), pp.19-26
DOI : 10.9708/jksci.2020.25.03.019
Publisher : The Korean Society Of Computer And Information
Research Area : Engineering > Computer Science
Received : December 31, 2019
Accepted : February 24, 2020
Published : March 31, 2020

Avlokulov Anvar ¹, Young Im Cho ¹

¹가천대학교

Accredited

ABSTRACT

Steel production requires high-quality surfaces with minimal defects. Therefore, the detection algorithms for the surface defects of steel strip should have good generalization performance. To meet the growing demand for high-quality products, the use of intelligent visual inspection systems is becoming essential in production lines. In this paper, we proposed a ShuffleDefectNet defect detection system based on deep learning. The proposed defect detection system exceeds state-of-the-art performance for defect detection on the Northeastern University (NEU) dataset obtaining a mean average accuracy of 99.75%. We train the best performing detection with different amounts of training data and observe the performance of detection. We notice that accuracy and speed improve significantly when use the overall architecture of ShuffleDefectNet.

KEYWORDS

Defect detection, Deep Learning, ShuffleNet, Light-weight modules

Citation status

* References for papers published after 2023 are currently being built.

[journal] R. Rajkolhe / Defects, causes and their remediesin casting process: A review 2(3) : 375~383

[journal] X. Li / Improving automatic detection of defects in castings by applying wavelet technique 53(6) : 1927~1934

[journal] S. Ghorai / Automatic defect detection on hot-rolled flat steel products 62(3) : 612~621

[journal] LeCun, Y. / 2015 / Deep learning / nature 521(7553) : 436~

[other] Ningning Ma / 2018 / ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design

[other] Sandler, M. / 2018 / L.C.:Inverted residuals and linear bottlenecks: Mobile networks for classification, detection, and segmentation / arXiv preprint arXiv:1801.04381

[other] Zhang, X. / 2017 / Shufflenet: An extremely efficient convolutional neural network for mobile devices / arXiv preprint arXiv:1707.01083

[other] Chollet, F. / 2016 / Xception: Deep learning with depthwise separable convolutions / arXiv preprint

[confproc] Ciresan, D. / 2012 / Multi-column deep neural networks for im-age classification / Proc. of Conference on Computer Vision and Pattern Recognition (CVPR) : 3642~3649

[journal] Arel, I. / 2010 / Deep machine learning - a new frontier inartificial intelligence research / IEEE Computational Intelligence Magazine 5(4) : 13~18

[confproc] Simard, P.Y. / 2003 / Best practices for convolutional neural networks applied to visual document analysis / Proc. of International Conference on Document Analysis and Recognition (ICDAR) : 958~963

[confproc] Coates, A. / 2011 / An analysis of single-layer networks in unsupervised feature learning / Proc. of International Conference on Artificial intelligence and statistics (AISTATS)

[confproc] Soukup, D. / 2014 / Convolutional neural networks for steel surface defect detection from photometric stereo images / Advances in Visual Computing 8887 : 668~677

[journal] Ke, W. / 2016 / Banknote Image Defect Recognition Method Based on Convolution Neural Network / International Journal of Security and Its Applications 10(6) : 269~280

[confproc] Faghih-Roohi, S. / 2016 / Deep convolutional neural networks for detection of rail surface defects / 2016 International Joint Conference on Neural Networks (IJCNN) : 2584~2589

[journal] 박제강 / 2016 / Machine Learning-Based Imaging System for Surface Defect Inspection / International Journal of Precision Engineering and Manufacturing-Green Technology / 한국정밀공학회 3(3) : 303~310

[journal] Weimer, D. / 2016 / Design of deep convolutional neural network architectures for automated feature extraction in industrial inspection / CIRP Annals - Manufacturing Technology 65(1) : 417~420

[journal] K. Song / 2013 / A noise-robust method based on completed local binary patterns for hot-rolled steel strip surface defects / Appl. Surface Sci 285 : 858~864

[other] Sandler, M. / 2018 / Inverted residuals and linear bottlenecks: Mobile networks for classification, detection, and segmentation / arXiv preprint arXiv:1801.04381

[confproc] A. Krizhevsky / 2012 / Imagenet classification with deep convolutional neural networks / Advances in neural information processing systems : 1097~1105

[confproc] S. Xie / 2017 / Aggregated residual transformations for deep neural networks / Computer Vision and Pattern Recognition (CVPR), 2017 IEEE Conference on / IEEE : 5987~5995

[journal] Yu He / 2020 / An End-to-End Steel Surface Defect Detection Approach via Fusing Multiple Hierarchical Features / IEEE Transactions on Instrumentation and Measurement / Institute of Electrical and Electronics Engineers (IEEE) 69(4) : 1493~1504

[journal] K. Song / 2013 / A noise-robust method based on completed local binary patterns for hot-rolled steel strip surface defects / Appl. Surface Sci 285 : 858~864

[journal] M. Xiao / 2017 / An evolutionary classifier for steel surface defects with a small sample set / EURASIP J. Image Vid. Process 2017(1) : 48~61

[confproc] P. Chen / 2016 / Is over feat useful for image-based surface defect classification tasks? / Proc. IEEE Int. Conf. Image Process. (ICIP) : 749~753

[journal] R. Ren / 2018 / A generic deep-learning-based approach for automated surface inspection / IEEE Trans. Cybern 48(3) : 929~940

[journal] Y. Lecun / 2015 / Deep learning / Nature 521(7553) : 436~444

This paper was written with support from the National Research Foundation of Korea.

KJCKorea
Journal Central

Journal of The Korea Society of Computer and Information 2023 KCI Impact Factor : 0.65

Automatic Metallic Surface Defect Detection using ShuffleDefectNet

ABSTRACT

KEYWORDS

Citation status

* References for papers published after 2023 are currently being built.

Journal of The Korea Society of Computer and Information 2023 KCI Impact Factor : 0.65

Automatic Metallic Surface Defect Detection using ShuffleDefectNet

ABSTRACT

KEYWORDS

Statistics

Tools

Issue List

Citation status

KCI Citation Counts (1)

REFERENCES (27) * References for papers published after 2023 are currently being built.

Search PDF

Citation

* References for papers published after 2023 are currently being built.