Journals / Gazi Üniversitesi Mühendislik Mimarlık Fakültesi Dergisi / 2023 / Cilt: 38 - Sayı: 2

InceptionV3 Based Enriched Feature Integration Network Architecture for Pixel-Level Surface Defect Detection

Piksel seviyesinde yüzey hata tespiti için InceptionV3 tabanlı zenginleştirilmiş öznitelik entegrasyon ağ mimarisi

Pages
721–732
Publication Date
2022-10-07
DOI
—
🏛️Bibliographic metadata and full text access for this article are provided via DergiPark.
DergiPark Record ↗
✨

AI Summary & Key Takeaways

AI Insight
⚡ TL;DR(Too Long; Didn't Read)

Keywords: Pixel-Level Surface Defects Detection, Convolutional Neural Network, Squeeze and Excitation Block, Feature Pyramid Networks

📌 3 Key Takeaways

🎯
Objective & Hypothesis:In this study, InceptionV3 based Enriched Feature Integration Network (Inc-EFIN) architecture was developed for automatic detection of surface defects.
📊
Key Findings & Data:In the proposed architecture, features of all levels of the InceptionV3 network architecture with the same height and width are extracted and combined.
💡
Conclusion & Impact:Keywords: Pixel-Level Surface Defects Detection, Convolutional Neural Network, Squeeze and Excitation Block, Feature Pyramid Networks
🔬Design:Research Article
📍Scope:Comprehensive Sample
⚙️Variables:Piksel Seviyesinde yüzey kusurları algılama · Evrişimsel sinir ağı · Sıkma ve Uyarma Bloğu

Abstract

In this study, InceptionV3 based Enriched Feature Integration Network (Inc-EFIN) architecture was developed for automatic detection of surface defects. In the proposed architecture, features of all levels of the InceptionV3 network architecture with the same height and width are extracted and combined. As a result of merging, 5 feature maps with different dimensions were obtained. Channel-Based Squeeze and Excitation (CSE) block has been applied to reveal important details in these feature maps. The CSE block strengthens important details by examining the channels in the feature map. In Feature Pyramid Network (FPN) module, information from low-level feature maps containing spatial details were transferred to high-level feature maps containing semantic details. Then, for the final feature map in the proposed architecture, 4 different feature maps obtained at the end of the FPN module were combined using the Feature Integration and Signification (FIS) module. The feature map combined in the FIS module was passed through the Spatial and Channel-based Squeeze and Excitation (SCSE) block, enhancing the spatial and semantic information that may be important for defect detection in the best way. Defect detection results were obtained by using convolution and sigmoid layers in the last layer of the Inc-EFIN architecture. MT, MVTec-Texture, and DAGM datasets were used to calculate the pixel-level defect detection success of the Inc-EFIN architecture. In experimental studies, Inc-EFIN architecture achieved higher performance than the latest technologies in the literature with 77.44% mIoU, 81.2% mIoU and 79.46% mIoU performance results in MT, MVTec-Texture and DAGM datasets, respectively. Keywords: Pixel-Level Surface Defects Detection, Convolutional Neural Network, Squeeze and Excitation Block, Feature Pyramid Networks

Özet

Bu çalışmada, otomatik yüzey hata tespiti için InceptionV3 tabanlı Zenginleştirilmiş Öznitelik Entegrasyon Ağ (Inc-ZÖEA) mimarisi geliştirilmiştir. Önerilen mimaride, InceptionV3 ağ mimarisinin her seviyesindeki öznitelikleri aynı yükseklik ve genişliğe sahip öznitelikler çıkartılmış ve birleştirilmiştir. Birleştirme sonucunda farklı boyutlara sahip olan 5 öznitelik haritası elde edilmiştir. Bu öznitelik haritalarındaki önemli detayları ortaya çıkartmak için Kanal Bazlı Sıkma ve Uyarlama (KSU) bloğu uygulanmıştır. KSU bloğu, öznitelik haritasındaki kanalları inceleyerek önemli ayrıntıları güçlendirmektedir. Öznitelik Piramit Ağ (ÖPA) modülünde mekânsal detayları içeren düşük seviyeli öznitelik haritalarındaki bilgiler, anlamsal detayları içeren yüksek seviyeli öznitelik haritalarına aktarılmıştır. Daha sonra önerilen mimaride nihai öznitelik haritası için Öznitelik Entegrasyon ve Anlamlandırma (ÖEA) modülü kullanılarak ÖPA modülünün sonunda elde edilen 4 farklı öznitelik haritaları birleştirilmiştir. ÖEA modülünde birleştirilen öznitelik haritası Mekânsal ve Kanal Bazlı Sıkma ve Uyarlama (MKSU) bloğundan geçirilerek hata tespiti için önemli olabilecek mekânsal ve anlamsal bilgiler en iyi şekilde güçlendirilmiştir. Inc-ZÖEA mimarisinin son katmanında evrişim ve sigmoid katmanları kullanılarak hata tespit sonucu elde edilmiştir. Inc-ZÖEA mimarisinin piksel seviyesinde hata tespit başarısını ölçmek için MT, MVTec-Doku ve DAGM veri setleri kullanılmıştır. Deneysel çalışmalarda, MT, MVTec-Doku ve DAGM veri setlerinde sırası ile Inc-ZÖEA mimarisi %77,44 mIoU, %81,2 mIoU ve %79,46 mIoU başarım sonuçları ile literatürde yer alan son teknolojilere göre daha yüksek başarımlar sağlamıştır

🔖

Cite This Article

ÜZEN, H., TÜRKOĞLU, M., ARI, A., HANBAY, D. (2022). InceptionV3 Based Enriched Feature Integration Network Architecture for Pixel-Level Surface Defect Detection. Gazi Üniversitesi Mühendislik Mimarlık Fakültesi Dergisi, Cilt: 38 - Sayı: 2, 721–732. https://ajindex.com/articles/667483

📄 Full Text PDF Viewer

📄

PDF görüntüleyici hazırlanıyor...

References (13)

  1. K. Hanbay, M. F. Talu, and Ö. F. Özgüven, Fabric defect detection systems and methods—A systematic literature review, Optik 127(24), 11960–11973, 2016
  2. H. Dong, K. Song, Y. He, J. Xu, Y. Yan, and Q. Meng, PGA-Net: Pyramid Feature Fusion and Global Context Attention Network for Automated Surface Defect Detection, IEEE Transactions on Industrial Informatics 16(12), 7448–7458, 2020
  3. M. H. Karimi and D. Asemani, Surface defect detection in tiling Industries using digital image processing methods: Analysis and evaluation, ISA Transactions 53(3), 834–844, 2014
  4. S. R. Aghdam, E. Amid, and M. F. Imani, A fast method of steel surface defect detection using decision trees applied to LBP based features, Proceedings of the 2012 7th IEEE Conference on Industrial Electronics and Applications, ICIEA 2012 1447–1452, 2012
  5. J. A. Tsanakas, D. Chrysostomou, P. N. Botsaris, and A. Gasteratos, Fault diagnosis of photovoltaic modules through image processing and Canny edge detection on field thermographic measurements, http://dx.doi.org/10.1080/14786451.2013.826223 34(6), 351–372, 2013
  6. K. L. Mak, P. Peng, and K. F. C. Yiu, Fabric defect detection using morphological filters, Image and Vision Computing 27(10), 1585–1592, 2009
  7. X. Bai, Y. Fang, W. Lin, L. Wang, and B. F. Ju, Saliency-based defect detection in industrial images by using phase spectrum, IEEE Transactions on Industrial Informatics 10(4), 2135–2145, 2014
  8. G. Liu and X. Zheng, Fabric defect detection based on information entropy and frequency domain saliency, The Visual Computer 2020 37:3 37(3), 515–528, 2020
  9. X. Dong, C. J. Taylor, and T. F. Cootes, A Random Forest-Based Automatic Inspection System for Aerospace Welds in X-Ray Images, IEEE Transactions on Automation Science and Engineering, 2020
  10. L. Qiu, X. Wu, and Z. Yu, A High-Efficiency Fully Convolutional Networks for Pixel-Wise Surface Defect Detection, IEEE Access 7, 15884–15893, 2019
  11. J. Cao, G. Yang, and X. Yang, A Pixel-Level Segmentation Convolutional Neural Network Based on Deep Feature Fusion for Surface Defect Detection, IEEE Transactions on Instrumentation and Measurement 70,, 2021
  12. H. Firat, 3B ESA Tabanlı ResNet50 Kullanılarak Hiperspektral Görüntülerin Sınıflandırılması Classification of Hyperspectral Images Using 3D CNN Based ResNet50, 2021 29th Signal Processing and Communications Applications Conference, SIU) 6–9, 2021
  13. H. Uzen, H. Firat, A. Karci, and D. Hanbay, Automatic Thresholding Method Developed w