應用YOLOV5於表面瑕疵檢測

Applying YOLOV5 in Surface Defect Detection

劉軒瑋(H. W. Liu) ； 洪鈺欣(Y. H. Hung) ； 林師弘(S. H. Lin)

《危機管理學刊》 21卷1期 (2024/03) Pp. 27-38

https://doi.org/10.6459/JCM.202403_21(1).0004

摘要

近年來物件偵測模型已被廣泛應用於工業領域中，相關企業使用AOI針對螺絲影像進行瑕疵檢測，但因拍攝影像中含有不完整或多個螺絲影像，導致AOI檢測誤判及無法分類瑕疵種類的問題，為了解決上述問題，本研究提出深度學習框架偵測螺絲瑕疵，首先進行影像前處理，並提取出完整的螺絲影像，最終建立YOLOv5模型進行螺絲帽頭瑕疵檢測。

關鍵字

深度學習 ； 自動化工業檢測 ； 瑕疵檢測 ； YOLOv5

並列摘要

Object detection models have been widely used in the industrial field, and related companies use AOIs for screw image defect detection. However, the captured image contains incomplete or multiple screw images, which results in AOI detection false kill, and the type of defects cannot be classified. This study proposes a deep learning framework for detecting screw defects by first performing image preprocessing and extracting complete screw images and, finally, building the YOLOv5 model for screw head defect detection.

並列關鍵字

Deep Learning ； Automated Industrial Inspection ； Defect Detection ； YOLOv5

參考文獻

紀翔瀛 (2022). “產業技術評析- 2022 年扣件市場與技術發展,” 2022 年 1 月 5 日, 取自 https://reurl.cc/Rv9Yj6

Lin, D., Li, Y., Prasad, S., Nwe, T. L., Dong, S., & Oo, Z. M. (2021). “CAM-guided Multi-Path Decoding U-Net with Triplet Feature Regularization for Defect Detection and Segmentation,” Knowledge-Based Systems, 228, 107272.

Lu, B., Xu, D., & Huang, B. (2022). “Deep-learning-based anomaly detection for lace defect inspection employing videos in production line,” Advanced Engineering Informatics, 51, 101471.

Kong, Q., Wu, Z., & Song, Y. (2022). “Online detection of external thread surface defects based on an improved template matching algorithm,” Measurement 195, 111087.

Avola, D., Cascio, M., Cinque, L., Fagioli, A., Foresti, G. L., Marini, M. R., & Rossi, F. (2022). “Real-time deep learning method for automated detection and localization of structural defects in manufactured products,” Computers & Industrial Engineering, 172, 108512.