基於深度卷積神經網路於脫蠟鑄件噴砂缺陷檢測之研究

The Study of Sandblasting Defect Inspection on Investment Castings Based on Deep Convolutional Neural Network

伍俊嘉(Jun-Jia Wu) ； 黃培興(Pei-Hsing Huang)

《鑄造工程學刊》 48卷1期 (2022/06) Pp. 25-36

摘要

由於脫蠟鑄件噴砂後，表面常殘留毛面或殼模等雜質及缺陷，需再經由人工耗時的二次檢測，本研究使用自動化光學檢測搭配深度學習進行脫蠟鑄件噴砂後之零件缺陷檢測，目的在減少工作人員成本、並避免長時間處於高噪音及空氣品質較差之工作環境。本研究深度學習以卷積神經網路（convolutional neural network, CNN）為主要架構，針對資料集使用AlexNet、VGG-16、GoogLeNet及ResNet-34等四種經典卷積神經網路模型進行訓練及預測分類有無缺陷，並將四種模型的預測結果進行綜合比較。研究結果顯示，在預測分類問題中，除了ResNet-34之外，AlexNet、VGG-16、GoogLeNet v1在辨識上均可準確地分類出有缺陷與無缺陷。其中，AlexNet為本研究辨識脫蠟鑄件是否有缺陷的最佳模型，其良品的預測準確率為99.53%、不良品的預測準確率為100.00%，檢測速度約1毫秒左右。另外，本論文設計一圖形使用者介面，此介面結合機器視覺、卷積神經網路及物件追蹤等技術，有助於使用者操作及監控脫蠟鑄件經噴砂工法後，表面是否殘留毛面或殼模等雜質的存在。

關鍵字

自動光學檢測 ； 脫蠟鑄件 ； 噴砂缺陷檢測 ； 卷積神經網路 ； 深度學習

並列摘要

In this study, the automated optical inspection (AOI) in coordination with deep learning was employed to detect the defects such as burrs or residues of ceramic shell mold on the surface of in sandblasted investment casting parts. The objective is to reduce the errors caused by long hours of manual inspection and to prevent personnel from the work environments with loud noise and poor air quality lasted long periods of time. In this study the deep learning framework was mainly based on convolutional neural networks (CNNs). Four classic CNN models, namely, AlexNet, VGG-16, GoogLeNet, and ResNet-34, were applied to the datasets for training, predicting, and classifying whether there are defects and their results were compared. The results revealed that in terms of the classification prediction, AlexNet, VGG-16, and GoogLeNet v1 could accurately determine the defects, whereas ResNet-34 could not. AlexNet was the most accurate in detecting detects on the investment casting parts in this study; it presented a prediction accuracy of 99.53% for good products and 100.00% for the defective ones. A graphical user interface (GUI) including machine vision (MV), CNNs, and object tracking was also designed and can assist users for detecting defects such as burrs surfaces or residues of shell mold contained on the surfaces of sandblasted investment casting parts.

並列關鍵字

Automated optical inspection ； investment castings ； sandblasting defect detection ； convolutional neural network ； deep learning

參考文獻

A. Korodi, D. Anitei, A. Boitor, and I. Silea, "Image-Processing-Based Low-Cost Fault Detection Solution for End-of-Line ECUs in Automotive Manufacturing," Sensors, vol. 20, no. 12, p. 3520, 2020.

Q. Luo et al., "Automated Visual Defect Classification for Flat Steel Surface: A Survey," IEEE Transactions on Instrumentation and Measurement, vol. 69, no. 12, pp. 9329-9349, 2020.

M. Alencastre-Miranda, R. M. Johnson, and H. I. Krebs, "Convolutional Neural Networks and Transfer Learning for Quality Inspection of Different Sugarcane Varieties," IEEE Transactions on Industrial Informatics, vol. 17, no. 2, pp. 787-794, 2020.

S. Dorafshan and H. Azari, "Deep learning models for bridge deck evaluation using impact echo," Construction and Building Materials, vol. 263, p. 120109, 2020.

C.-F. J. Kuo, T.-y. Fang, C.-L. Lee, and H.-C. Wu, "Automated optical inspection system for surface mount device light emitting diodes," Journal of Intelligent Manufacturing, vol. 30, no. 2, pp. 641-655, 2019.