工業界中常使用的金屬材料常需經電鍍處理以強化其抗腐蝕的能力，但電鍍處理常導致金屬材料表面的反光性變強，若表面上出現刮痕、撞痕等瑕疵，則需仰賴人力從不同角度去觀看金屬片，以找出瑕疵，此將耗費大量時間。為克服此問題，本研究著眼於使用電腦軟體搭配機器視覺來自動化辨識高反光性金屬表面上的瑕疵。
本研究主要分成三大部分：瑕疵之分類方式、瑕疵之打光方式及瑕疵之軟體檢測，其中“瑕疵之分類方式”將金屬片上常出現的瑕疵區分為凹陷、鍍膜不均及刮痕三大類；“瑕疵之打光方式”則是以實驗的角度測試出何種瑕疵需使用何種檢測光源及何種角度照射，方可獲得最清晰的瑕疵輪廓影像；“瑕疵之軟體檢測”則是開發電腦檢測軟體，由影像中辨識出瑕疵，並將瑕疵清楚的標示在影像中。
本論文所提出之打光及瑕疵檢測方法對於凹陷可達到85%檢測成功率、刮痕可達到86%，鍍膜不均的部分，若屬大範圍的鍍膜不均，則可達到100%檢測成功率。本研究亦以此瑕疵檢測方法對不同尺寸的金屬片進行瑕疵檢測，驗證其泛用性。

Metal materials in today’s industry must be electroplated first before used, so that its corrosion resistance could be highly improved. But this process will make the metal surface very reflective. If some tiny dents or scratches appear on the surface, lots of technicians are needed to rotate and flip the metal to find the defects, which consumes a large amount of manpower and time. In order to solve this problem, this thesis attempts to use computer programs in together with machine vision to automatically detect defects on metal surface coatings.
This study consists of the following three research issues: defining the criteria to classify various defects, proposing the configuration of optical lighting, and developing computer codes for defect inspection. The first issue “defining the criteria to classify various defects” studies all kinds of defects that appear on the metal samples first, and then classifies them into dent, uneven coating and scratch. The second issue “proposing the configuration of optical lighting” experiments different optical arrangements to find out a good way to obtain clear images for each type of defect. The third issue “developing computer codes for defect inspection” applies computer programs to automatically distinguish defect types and clearly mark the types on the images.
Optical lighting along with image processing methods proposed in this research can achieve 85% detection success rate for dent, 86% for scratch, and 100% for uneven coating if the uneven area is huge. Furthermore, this thesis uses the same method to detect defects in different sizes of metal samples in order to verify its multiusability.

[1] https://opencv.org/
[2] Dufour, M.L. and Samson, M. (1992), “On-line detection of surface defects on steel billets using multiple grazing-incidence light sources,” Optics, Illumination, and Image Sensing for Machine Vision, Vol. 1614, pp. 35-44.
[3] Fu, S., Cheng, F. and Tjahjowidodo, T. (2020), “Surface Topography Measurement of Mirror-Finished Surfaces Using Fringe-Patterned Illumination,” Metals, Vol. 10, Issue. 1, pp. 1-11.
[4] Xu, L.M., Yang, Z.Q. and Jiang, Z.H. (2017), “Light source optimization for automatic visual inspection of piston surface defects,” Internation Journal of Advanced Manufactureing Technology, Vol. 91, pp. 2245-2256.
[5] 莊宜橙，「玻璃基板表面刮痕檢測方法之研究」（2010），碩士論文，國立高雄應用科技大學機械與精密工程研究所，高雄市。
[6] Hsu, Q.C., Ngo, N.V. and Ni, R.H. (2019), “Development of a faster classification system for metal parts using machine vision under different lighting environments,” The International Journal of Advanced Manufacturing Technology, Vol. 100, Issue 9-12, pp. 3219-3235.
[7] Newman, T.S. and Jain, A.K. (1995), “A Survey of Automated Visual Inspection,” Computer Vision and Image Understanding, Vol. 61, Issue 2, pp. 231-262.
[8] Akdemir, B. and Ozturk, S. (2017), “Detection of PCB Soldering Defects using Template Based Image Processing Method,” International Journal of Intelligent Systems and Applications in Engineering, Vol. 5, No. 4, pp. 269-273.
[9] Tsai, D.M. and Huang, C.K. (2019), “Defect Detection in Electronic Surfaces Using Template-Based Fourier Image Reconstruction,” IEEE Transactions on Components, Packaging and Manufacturing Technology, Vol. 9, No. 1, pp. 163-172.
[10] Yun, J.P., Kim, D., Kim, K., Lee, S.J., Park, C.H. and Kim, S.W. (2017), “Vision-based surface defect inspection for thick steel plates,” Optical Engineering, Vol. 56, Issue 5, pp. 1-12.
[11] Otsu, N. (1979), “A Threshold Selection Method from Gray-Level Histograms,” IEEE Transactions on Systems, Man, and Cybernetics, Vol. 9, No. 1, pp. 62-66.
[12] Liu, M., Zhuang, R., Guo, X. and Zhao, J. (2020), “Application of improved Otsu threshold segmentation algorithm in mobile phone screen defect detection,” Chinese Control And Decision Conference, August 22-24, 2020, Hefei, China.
[13] Canny, J. (1986), “A Computational Approach to Edge Detection,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. PAMI-8, No. 6, pp. 679-698.
[14] https://www.google.com.tw/search?q=%E5%90%8C%E8%BB%B8%E5%85%89&hl=zh-TW&sxsrf=ALeKk01JreEME-pDwLu35QeacwCQcaZZ6A:1619599755742&source=lnms&tbm=isch&sa=X&ved=2ahUKEwix1vmgx6DwAhVyNKYKHT0vBhQQ_AUoAXoECAEQAw&biw=1920&bih=937#imgrc=xtKbc0fvE20IIM
[15] https://login.taobao.com/member/login.jhtml?redirectURL=https%3A%2F%2Fworld.taobao.com%2Fitem%2F615680666197.htm%3Fspm%3Da21wu.10013406-tw.0.0.187f41f1ZFOhUR&uuid=913d01f44b5b14db3811088295c201cb
[16] 陳宗達，「CMOS玻璃蓋片自動光學檢測機台之設計及開發」（2005），碩士論文，國立交通大學工業工程與管理系所，新竹市。
[17] https://www.china.cn/shulihuajiaoxueqc/4378643928.html
[18] 李立宗，科班出身的AI人必修課：OpenCV影像處理，深智數位出版，台北市，2019。
[19] https://www.tjc.com.tw/products-detail/id/7/title/FL_%E6%B0%B4%E5%B9%B3%E7%92%B0%E5%9E%8B%E7%87%88(90%C2%B0)
[20] https://www.baslerweb.com/en/products/software/basler-pylon-camera-software-suite/pylon-viewer/
[21] https://blog.jetbrains.com/pycharm/2020/05/pycharm-2020-1-1/
[22] https://www.python.org/