本論文旨在自動化工件辨識與檢測技術應用於多工作台(Pallet Pool System，PPS)自動化生產機台，以改善人為疏失對於機台生產線的干擾，並減少其人力需求。研究中針對多工作台自動化生產機台為情境，進行設計並提出基於電腦視覺的兩大機制：(1)在加工前確認此工件是否為加工目標，(2)在加工後確認加工的正確性，並另外規劃與設計CNC機台與電腦自動化連線流程。
　　研究中主要以工件型號辨識與加工程序檢測為主軸，首先使用者擷取加工前與加工後的標準件影像，做為加工前模板影像與加工後模板影像並儲存至模板影像資料庫。工件型號辨識中，在加工前擷取實際影像，與加工前模板影像進行特徵點匹配，並評估匹配結果。加工程序檢測中，在加工後採用參考件比較法，比較加工後模板影像與加工後工件影像之間的差異，檢測出工件的瑕疵部位。此外研究中亦建置系統人機介面，提供使用者了解辨識與檢測狀況。
　　實驗中，採用數位相機擷取工件影像，在離線狀態下對工件型號辨識與加工程序檢測進行時間與準確度之測試，並另外模擬斷刃殘留、鑽孔加工失敗等兩種狀況，評估加工程序檢測的效果。在機台與電腦自動化連線部份，則以網路攝影機進行流程測試。
　　本研究主要貢獻，分述以下三點如下：1.導入工件型號辨識，避免上料時錯放工件。2.導入加工程序檢測，提供機台自動檢測加工正確性，以減少工件檢測時的人力需求。3.規劃與設計機台與電腦自動化連線，賦予機台自動化辨識與檢測之功能。

PPS (Pallet Pool System) automatic production line is a highly automatic loading system which integrates CNC machine tool and storage areas. To prevent the PPS automatic production line from operator errors, a vision-based identification and inspection system for PPS automatic production line is presented in this research. Two main functions are developed: 1. A feature-based identification method is proposed to identify whether the workpiece matches machining mission before machining. 2. A Reference Matching inspection method is used to compare template image and workpiece image to inspect the defective area on the workpiece after machining. In addition, the design and the planning of automatic connection between a machine tool and a computer are proposed. In the experiment, Webcam was adopted to test information flow of automatic connection. A digital camera was adopted to test the system in an off-line experiment. In this research, three main advantages of this developed system are obtained: 1. Implement workpiece identification to avoid setting a wrong workpiece while loading. 2. Implement workpiece inspection to automatically inspect the machining error and reduce the requirement of manpower. 3. Machine is provided with the function of automatic identification and inspection by the design of automatic connection between a machine tool and a computer.

[1] 行政院科技會報,”我國2020智慧型自動化產業發展策略,” http://www.bost.ey.gov.tw/Upload/UserFiles/ /議題1_我國2020智慧型自動化產業發展策略.pdf , March 13,2016.
[2] Browne, Jim, et al. "Classification of flexible manufacturing systems.", The FMS magazine , vol. 22, pp. 114-117., 1984.
[3] H. Golnabi and A. Asadpour, "Design and application of industrial machine vision systems", Robotics and Computer-Integrated Manufacturing , vol. 23, no. 6, pp. 630-637, 2007.
[4] DMG MORI COMPANY LIMITED,”Automation-Reference-Booklet,” http://us.dmgmori.com/blob/262502/0f75eb9826e29ba01451de7ac23e6b32/ps0uk14-automation-reference-booklet-pdf-data.pdf, March 13,2016.
[5] Tongtai Machine&Tool Co.,Ltd,”臥式加工中心機 HA-400II/500II,” http://www.tongtai.com.tw/cht/product/showinfo-64.html, March 07,2016.
[6] Okuma America|CNC Machine Tools|CNC Controls,” Pallet Pools ,” http://www.okuma.com/pallet-pools-2, March 07,2016.
[7] 邱國育　洪裕筆　蔡昌裕,”智慧自動化－視覺整合在雷射切割機台的應用” ,機械工業雜誌104年2月號，P100.
[8] 黃國龍, "工件尺寸之自動影像檢測系統之研究",　碩士,國立雲林科技大學機械工程系研究所, 2013.
[9] P. Rummel and W. Beutel, "Workpiece recognition and inspection by a model-based scene analysis system", Pattern Recognition , vol. 17, no. 1, pp. 141-148, 1984.
[10] W. Yan, F. Weiping, Z. Hong and L. Yuanyue, "A Method for Recognition of Work-Pieces Based on Improved SIFT Characteristics Matching," Software Engineering, 2009. WCSE '09. WRI World Congress on, Xiamen, 2009, pp. 29-33.
[11] X. Wang, W. Fu and D. Zhu, "Research on recognition of work-piece based on improved SIFT algorithm," Electronic Measurement & Instruments, 2009. ICEMI '09. 9th International Conference on, Beijing, 2009, pp. 1-417-1-421.
[12] D. Lowe, "Distinctive Image Features from Scale-Invariant Keypoints", International Journal of Computer Vision , vol. 60, no. 2, pp. 91-110, 2004.
[13] L. Juan and O. Gwun, "A comparison of sift, pca-sift and surf.", International Journal of Image Processing , vol. 34, pp. 143-152, 2009.
[14] 劉建國, "光學影像檢測模組之開發與應用", 碩士, 國立台灣大學機械工程系研究所, 2002.
[15] 范光照, et al. “環形粉末冶金件之自動光學檢測,”粉末冶金會刊., pp.1-7,2008.
[16] T. Alexandropoulos, S. Boutas, V. Loumos and E. Kayafas, "Template-guided inspection of arbitrarily oriented targets", IET Comput. Vis., vol. 2, no. 3, p. 150, 2008.
[17] Ibrahim, Ismail, et al, "A printed circuit board inspection system with defect classification capability", Int. J. Innovative Manage. Inf. Prod 3.1 2012.
[18] C. Harris and M. Stephens, "A Combined Corner and Edge Detector", Procedings of the Alvey Vision Conference 1988, 1988.
[19] 井上誠喜等 合著，吳上立，林明德　譯，”C語言數位影像處理”，全華圖書,2009
[20] Y. Cheng, "Image Registration Based on Mutual Information and Harris Corner Points", AMM , vol. 220-223, pp. 2199-2203, 2012.
[21] Y. Liu, H. Chen, Y. Guo, W. Sun and Y. Zhang, "The Research of Remote Sensing Image Matching Based on the Improved Harris Corner Detection Algorithm", AMR , vol. 271- 273, pp. 201-204, 2011.
[22] N. Otsu, "A Threshold Selection Method from Gray-Level Histograms", IEEE Transactions on Systems, Man, and Cybernetics, vol. 9, no. 1, pp. 62-66, 1979.
[23] Ardunio,” Arduino Pro Mini,”
https://www.arduino.cc/en/Main/ArduinoBoardProMini, June 25,2016.