Please use this identifier to cite or link to this item:
標題: TFT-LCD包裝條碼缺陷視覺檢測技術
TFT-LCD Packing Barcode Defect Inspection by AOI
作者: 王清俊
Ching-Jiun Wang
關鍵字: TFT-LCD
computer vision
barcode defect
引用: [1.] 曾俊舜 (2008) , '條碼快速定位技術之研發' ,國立臺灣海洋大學碩士論文。 [2.] 蘇彥志 (2000) , '條碼定位使用四向梯度運算',大同大學碩士論文. [3.] 陳嵩月 (2009) , '應用形狀特徵於有效濾除文字圖騰之即時一維條碼定位系統',國立台灣科技大學碩士論文. [4.] 鄭秀文 (2004) ,'扭曲條碼辨識系統',大同大學碩士論文. [5.] R. Mufiiz, and I. Junco, and A. Otero,' A Robust Software Barcode Reader using the Hough Transfer', Proceedings of the 1999 International Conference on the Information Intelligence and System, pp.313-319,1999. [6.] Arnould, S., and Awcock, G.J., and Thomas, R., 'Remote barcode localization using mathematical morphology,' Seventh International Conference on Image Processing and Its Application, 2, pp. 642-646(1999). [7.] M. K. Hu ' Visual pattern recognition by moment invariants.' IRE Trans. Information Theory. Vol.8.pp.179-187, 1962. [8.] S. O. Belkasim, and M. Shridhar, and M. Ahmadi,' Pattern recognition with moment invariants: a comparative study and new result.' Pattern Recognition,vol.24,pp.1117-1138,1991. [9.] J. Flusser, and T. Suk, and S. Saic,' Recognition of blurred image by the method of moment,' IEEE Trans. Image Process, vol.5.pp. 533-538,1996. [10.] Jan Flusser, and Tomas Suk,' Affine moment variants: A new tool for character recognition' Pattern Recognition,vol.15, pp.433-436, 1994. [11.] Alberto Tomita, and Jr. Tsuyoshi Ebina, and Rokuya Ishii 'An Image Process Method for Use in a GUI for visually Impaired.' Industrial Electronics, Control and Instrumentation, 1997. IECON 97. 23rd International Conference on, vol.1, pp: 264 -268, 1997. [12.] Gonzalez. Woods(2002), '數位影像處理Digital Image Processing 3/e' (繆紹綱譯), 普林斯頓國際有限公司(2009). [13.] OpenCV 2.4.6 documentation, Retrieved from [14.] OpenCV 2.4.6.documentation>> Camera Calibration and 3D Reconstruction, Retrieved from [15] Gary Bradski ,and Adrian Kaehler, 'O''Reilly Learning OpenCV ',Retrieved from [16] 康多利有限公司(2000),'Barcode條碼 ',取自 [17] R.C Hibbeler. 'ENGINEERING MECHANICS-STATICS 12/E ANADAPTER VERSION', 歐亞書局(2009). [18] Keyence ,' SR-650 小型二維條碼讀取器產品規格 ',取自
摘要: 條碼系統於1974年被發展出來後,發展至今已廣泛應用於工業產品、商業資料與醫療保險資料,於TFT-LCD包裝製程上條碼對於紀錄產品特徵,如生產年份、產品等級、產品代碼,有著極其重要的一角,而縱觀條碼的產生多是由條碼機使用碳帶由印字頭加熱印在標籤紙上,如果當印字頭上累積過多的粉塵或印字頭不當加熱溫度過高,列印出來的條碼往往產生缺陷如,斷針、模糊,或是人為操作錯誤產生異常情形如,條碼位置偏移、條碼重複。本論文使用影像處理技術來攔檢條碼機產生的異常與人為產生的異常,避免產品出貨後造成的客訴案件如,條碼讀取不良、條碼位置偏移、條碼重複。 本系統硬體項目包括一台條碼印表機、一台雷射條碼讀取裝置並有影像拍照輸出功能、一台電腦、一台機器手臂(Robot)與放置TFT-LCD的牙叉平台,另外軟體項目包括條碼讀取設定軟體AutoID,與本論文開發的檢測軟體。 本論文的研究是針對TFT-LCD於包裝製程玻璃表面貼附條碼,當讀取條碼時產生的影像檔來進行研究,利用電腦視覺及影像處理的技術來進行條碼列印後產生的影像來辨識其缺陷,以期望達到預警的功能,提早發現條碼缺陷的品質異常,避免產品出貨到客戶端產生客訴的情形,本論文研究的內容先介紹TFT-LCD 包裝製程與條碼缺陷分析,後續對本文使用的電腦視覺理論及影像處理實作一一說明,其次說明本文的研究結果,最後定義本文的結論與未來展望。
Since 1974, the Barcode system has been developed, it was widely applied for industry product, business data and private medical insurance, just as recorded production year, grade and code it was played an important role in the TFT-LCD packing process. Normally the Barcode is produced by Barcode machine and its process is heating the header and then ribbon print on the label. If there were many particle of dirt on the header surface or abnormally heated the header, the ribbon printed broken or blur the Barcode which were classified to NG case. In addition, the other NG case were label offset or duplicate, which were operation miss by user. To avoid customer complain this thesis use image process technology to check the barcode image NG case immediately. The hardware system includes a barcode printer machine, a Barcode reader which contain image output, a computer, a Robot and a platform which put TFT-LCD on it. The software system included Barcode setup software “AutoID” and developed detective software by myself. This thesis research is for using image process technology to check the barcode image NG case in the TFT-LCD packing process. The purpose is wished to find the defect Barcode label before it pass throw into the customer and prevent customer complaining it. In this paper, the major research started form TFT-LCD packing process characteristics and label NG image case analysis. And then the computer vision and image process explained and implemented it .Secondly, the result of analyzes the barcode image .Finally, the conclusion of this thesis research and future development.
文章公開時間: 2017-01-23
Appears in Collections:機械工程學系所



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.