臺灣博碩士論文加值系統

光碟產業使用許多自動光學檢測（Automatic Optical Inspection • AOI）設備以確保所生產之光碟沒有瑕疵，然而大部份的AOI設備都是仰賴進口而且價格相當昂貴。除此之外，目前並沒有任何一套設備可以將瑕疵品進一步做分級的工作。有鑑於此，本研究發展一套以線掃瞄取像技術為基礎之光碟片瑕疵偵測與分級系統，冀望能夠降低光碟製造業者購置AOI設備的成本，以提高產業的競爭力。
為了達成上述目標，本研究分兩階段進行。第一階段為硬體設備的規劃，其首要目標是將出現在光碟上的所有瑕疵完整且真實的呈現出來。本階段是本研究成敗的關鍵，因為如果影像中所出現的瑕疵失真或不完整，則後續的工作將變得毫無意義可言。儘管光碟片屬於高反射體，本研究採用亮場成像原理，同時攫取光碟片正反兩面的影像以節省時間。本研究所建構之硬體設備是由兩台線掃瞄攝影機、兩塊影像攫取卡、兩台線性光纖光源、一個步進馬達驅動之旋轉平台所組成。第二階段是利用影像濾波、梯度分析、影像分割、形態處理、物件分析、邊界偵測、及特徵抽取與分析等影像處理技術，發展一套瑕疵檢測程式，將瑕疵分割出來並予以分類。實驗結果顯示，本研究所發展之AOI系統可成功檢查出光碟表面上的瑕疵，包含刮痕、流星、針孔、氣泡、條紋、及沾污等。就P4 3.0G之個人電腦及2048 pixels解析度之線掃瞄攝影機而言，在29 μm的空間解析度下檢測時間約需0.765秒；在117 μm的空間解析度下，檢測時間則只需0.172秒。

Optical disc industry uses many automated optical inspection (AOI) systems to ensure their products defect-free. However, most of them are imported and very expensive. Besides, none of them is capable of providing grading function for the defected discs. In view of this, the object of the research is to develop a linescan-based flaw detection and classification system to reduce the cost for purchasing AOI equipment and increase the competitiveness of domestic DVD industry.
To achieve the above-mentioned objectives, a two-step approach is adopted. The first step is to establish an apparatus to acquire images that will truly show all the defects on the surface of optical discs. This step is crucial to the success of the study, because if the captured defects are incomplete or camera cannot even perceive defects at all, the developed system will be useless. In spite of the high reflecting surface of the optical discs, we apply bright-field imaging technique to capture images of both sides simultaneously to reduce inspection time. The developed imaging system consists of two line-scan cameras, two image grabbers, two line-type fiber-optic light sources, and a rotation stage driven by a stepping motor. The second step is to develop a flaw-detection program by using image-processing techniques such as filtering, segmentation, blob analysis, edge detection, feature extraction and analysis, etc. to detect and classify defects. The experimental results show that the developed AOI system is successful in inspecting optical discs having specular surfaces for defects such as scratches, comets, pinholes, bubbles, streaks, etc. For a P4 3.0G personal computer and a 2048 pixels line-scan camera, if the desired resolution is 29 μm, it takes 0.765 seconds to complete the inspection. On the other hand, if the desired resolution is 117 μm, it takes only 0.172 seconds.

摘 要 i
Abstract ii
誌 謝 iii
目 錄 iv
圖 目 錄 vi
表 目 錄 viii

1. 前言 1
1.1 研究背景 1
1.2 研究動機與目的 7
1.3 文獻回顧 7
1.3.1 現有檢測機台 7
1.3.2 專利 10
1.3.3 國內外研究 11
2. 研究設備與方法 14
2.1 檢測硬體設備之配置 14
2.1.1 線掃瞄攝影機 15
2.1.2 旋轉機構與取像高度 17
2.1.3 線型光纖光源與照明方式 21
2.2 檢測軟體程式及檢測流程 23
2.3 參數設定與位置校正 24
2.3.1 參數設定 25
2.3.2 位置校正 27
3. 影像攫取 31
3.1 半透明DVD 32
3.2 鏡反射DVD 34
3.3 亮暗場照明方式之選擇 35
4. 瑕疵偵測與分類 38
4.1 瑕疵偵測 38
4.1.1 平均灰階值瑕疵檢測法 39
4.1.2灰階標準差瑕疵檢測法 40
4.1.3平均梯度瑕疵檢測法 41
4.1.4多閥值瑕疵檢測法 42
4.1.5 瑕疵偵測法的比較 43
4.1.5.1 瑕疵完整性 43
4.1.5.2 處理時間 47
4.1.5.3 參數易調性 49
4.1.5.4 瑕疵檢測法之選用準則 49
4.2 瑕疵分類與光碟分級 51
4.3 顯示及輸出檢測結果 55
4.3.1結果的顯示 55
4.3.2 輸出結果 59
5. 研究結果與討論 60
5.1 研究結果 60
5.1.1 Dye樣本實驗結果 60
5.1.2 Bond DVD樣本實驗結果 67
5.2 討論 70
6. 結論 72
6.1 結論 72
6.2 未來展望 73
參考文獻 75

1.光電產業2004年回顧與展望，工研院IEK電子組，2005。
2.記錄型光碟片產業2004年回顧與2005年展望，DVD通訊，第96期，2005。
3.Dr. Schenk, “Inspection and Measurement Solutions for CD-R and DVD-R,” Technical White Paper, Dr. Schenk, 2002.
4.Dr. Schenk, “Disk Inspection System Quick User Manual,” Dr. Schenk, 2001.
5.Disc Inspection Technology, http://www.disctech.co.jp/index-e.htm
6.DaTARIUS, ”Compact Disc Optitest User Manual Version 1.3,” DaTARIUS, 1994.
7.Basler Vision Technologies, http://www.baslerweb.com/index.html
8.M.E. Gardiner and D.W. Kuntz, “Optical Disc Measurement by Refraction,” U.S. Patent No. 4456375, Jun. 26, 1984.
9.W.R. Mueller, L.R. Gensel, and P.V. Hoof, “Optical Disc Inspection Equalization System and Method,” U.S. Patent No. 6154275, Nov. 28, 2000.
10.W.R. Mueller, L.R. Gensel, and P.V. Hoof, “Optical Disc Inspection Equalization System and Method,” U.S. Patent No. 5940174, Aug. 17, 1999.
11.K. Yoneda, “Unit for Inspecting a Surface,” U.S. Patent No. 6538729, Mar. 25, 2003.
12.C.S. Lin, “Evaluation of Defects on an Optical Disc Master Plate,” Optics and Laser Technology, Vol. 29, No. 8, pp. 425-432, 1997.
13.梁俊傑，“DVD與VCD光碟表面瑕疵檢測系統之開發”，碩士論文，國立屏東科技大學機械工程系，2001。
14.G.R. Mather, Jr., “Optical Disc Inspection by Linear Array CCD Systems,” Midwest Symposium on Circuits and Systems, Vol. 2, pp. 1447-1455, 1993.
15.P. LHostis, F. Byers, F. Podio, and X. Tang, “Microscopic Image Analysis of Defect Areas in Optical Disks,” Proceedings of SPIE - The International Society for Optical Engineering, Vol. 3806, pp. 106-112, 1999.
16.E. Vidal, P. Andersen, J. Stoustrup, and T.S. Pedersen, “A Study on the Surface Defects of a Compact Disk,” Proceedings of the 2001 IEEE International Conference on Control Applications, pp. 101-104, 2001.
17.D. Toth, A. Condurache, and T. Aach, “A Two-stage-classifier for Defect Classification in Optical Media Inspection,” Proceedings of the 2002 International Conference on Pattern Recognition, Vol. 4, pp. 373-376, 2002.
18.R. Stojanovic, P. Mitropulos, C. Koulamas, Y. Karayiannis, S. Koubias, and G. Papadopoulos, “Real-time Vision-based System for Textile Fabric Inspection,” Real-Time Imaging, Vol. 7, No. 6, pp. 507-518, 2001.
19.L. Hepplewhite and T.J. Stonham, “Magnetic Disk Inspection using Texture Recognition,” IEE Colloquium on Texture Classification: Theory and Applications, pp. 8/1-8/4, 1994.
20.W. Niblack, “An Introduction to Digital Image Processing,” Prentice Hall, 1986.
21.S.D. Yanowitz and A.M. Bruckstein, “A New Method for Image Segmentation,” Computer Vision, Graphics, & Image Processing, Vol. 46, No. 1, pp. 82-95, 1989.
22.J. Bernsen, “Dynamic Thresholding of Grey-level Images,” Proceedings - International Conference on Pattern Recognition., pp. 1251-1255, 1986.
23.I. Blayvas, A.Bruckstein, and R. Kimmel, “Efficient Computation of Adaptive Threshold Surfaces for Image Binarization,” Pattern Recognition, Vol. 39, No. 1, pp. 89-101, 2006.
24.楊禮鴻，“以機器視覺為基礎之液晶面板點燈檢測法”，碩士論文，中華大學機械與航太工程研究所，2005。
25.林憲忠，“積層陶瓷電容瑕疵檢測與分類”，碩士論文，中華大學機械與航太工程研究所，2004。
26.G. Wang and T.W. Liao, “Automatic Identification of Different Types of Welding Defects in Radiographic Images,” NDT and E International, Vol. 35, No. 8, pp. 519-528, 2002.
27.M.R. Driels and D.J. Nolan, “Automatic Defect Classification of Printed Wiring Board Solder Joints,” IEEE transactions on components, hybrids, and manufacturing technology, Vol. 13, No. 2, pp. 331-340, 1990.
28.C.J. Du and D.W. Sun, “Pizza Sauce Spread Classification Using Colour Vision and Support Vector Machines,” Journal of Food Engineering, Vol. 66, No. 2, pp. 137-145, 2005.
29.X. Zhang, C. Krewet, and B. Kuhlenkotter, “Automatic Classification of Defects on the Product Surface in Grinding and Polishing,” International Journal of Machine Tools and Manufacture, vol. 46, no. 1, pp. 59-69, 2006.
30.莊富傑，“導線架瑕疵之偵測與分類使用類神經網路”，碩士論文，中華大學機械與航太工程研究所，1999。
31.陳冠字，“紋理分析在彩色影像分類上之應用”，碩士論文，中華大學機械與航太工程研究所，2002。
32.N. Otsu, “Threshold Selection Method from Gray-Level Histograms,” IEEE Trans. Syst. Man. Cybern., Vol. SMC-9, No. 1, pp. 62-66, 1979.
33.K.T. Gribbon and D.G. Bailey, “A Novel Approach to Real-time Bilinear Interpolation,” Second IEEE International Workshop on Electronic Design, Test and Applications, pp. 126-131, 2004.