臺灣博碩士論文加值系統

有機發光層是有機發光二極體的主要發光層，該層為透明薄膜層，若要利用人工檢測方式極為困難且容易失誤，故此目前極需建立一系統，使得專業檢測能更具標準化與效率性。所以本論文針對有機發光層瑕疵外觀，應用影像處理來開發一套自動檢測系統。而檢測的有機發光層瑕疵可分為四種類型：包括刮痕、氣泡、塗佈不均、雜點等四類。因有機發光層為透明層所以需要利用紫外光線激發瑕疵樣本，瑕疵部分會被激發顯影，再進而擷取該瑕疵影像。在影像處理過程中，先利用中央加權中值濾波器減少脈衝雜訊，由於瑕疵影像灰階差異不大，無法以一固定門檻值完整分割所有瑕疵類別，故使用統計式門檻值決定灰階值大小，以選擇一個或兩個最佳門檻值來分割出瑕疵區域，再配合形態學中的閉合運算使瑕疵輪廓更平滑完整，並選擇瑕疵厚度、質心灰階對比和緊緻性做為瑕疵特徵。然而在瑕疵樣本方面，搜集80比瑕疵樣本，利用模糊類神經建立資料庫且分類，並規劃以不同數量之訓練樣本與固定的測試樣本進行實驗，結果顯示在訓練樣本為30筆以上時，其辨識率皆可達到100%，驗證了模糊類神經網路在訓練樣本足夠的資料庫下，可獲得相當準確的辨識率，成功被應用於OLED有機發光層瑕疵自動檢測系統。

The organic emitting layer is a transparent film layer. It’s difficult to inspect the emitting layer defects by the human and easily leads to fault inspection. The standardized OLED (Organic Light Emitting Diode) defect detection system must be built. This thesis applies image processing and appropriate classifiers to develop an image inspection system for Organic emitting layer defects. The organic emitting layer defects can be divided into four groups: scratch, air bubble, dust particle and coating nonuniformity. Because the organic emitting layer is a transparent film layer, we use ultraviolet light (UV) to excitive the defect sample. The defect area will be excited and we capture to be digital image. In image processing, the center weighted median filter is used to reduce the impulse noise of images. Since the gray values of all the defects are close to each other, we cannot segment entire defect-areas of four defects by a fixed threshold value. Therefore, the statistical threshold value decision method is used to choose one or two optimal threshold values with the difference of gray values in image segmentation, and then we use the closing operator in morphology to smooth the contour of defects.
The compactness, thickness difference and centroid of gray scale contrast are selected as defect features. We divide 80 defect samples into two groups, which are the different amount of training samples and the fixed amount of testing samples for experiment. Finally, we have a database with the Fuzzy Neural Network (FNN), which is trained by the training samples, and recognize the testing samples with the database. The result shows that the recognition rate is 100% when the amount of training samples is more than or equal to 30：The FNN can achieve a high recognition rate with enough training samples in the database, and it can be successfully applied to the inspection of the organic emitting layer defects of OLED.

目錄
摘要 I
ABSTRACT III
誌謝 III
目錄 VI
圖目錄 X
表目錄 XII
第1章 緒論 1
1.1 研究動機與目的 1
1.2 研究步驟與方法 3
1.3 論文架構 4
第2章 文獻回顧 5
2.1 文獻回顧 5
2.1.1 有機發光層發光層螢光效應 5
2.1.2 影像品質提升 5
2.1.3 影像分割 6
2.1.4 特徵值擷取 7
2.1.5 分類器 7
第3章 有機發光層 9
3.1 有機發光二極體原理與結構 9
3.1.1 有機發光二極體 9
3.1.1.1 有機發光二極體優點 10
3.1.2 發光原理 10
3.2 有機發光層種類 12
3.3 有機發光層製造方法 13
3.3.1 有機發光二極體製造流程 13
第4章 數位影像處理 17
4.1 數位影像處理分析架構 17
4.2 空間濾波 18
4.2.1 低通濾波器 19
4.2.1.1 平滑濾波器 19
4.1.1.2中值濾波器 21
4.1.1.3中央加權中值濾波器 22
4.2.2 高通濾波器 22
4.2.2.1 拉普拉斯算子 23
4.2.2.2 索貝爾算子 24
4.3 影像分割 25
4.3.1 門檻值法 25
4.3.2 統計式門檻值決定法 26
4.4 形態學 29
4.4.1 標記化 29
4.4.2 侵蝕 31
4.4.3 膨脹 31
4.4.4 斷開 32
4.4.5 閉合 33
第5章 分類器原理 34
5.1 糢糊類神經網路 34
5.1.1 糢糊理論 34
5.1.2 語意變數糢糊集合 36
5.1.3 歸屬函數的種類 37
5.2 模糊類神經網路基本原理 39
5.2.1 各層神經元演算法 40
第6章 實驗過程 44
6.1 實驗軟硬體架構 44
6.2 硬體設備 44
6.3 應用軟體 45
6.4 實驗步驟與流程 47
6.5 有機發光層瑕疵樣本 49
6.6 瑕疵影像分割 51
6.7 瑕疵影像的形態運算 52
6.8 瑕疵特徵擷取 53
6.9 瑕疵分類 58
6.9.1 瑕疵分類器參數設定 58
6.9.2 瑕疵樣本設計 61
6.9.3 瑕疵分類結果與討論 63
第7章 結論 65
7.1 結論 65
參考文獻 67

參考文獻
[1]F. J. Lin, R. J. Wai and H. P. Chen, “A PM synchronous servo motor drive with an on-line trained fuzzy neural network controller,”IEEE Trans. Energy Conversion, vol. 13, no. 4, pp. 319-325, 1998.
[2]F. J. Lin, W. J. Hwang and R. J. Wai, “A supervisory fuzzy neural network control system for tracking periodic inputs,” IEEE Trans. Fuzzy Systems, vol. 7, no.1, pp. 41-52, 1999.
[3]Y. C. Chen and C. C. Teng, “A model reference control structure using a fuzzy neural network,” Fuzzy Sets and Systems, vol. 73, pp. 291-312, 1995.
[4]L. X. Wang, “A Course in Fuzzy Systems and Control,”New Jersey: 122 Prentice-Hall, 1997.
[5]K. S. Narendra and K. Parthasarathy, “Identification and control of dynamical systems using neural networks,” IEEE Trans. Neural Networks, vol. 1, no. 1, pp. 4-27, 1990.
[6]J. Zhang and A. J. Morris, “Recurrent neural-fuzzy networks for nonlinear process modeling,” IEEE Trans. Neural Networks, vol. 10, no. 2, pp. 313-326, 1999.
[7]C. Hecht and G. Dishon, “Automatic optical inspection (AOI),” Proceedings of the 1990 40th Electronic Components and Technology Conference, vol. 1, pp. 659-661, 1990.
[8]楊富名, “紫外光照射處理對有機半導體電特性影響之研究,” 國立彰化師範大學光電科技研究所, 2007.
[9]C. J. Huang, C. F. Wu and C. C. Wang, “Image processing techniques for wafer defect cluster identification,” IEEE Design & Test of Computers, vol. 19, no. 2, pp. 44-48, 2002.
[10]S. J. Ko and Y. H. Lee, “Center weighted median filters and their applications to image enhancement,” IEEE Trans. on Circuits and Systems, vol. 38, no. 9, pp. 984-993, 1991.
[11]M. Unser, A. Aldroubi and M. Eden, “Polynomial spline signal approximations: filter design and asymptotic equivalence with shannon’s sampling theorem,” IEEE Trans. Information Theory, vol. 38, no. 1, pp. 99-103, January 1992.
[12]劉鴻明,“Splines 應用於影像內插法之研究,” 中原大學資訊工程學系碩士學位論文, 2005.
[13]N. Otsu, “A threshold selection method form gray level histogram,” IEEE Trans. on System ,Man ,and Cybernetics , Sem 8,1978, pp. 62-66.
[14]J. N. Kapur, P. K. Sahoo and A. K. C. Wong, “A new method for gray-level picture thresholding using the entropy of the histogram,” Computer Vision, Graphics and Image Processing, vol. 29, no. 3, pp. 273-285, 1985.
[15]T. Markiewicz, L. Moszczynski, and S. Osowski, “Myelogenous leukemia cell image preprocessing for feature generation,” V Int. Workshop –Computational 72 Methods in Electrical Engineering, Jaznowiec, 2003, pp.70-73.
[16]M. Partio, B. Cramariuc “Rock texture retrieval using gray level Co-Occurrence matrix,” Proceedings of the 5th Nordic Signal Processing Symposium, 2002.
[17]李玫憶, “細胞次結構影像辨識系統,”私立中原大學醫學工程系碩士學位論文,2005.
[18]P. Pope, H. P. Kallmann and P. J. Magnante, J. Chem. Phys. , 38, 2042 ( 1963 ).
[19]C. W. Tang and S. A. VanSlyke, Appl. Phys. Lett. , 51, 913 ( 1987 ).
[20]K. Sakurai, A. Onoyama, H. Ishii, K. Oka and K. Yamanishi, “Capture rate enhance method of 0.1-μm level defects by pattern-matching inspectors,” IEEE Trans. on Semicond. Manuf., vol. 13, no. 4, pp. 434-441, 2000.
[21]F. J. Lin and C. H. Lin, “On-line gain-tuning IP controller using RFNN,” IEEE Trans. Aerospace and Electronic Systems, vol. 37, no. 2, pp. 655-670, 2001.
[22]C. T. Su, T. Yang and C. M. Ke, “A neural-network approach for semiconductor wafer post-sawing inspection,” IEEE Trans. on Semicond. Manuf., vol. 15, no 2, pp. 260-266, 2002.
[23]游清彥,“電漿輔助化學氣相沉積法製備OLED 元件氣體阻障層之研究,”私立中原大學化學工程學系碩士學位論文,2004
[24]鐘國亮, “偏光膜外觀瑕疵之影像檢測系統開發,” 2003
[25]C. Hecht and G. Dishon, “Automatic optical inspection (AOI),” Proceedings of the 1990 40th Electronic Components and Technology Conference, vol. 1, pp. 659-661, 1990.
[26]陳建堂, “偏光膜外觀瑕疵之影像檢測系統開發,” 國立台灣科技大學高分子工程研究所碩士學位論文,2008.
[27]戴慶輝,“晶片電阻與標籤字元辨識系統之開發,”國立屏東科技大學機械研究所碩士班論文,2000.
[28]陳普中,“紋理分析於瑕疵定位及影像檢索之研究,”國立臺灣科技大學電機工程系碩士學位論文,2005.
[29]陳靜怡,“影像處理及類神經網路於微細胞核自動計數之應用,”元智大學資訊管理學系碩士班碩士論文,2005.
[30]王文賓,“複雜文件影像的文字抽取技術,”,國立交通大學電機資訊學院電機與控制學程碩士論文,2003.
[31]黃柏凱,“以基因演算法為基礎之遞迴式模糊類神經網路控制線型感應馬達驅動系統,”私立中原大學電機工程學系碩士學位論文,2003.
[32]鄭光宏,“應用影像視覺於超薄型表面載式電感器之線上自動檢測,”大葉大學機電自動化研究所碩士班碩士論文,2004.
[33]洪智毅,“交疊錢幣辨識之研究,”南台科技大學資訊工程研究所碩士學位論文,2007.
[34]陳志強,“OLED有機發光二極體顯示器技術,”全華圖書股份有限公司,2006.