進階搜尋


 
系統識別號 U0026-0812200913524350
論文名稱(中文) 使用遞迴模糊蜂巢式細胞神經網路於缺陷檢測
論文名稱(英文) Defect Inspection Using Recurrent Fuzzy Cellular Neural Networks
校院名稱 成功大學
系所名稱(中) 電機工程學系碩博士班
系所名稱(英) Department of Electrical Engineering
學年度 95
學期 2
出版年 96
研究生(中文) 顏仲甫
研究生(英文) Chung-fu Yen
電子信箱 n2694113@mail.ncku.edu.tw
學號 n2694113
學位類別 碩士
語文別 英文
論文頁數 51頁
口試委員 指導教授-王振興
口試委員-邱瀝毅
口試委員-蔣榮先
中文關鍵字 缺陷檢測  蜂巢式細胞神經網路 
英文關鍵字 Defect inspection  Cellular neural networks 
學科別分類
中文摘要 使用人類的視覺來處理影像和辨識影像是有限的。在電子工廠的生產線上,對於產品的缺陷檢測是一個重要的過程,然而使用人工的方式來檢查產品的錯誤,是一種不切實際的方式。因此,針對產品缺陷檢測,我們需要開發一些藉由電腦取代人力的方法。我們提出了一個整合平行蜂巢式細胞類神經網路(CNNs)的架構利用產品的影像來解決缺陷檢測處理的問題。我們的架構不同於一般遞迴模糊蜂巢式細胞神經網路;在架構學習部份,我們利用K-means 群聚演算法來做為模糊規則的歸屬函數之分類。在參數學習部份,我們利用ordered derivatives 之遞迴參數學習法來獲得遞迴模糊蜂巢式細胞神經網路中模板的最佳化參數。為了驗證本論文所提出的架構,我們測試許多不同的缺陷並增加影像檢測的困難度,並與基因演算法訓練所得之蜂巢式細胞神經網路的模板做比較。由模擬結果證明了我們所提出的架構,對於偵測錯誤區塊困難度較高的影像有較好的表現。
英文摘要 The use of human vision for defect inspection from product images is limited to a certain quality level. In electronics industrial production lines, it is important to inspect the products for defects. It is feasible to check for product defects in the production lines by artificial means. Therefore, there is a need to develop methods using computer intelligence to replace manpower for product defect identification. We propose a framework to integrate a set of CNNs in parallel for solving defect identification as image processing problems. Our framework was modified from a generic recurrent fuzzy cellular neural network (RFCNN) that consists of a set of fuzzy IF–THEN rules. We employ a k-means algorithm for constructing the antecedent and consequent parts in the structure learning. To obtain the parameters of CNN templates, we derive a recurrent parameter learning algorithm based on ordered derivatives. We name our network modified RFCNN, mRFCNN. To validate the effectiveness of the proposed mRFCNN, we experiment different types of defects and compare our approach with a conventional defect inspection method, wherein CNN templates are trained by genetic algorithms (GACNN). The results of the experiments conclude that mRFCNN, compared to GACNN, has superior performance on more difficult image processing tasks.
論文目次 CHINESE ABSTRACT i
ABSTRACT ii
Acknowledgement iii
LIST OF FIGURES vi
1 Introduction 1-1
1.1 Motivation 1-1
1.2 Literature Survey 1-2
1.3 Purpose of the Study 1-5
1.4 Organization of the Thesis 1-6
2 Cellular Neural Networks and Genetic Algorithms for CNN Template Learning 2-1
2.1 The Structure of Cellular Neural Networks 2-1
2.2 The CNN Templates for Image Processing 2-5
2.2.1 Average Template 2-5
2.2.2 Edge Detection Template 2-6
2.2.3 Isolated Black Pixel Extraction 2-7
2.2.4 Convex Corner Detection Template 2-8
2.3 Genetic Algorithms 2-9
2.4 Genetic Algorithms for CNN Template Learning 2-13
3 Modified Recurrent Fuzzy Cellular Neural Networks 3-1
3.1 Structure of RFCNN 3-1
3.2 Structure and Parameters of mRFCNN Learning Algorithms 3-5
3.2.1 Structure Learning Algorithm of mRFCNN using K-means Clustering 3-6
3.2.1.1 K-means Clustering Algorithm 3-6
3.2.1.2 Input-Output Space Partitioning 3-7
3.2.2 Parameter Learning Algorithm of mRFCNN by Ordered Derivative Calculus 3-7
4 Simulation Results and Discussion 4-1
4.1 Simulation Results 4-1
4.2 Discussion 4-11
5 Conclusions and Future Work 5-1
References
參考文獻 [1] T. Amano, “Correlation based image defect detection,” in Proc. 18th Int’l. Conf. Pattern Recognition, vol. 1, pp. 163-166, 2006.
[2] C. J. Lu and D. M. Tsai, “Automatic defect inspection for LCDs using singular value decomposition,” Int. J. Adv. Manuf. Technol., vol. 25, no. 1-2, pp. 53-61, 2005.
[3] L. O. Chua and L. Yang, “Cellular neural networks: Theory,” IEEE Trans. Circuits and Systems, vol. 35, no. 10, pp. 1257-1272, 1988.
[4] L. O. Chua and L. Yang, “Cellular neural networks: Applications,” IEEE Trans. Circuits and Systems, vol. 35, no. 10, pp. 1273-1290, 1988.
[5] Cs. Rekeczky, T. Roska, and A. Ushida, “CNN-based difference-controlled adaptive nonlinear image filters,” International Journal of Circuit Theory and Applications, vol. 26, no. 4, pp. 375-423, 1998.
[6] Cs. Rekeczky, A. Tahy, Z. Vegh, and T. Roska, “CNN based spatio-temporal nonlinear filtering and endocardial boundary detection in echocardiography,” International Journal of Circuit Theory and Applications, vol. 27, no. 1, pp. 171-207, 1999.
[7] F. Zou, S. Schwarz, and J. A, Nossek, “Cellular neural network design using a learning algorithm,” in Proc. IEEE Int’l. Wkshp. Cellular Neural Networks Application, pp. 73-82, 1990.
[8] T. Kozek, T. Roska, and L. O. Chua, “Generic algorithm for CNN template learning,” IEEE Trans. Circuits and Systems-I, vol. 40, no. 6, pp. 392-402, 1993.
[9] D.-C. Lim, S.-Y. Yun, B.-Y. Jung, and C. K. Hong, “Optical inspection method of lead frame using mathematical morphology,” in Proc. SPIE – The International Society for Optical Engineering, vol. 4564, pp. 107-114, 2001.
[10] S. C. Kim and T. J. Kang, “Automated defect detection system using wavelet packet frame and Gaussian mixture model,” Journal of the Optical Society of America A: Optics and Image Science, and Vision, vol. 23, no. 11, pp. 2690-2701, 2006.
[11] D.-C. Lim, D.-G. Seo, and D.-H. Jeong, “Defect classification for the inspection of TFT LCD glass,” in Proc. SPIE – The International Society for Optical Engineering, vol. 6051, pp. 60510F-1-60510F-6, 2005.
[12] F. Colodro and A. Torralba, “Cellular neuro-fuzzy networks (CNFNs) a new class of cellular networks,” in Proc. Fifth IEEE Int’l. Conf. Fuzzy Systems, vol. 1, pp. 8-11, 1996.
[13] T. Roska and L. O. Chua, “The CNN universal machine: An analogic array computer,” IEEE Trans. Circuits and Systems-II, vol. 40, no. 3, pp. 163-173, 1993.
[14] T. Yang, L. B. Yang, C. W. Wu, and L. O. Chua, “Fuzzy cellular neural networks: Theory,” in Fourth IEEE Int’l Workshop Cellular Neural Networks and Their Applications, pp. 181-186, 1996.
[15] T. Yang, L. B. Yang, C. W. Wu, and L. O. Chua, “Fuzzy cellular neural networks: Applications,” in Proc. CNNA-96, pp. 225-230, 1996.
[16] M. Balsi and F. Voci, “Implementation of fuzzy rule based image processing on the CNN universal machine,” in European Conference Circuits Theory and Design (ECCTD), pp. 1167-1170, 1999.
[17] F. Zou and J. A. Nossek, “Design method for cellular neural network with linear relaxation,” in Proc. IEEE Int’l. Symp. Neural Networks, vol. 3, pp. 1323-1326, 1991.
[18] C. Guzelis and S. Karamahmut, “Recurrent perceptron learning algorithm for CNNs with application to edge detection,” in Proc. IEEE Int’l. Conf. Neural Networks, vol. 2, pp. 1134-1139, 1995.
[19] C. T. Lin, C. L. Chang, and W. C. Cheng, “A recurrent fuzzy cellular neural network system with automatic structure and template learning,” IEEE Trans. Circuits and Systems-I, vol. 51, no. 5, pp. 1024-1035, 2004.
[20] A. Hyvarinen, J. Karhunen, and E. Oja, Independent Component Analysis. New York: Wiley, 2001.
[21] C. Jutten and J. Herault, “Independent components analysis (ICA) versus principal components analysis,” in Signal Processing IV: Theories and Applications, EUSIPCO-88, J. Lacoume et al., Eds, Amsterdam, The Netherlands: Elsevier, pp. 643–646, 1988.
[22] J. B. MacQueen, “Some methods for classification and analysis of multivariate observations,” in Proc. 5th Berkeley Symposium Mathematical Statistics and Probability, pp. 281-297, 1967.
[23] T. Kanungo, D. M. Mount, N. S. Netanyahu, C. D. Piatko, R. Silverman, and A. T. Wu, “An efficient k-means clustering algorithm: Analysis and implementation,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 24, no. 7, pp. 881-892, 2002.
[24] J. H. Holland, Adaptation in natural and artificial systems. Ann Arbor: University of Michigan Press, 1975.
論文全文使用權限
  • 同意授權校內瀏覽/列印電子全文服務,於2009-08-07起公開。
  • 同意授權校外瀏覽/列印電子全文服務,於2009-08-07起公開。


  • 如您有疑問,請聯絡圖書館
    聯絡電話:(06)2757575#65773
    聯絡E-mail:etds@email.ncku.edu.tw