本論文提出一套基於自組織映射圖網路(Self-Organizing Map Neural Network)的被動式自動對焦系統，利用收集的物體影像資料建立銳利値(Sharpness value)曲線，並從曲線中決定自組織映射圖網路的訓練範例，根據訓練範例學習出內在集群模型並利用其結果預測物體可能對焦位置，接著使用百分比下降搜尋法(Percentage Drop Search)找出正確對焦位置，並探討不同對焦搜尋方法間的搜尋次數、對焦速度及對焦穩定性，最後由實驗結果證明此對焦方法可達到快速、穩定的結果。本論文將此被動式自動對焦系統應用在渦卷及手機背殼的自動光學檢測(Automatic Optical Inspection, AOI)上。在渦卷自動光學檢測系統中，使用此被動式自動對焦系統對焦渦卷後，接著藉由模板匹配(Template matching)演算法及全景拼圖(Panoramic image stitching)技術進行渦卷定位與渦卷子影像合併，以便進行後續檢測。在手機背殼自動光學檢測系統中，使用此被動式自動對焦系統對焦手機背殼後，手機背殼檢測的部分則利用設計的濾波器進行背景雜訊濾除，接著使用改良式大津演算法(Otsu’s)進行分割前景與背景來萃取刮痕瑕疵。

A passive auto-focus system is proposed in this dissertation based on Self-Organizing Map Neural Network (SOM NN) for automatic optical inspections (AOI). Image data are collected to build sharpness value curves. From the curves, training samples for SOMNN are determined. An inner-cluster model is built regarding the training samples. Based on the model, candidate of the focused position on the object are predicted. Then using Percentage Drop Search to search the best focused position. By using different focus search methods, the number of searches, a speed of focusing, and result robustness are obtained for further investigations. From experimental results, the proposed system performed quickly and robustly. The applications of this proposed system are AOI of (1) a scroll profile and (2) a back shell of a cellphone. During (1), this proposed system is adopted before the scroll profile needs to be alignment with a template matching algorithm. Moreover, all piecewise images are combined with a panoramic image stitching method to produce a whole image of the scroll profile for further work. During (2), with our proposed method, auto-focusing on the back shell is also achieved. With a design filer, background noise is removed. Then, a back shell mura is extracted after foreground and background are dichotomized with an improved Otsu method.

[1] A. Adams, "The Camera," New York Graphic Society, Boston, 1980.
[2] H.U. Bauer and K. Pawelzik, "Quantifying the Neighborhood Preservation of Self-Organizing Feature Maps," IEEE Transaction Neural Networks, Vol. 3, No. 4, pp. 570-579, 1992.
[3] M. Basu, "Gaussian-Based Edge-Detection Methods-a Survey," IEEE Transaction on System, Man, and Cybernetics, Part C: Applications and Reviews, Vol. 32, No. 3, pp. 252 -260, 2002.
[4] J. Canny, "A Computational Approach to Edge Detection," IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 8, No. 6, pp. 679-698, 1986.
[5] N.N.K. Chern, P.A. Neow and J.M. H. Ang, "Practical Issues in Pixel-Based Autofocusing for Machine Vision," Proceedings of the IEEE International Conference on Robotics and Automation, Vol. 3, pp. 2791-2796, 2001.
[6] L.J. Dong, G. Yu, P. Qgunbona and W.Q. Li, "An Efficient Iterative Algorithm for Image Thresholding," Pattern Recognition Letters 29, pp. 1311–1316, 2008.
[7] M. Fischler and R. Bolles, "Random sample consensus: A Paradigm for Model Fitting Applications to Image Analysis and Automated Cartography", Proceedings: Image Understanding Workshop, pp. 71-88, 1980.
[8] J.L. Fan, B. Lei, "A Modified Valley-Emphasis Method for Automatic Thresholding," Pattern Recognition Letters 33, pp. 703-708, 2012.
[9] J. Gong, L. Li and W. Chen, "Fast Recursive Algorithms for Two-Dimensional Thresholding," Pattern Recognition 31, pp. 295–300, 1998.
[10] T. Hohonen, Self-Organizing Maps, Springer-Verlag, Berlin, 1995.
[11] J. He , R. Zhou and Z. Hong, “ Modified Fast Climbing Search Auto-Focus Algorithm with Adaptive Step Size Searching Technique for Digital Camera,” IEEE Transaction Consumer Electron., Vol. 49, No. 2, pp. 257 -262, 2003.
[12] R.A. Jarvis, "Focus Optimization Criteria for Computer Image Processing," Microscope, Vol. 24, No. 2, pp, 163-180, 1976.
[13] E. Jacobs, "Auto-Focus Cameras," Popular Science, Vol. 217, No. 6, pp. 96-99, 1980.
[14] B. Jahne, H. Scharr, and S. Korke, "Principles of Filter Design," in Handbook of Computer Vision and Applications, Vol. 2, pp. 125-151, 1999.
[15] E.P. Krotkov, "Active Computer Vision by Cooperative Focus and Stereo," New York: Springer-Verlag, 1989.
[16] H.Y. Kim, "Rotation-Discriminating Template Matching Based on Fourier Coefficients of Radial Projections with Robustness to Scaling and Partial Occlusion," Pattern Recognition, Vol. 43, No. 3, pp. 859-872, 2010.
[17] D.G. Lowe, "Object Recognition from Local Scale-Invariant Features," Proceeding Seventh International, International Conference Computer Vision, pp. 1150-1157, 1999.
[18] F. Morii, "An Image Thresholding Method using a Minimum Weighted Squared Distortion Criterion," Pattern Recognition 28, pp.1063–1071, 1994.
[19] S.K. Nayar and Y. Nakagawa, "Shape from Focus," Proceedings of the IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 4, No. 8, pp. 824-831, 1994.
[20] I.E. Sobel, "Camera Models and Machine Perception", PhD thesis, Stanford University, 1970.
[21] J.M. Tenenbaum, Accommodation in Computer Vision, Ph.D. thesis, Stanford University, 1970.
[22] C. Tomasi and R. Manduchi, "Bilateral Filtering for Gray and Color Images," Proceeding Sixth International, International Conference Computer Vision, pp. 839-846, 1998.
[23] H.C. Wei, Y.H. Chien, W.Y. Hsu, Y.C. Cheng and G.D.J. Su, "Controlling a MEMS Deformable Mirror in a Miniature Auto-Focusing Imaging System," IEEE Transaction Control System Technology, Vol. 20, No. 6, pp. 1592 -1596, 2012.
[24] H.H. Yao, "Fast Percentage Drop Search," Technical Reports, Tongtai Machine & Tool Company Limited, 2014.