台灣為電子資訊大國，舉凡各種電子產品如主機板、數位相機、手機乃至於晶片產品，台灣皆有製造。在其生產流程中，必有產品缺陷檢測。如晶片封裝時的印字不清、引腳斷裂，主機板元件如電容、電阻焊接不良，產品外殼烤漆不佳…等等，都必須及時發現，期能在產品出貨前，發現缺點、提升良率。
數十年來有關電腦視覺應用的論文不斷被提出，從人臉辨識、物體追蹤、數位影像處理乃至各種缺陷檢測自動化，各領域的研究人員無不費盡心思，在講求成本降低的工業生產領域中，各廠商亦紛紛投入電腦視覺研究，期望以此取代部份人工。台灣為一電子產品生產大國，對此一問題之重視自不在話下，本論文著眼於此，特別專注於一般電子產品生產流程中最需要的印刷電路板缺陷檢測。
本論文以有別於以往的檢測方式，使用最近廣為使用的”小波轉換”理論，作為演算法基本架構，該法可將空間域信號轉為頻率域信號，提供使用者可針對某特定頻率範圍做處理之能力，而演算法亦根據此特性將屬較低頻信號的缺陷分離出來。其最大優點是運算簡易、執行快速，只需將經過影像處理及閥值化後正確的影像與含有缺陷的影像同時輸入本演算法，便可求得缺陷之所在位置。

Taiwan is a well-known country of affluent electrical information; electronic products such as main board, digital camera, mobile phone, and SOC chip, can all be made by the technique in Taiwan. In the process of manufacturing, the need of defect checking has always been inevitable. Product defects might occur in the following cases, the blurred print, the splitting connector during the packing of the chip, and the ill-welding of those main board components such as the capacitor and the resistor. Those defects must be detected before being sent to the market so that good quality can be maintained.

In recent years, papers pertaining the applications of computer vision have been submitted. From the recognition of the human face, the object tracing, the digital image processing, as well as the automation of various defect checking, the researchers of all fields have tried their best to explore every possibility to their applications. In the fields of industrial manufacturing in which the reduction of the cost is strongly stressed, factory owners have invested in the research work of computer vision which can take over the human labor. Taiwan, as a major country in electronic manufacturing, puts no less stress on these fields of research. Having witnessed such a tendency of industrial manufacturing, we devote ourselves to the defect defection in printed circuit board that is one of the most important steps during the general process of producing electrical products.

Our research work uses special methods which are different from those former ones through the use of the current widely used theory, the wavelet transformation. As a basic construction of algorithm, this method can transform signals in time domain into signals in frequency domain, and can also provide the users the processing abilities for a specific range of frequency. Based on these properties, algorithm can be designed to pick up the defects of lower frequencies. Its excellency lies in its easy operation and speedy performance. Only when the two images, i.e. the processed thresholded image and the image with defects, are manipulated simultaneously by the algorithm, can we be certain to locate where the defects are.

[1] Bently, “The inspection: an Automatic optical printed circuit board (PCB) inspection”, Proc. Soc. Photo-Optical Instrumn. Engrs., 201, Page(s).37-47, 1979.

[2] Da Newman, G. Vachtsevanos, “On-line Inspection of Surface Mount Devices Using Vision and Infrared Sensors”, Autotest conference 95. Systems Readliness:Test Technology f7 the 21st Century,Page(s):376-384, 1995

[3] Eric J. Stollnitz, Tony D. DeRose, David H. Salesin, “Wavelets for Computer Graphics: A Primer, Part1“, IEEE Computer Graphics and Applications, Tutorial, May 1995

[4] E. K. Khwang, D. P. Mital, “A transputer-based Automated Visual Inspection system for Electronic Devices and PCBs” Optics and Laser in Engineering, Vol. 22, Page(s). 161-180,1995.

[5] J. S. Park, J. T. Tou, “A Solder Joint Inspection System for Automated Printed Circuit Board Manufacturing”, Proceedings of IEEE International Conf. Robotics and Automation, Vol. 2, Page(s)1290-1295, 1990.

[6] M. Moganti, F. C. Ercal, H. Dagli, S. Tsunekawa, “Automatic PCB inspection algorithms：a survey ”, Computer Vision and Image understanding, Vol. 63, No.2, Page(s):287-313, 1996

[7] M. Nikaido, “ Recognition of pattern defects of printed circuit board using
topological information” Electronics Manufacturing Technology Symposium, 1991., Eleventh IEEE/CHMT International , Page(s): 202 –206,16-18 Sep 1991.

[8] Mauro Hiromu Tatibana, Roberto De Alencar Lotufo, “Novel Automatic PCB Inspection Technique Based on Connectivity“,Computer Graphics and Image Processing 1997 Proceeding, X Brazilian Symposium, Page(s) 14-17 1997.

[9] Mitchell D. Swanson, Student Member, Ahmed H. Tewfik, “A Binary Wavelet Decomposition of Binary Images“, IEEE Transaction on Image Processing, Vol. 5, No. 12, December 1996.

[10] M. A. Sutton, M. Cheng, W. H. Peter, Y. J. Chao and S. R. Mchneill, “Application of an optimized digital correlation method to planar deformation analysis”, Image and Vision Computing, Vol. 4, No. 3, Page(s). 143-150,1986.

[11] R. Z. Liu, Y. Q. Shi, W. F. KoSonocky, F. P. Higgins, “Infrared Solder Joint Inspection on Surface Mount Printed Circuit Boards”, Proceeings of 38th Circuits and Systems, Midwest Symposium Usa, Vol. 1, Page(s) 145-148, 1996.

[12] Tetsuki Yaniguchi, Dariusz Kacprzak, Sotoshi Yamada, and Masayoshi Iwahara, “Wavelet-Based Processing of ECT Images for Inspection of Printed Circuit Board“, IEEE Transactions on magnetics, Vol. 37,No. 4, July 2001.

[13] W. E. McIntosh, “Automatic the Inspection of Printed Circuit Boards”, Robotic Today, Page(s):75-78, Jun.1983.

[14] Z. Ibrahim, Al-Attas , S.A.R. Aspar, Z., Mokji, M.M., “Performance evaluation of wavelet-based PCB defect detection and localization algorithm” Industrial Technology, 2002. IEEE ICIT '02. 2002 IEEE International Conference on , Vol: 1 Page(s): 226 –231 , 2002.

[15] Milan Sonka, Vaclav Hlavac, Roger Boyle, “image processing, Analysis and Machine Vision_Second Edition”，Brooks/Cole Publishing Company，USA，1999

[16] Maria Petrou, Panagiota Bosdofianni, “Image Processing_The Fundamentals”，Wiley，England，1999.

[17] 彭光裕，“應用電腦視覺技術於表面元件印刷電路板之自動檢測系統設計及開發”，國立交通大學碩士論文，2000

[18] 邱振洋，”應用電腦視覺於IC瑕疵之研究”，國立雲林科技大學，碩士論文，2000

[19] 彭瑞勳，”移動框法在電路板上IC的檢測”，國立交通大學碩士論文，2000

[20] 許嘉豪，”複雜紋理之瑕疵檢測”，國立台灣科技大學碩士論文，2000

[21] 許友南，”機器視覺為基礎之焊錫瑕疵偵測與分類“，元智大學碩士論文，2000

[22] 林千智，”獨立式動態影像判別與儲存系統”，國立成功大學碩士論文，2001

[23] 劉權霈，”應用電腦視覺技術於PCB自動檢測系統之設計及開發”，國立交通大學碩士論文，2001

[24] 許辰合，”印刷電路板基本元件圖像之萃取系統”，國立成功大學碩士論文，2002

[25] 單維彰，”凌波初步”，全華，台灣，1999

[26] 吳成柯,戴善榮,程湘君,雲立實 譯，”數位影像處理”，儒林，台灣，2000

[27] 陳同孝,張真誠,黃國峰，”數位影像處理技術”，松崗，台灣，2001