透過您的圖書館登入
IP:18.220.160.216
  • 學位論文

具位移容忍之快速印刷電路板線路瑕疵檢測法

A fast, shift-tolerable approach for PCB conductive paths inspection

指導教授 : 蔡篤銘
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


由於元件的微型化和密集化之發展趨勢,促依靠人工已經不可能對分佈細密的元件進行可靠而一致的檢測,經過多年的努力,自動光學檢測系統(Automated Optical Inspection; AOI)已成功地運用在印刷電路板(Printed Circuit Board; PCB)生產線上。在傳統的印刷線路檢測方法中,部份的檢測方法雖然可以檢測大多數的印刷電路板,但是需要事先儲存參考影像作為比對的依據,而且待測物對位移的容忍度也比較低。而另外不需要參考影像作為比對的檢測方法,雖然比較不易受位移的影響,但是卻存在著只能檢測有限瑕疵的缺失。 為了提高印刷線路瑕疵檢測的速度與位移容忍度,本研究所發展的檢測方法將分成二個階段。第一階段是屬於線外的學習階段,主要針對PCB板上線形線路、轉角、圓形銲墊及矩形銲墊等四種類別之線路進行分割,並紀錄每類印刷線路的位置與線路類別等相關資訊於標準影像參考表。第二階段則是線上的檢測階段,是根據學習階段所建立的標準影像參考表,對待測的印刷電路板影像依據每個線路之分割區域進行快速瑕疵檢測。本研究採用投影量的量測對線形線路、圓形銲墊與模板比對方法對矩形銲墊、轉角等方式對印刷線路進行瑕疵檢測,但與傳統模板比對法不同的是,本方法直接由待測物產生比對模板而不需事先儲存參考模板,在檢測效果上,對待測印刷電路板影像位移的容忍度接近最小線距的1/2,以及可以快速有效檢測出短路(Short)、斷路(Open)、線路突出(Spur)、缺口(Mousebite)、針孔(Pinhole)、線寬等重大瑕疵。

並列摘要


In this study we propose a fast and shift-tolerable machine vision scheme for detecting conductive paths on printed circuit boards (PCBs), especially the multiplayer PCBs that comprise straight line segments, interested corners between two line segments, circular and rectangular pads. The proposed method involves two processes: training and detecting. In the training process, a faultless PCB image is used to automatically construct a reference table that contain the inspection frame and location, method of inspection and specification of each segment in the PCB. In the detection process, each inspection frame obtained from the reference table allows the target segment under inspection to be shifted within , where is the line spacing. The inspection of line segments and circular pads are based on a simple computation of projection. The inspection of corners and rectangular pads are based on self-reference pattern matching, where the reference pattern are generated from the test PCB so that it is invariant to translation. Experimental results have shown that the proposed method is computationally efficient to detect the defects of open, short, mousebite, spurs, holes, line width and line spacing.

參考文獻


1. Borba, J. F. and J. Facon, 1995, “A printed circuit board automated inspection system,” Cir-cuits and Systems, Proceedings of the 38th Midwest Symposium, 1, pp. 69-72.
2. Chen, Y. T. and S. M. Thomas, 1989, “A simple approach for the estimation of circular arc center and its radius,” Computer Vision, Graphics, and Image Processing, Vol. 45, pp.362--370.
5. Gonzalez, R. C. and R. E. Woods, 1992, Digital Image Processing, Addison-Wesley, New York, NY.
8. Jain, A. K., N. K. Ratha and S. Lakshmanan, 1995, “Object detection using Gabor filter,” Pattern Recognition, Vol. 30, pp. 295-309.
9. Lee, D. T., 1978, “A computerized automatic inspection system for complex printed thick film patterns,” SPIE ─ Applied Electronic Imaging System, 143, pp.172-177.

延伸閱讀