此研究之目的在探討利用次像素插值方法，以提高直接比對法之可行性。我們將利用印刷電路板上之定位點，計算標準影像與待測影像間之平移及旋轉誤差，再經由座標轉換後，利用次像素插值方法計算出參考標準影像，再利用直接比對法將參考標準影像與待測影像直接比對，並將缺陷位置圈選出來。此研究利用三種次像素插值方法：傅立葉轉換、線性內插法、立方內插法，並比較三種插值方法之計算時間、誤差總和平均及誤差平方總和均方根。計算時間上以線性內插法最快，立方內插法次之，傅立葉轉換最慢；在誤差總和平均及誤差平方總和均方根方面，三者間之差異極小。系統測試之印刷電路板為已插件之合格及不合格印刷電路板，不合格印刷電路板包含缺陷種類有橋接、小元件缺件及大元件缺件。測試發現，差異影像除缺陷所造成之影像差異外，還有印刷線條位置之差異，以及反光元件及較高之元件頂端所產生之光線差異。利用適當之遮罩及閥值可以在高(324 pixels/mm2)解析度下，將三種缺陷之位置圈選出來。

This study demonstrates the subpixel interpolation methods to improve the practicable of direct comparison. Three subpixel interpolation methods, fourier transform, linear interpolation method and cubic interpolation method, are applied in this paper. The indices that can evaluate these three methods are concluded as the calculation time, the mean of error sum and the root mean square of error square sum of these three interpolation methods. According to the calculation time, the fastest one is the linear interpolation; instead, the fourier transform is the slowest one. Based on the mean of error sum and the root mean square of error square sum, the differences of these three interpolation methods are very small. Moreover, in the system test, PCBs that include no defects, bridge, missing of small component and missing of large component are tested. After test, in addition to the differences of defects, there are other differences like the print lines on the PCBs and the reflection of light. The position of defects can be marked after using appropriate mask and threshold under the high resolution (324 pixels/mm2).

1. 林育興，「應用機械視覺檢測印刷電路板表面黏著元件演算法之評估與比較」，國立清華大學動力機械研究所碩士論文，2003.
2. Teoh, E. K., Mital, D. P., Lee, B. W. and Wee, L. K.,” Automated Visual Inspection of Surface Mount PCBs,” Industrial Electronics Society, 1990. IECON '90., 16th Annual Conference of IEEE, Vol. 1, pp. 576 -580, Pacific Grove, CA USA, Nov. 1990.
3. Teoh, E. K. and Mital, D. P.,” A Transputer-based Automated Visual Inspection System for Electronic Devices and PCBs,” Optics and lasers in Engineering, Vol. 22, No. 3, pp. 161-180, 1995.
4. Demir, D., Birecik, S., Kurugollu, F., Sezgin, M., Bucak, I. O., Sankur, B. and Anarim, E.,”Quality Inspection in PCBs and SMDs Using Computer Vision Techniques,” 20th International Conference on Industrial Electronics, Control and Instrumentation, 1994. IECON '94., Vol. 2, pp. 857 -861, Bologna, Italy, Sept. 1994.
5. Loh, H. H. and Lu, M. S.; “Printed Circuit Board Inspection Using Image Analysis,” International IEEE/IAS Conference on Industrial Automation and Control: Emerging Technologies, pp. 673 -677, Taipei, Taiwan, May, 1995.
6. 彭光裕，「應用電腦視覺技術於表面黏著元件印刷電路板之自動檢測新系統設計及開發」，國立交通大學工業工程與管理學系碩士論文，2000.
7. Wu, W. Y., Wang, M. J. and Liu, C. M., “Automated inspection of printed circuit boards through machine vision,” Computers in industry, Vol. 28, No. 2, pp. 103-111, May, 1996.
8. Ibrahim, Z., Al-Attas, S. A. R. and Aspar, Z., “Analysis of the　wavelet-based image difference algorithm for pcb inspection,” Proceedings of the 41st SICE Annual Conference, Vol. 4, pp. 2108 -2113, Aug. 2002.
9. 楊榮華，「應用灰階共變異矩陣之多重指標於瑕疵檢測」，元智大學工業工程與管理研究所碩士論文，2002.
10. Keys, R., “Cubic Convolution Interpolation for Digital Image Processing,” IEEE Transactions on Acoustics, Speech, and Signal Processing, Vol. 29, No. 6, pp. 1153 – 1160, Dec. 1981.
11. Reichenbach, S. E. and Geng, F., “Improved cubic convolution for two dimensional image reconstruction,” Nuclear Science Symposium Conference Record, Vol. 3, pp. 1775 – 1778, Nov. 2001.
12. Albiol, A., and Serra, J., “Morphological Image Enlargements,” Journal of Visual Communication and Image Representation, Vol.8, No. 4, pp. 367-383, 1997.
13. Chuah, C. S., and Leou, J. J., “An adaptive image interpolation algorithm for image/video processing,” Pattern Recognition, Vol. 34, pp. 2383-2393, 2001.
14. Chang, S.G., Cvetkovic, Z. and Vetterli, M., “Resolution enhancement of images using wavelet transform extrema extrapolation,” International Conference on Acoustics, Speech, and Signal Processing, Vol. 4, pp. 2379 – 2382, Detroit, MI USA, May, 1995.
15. 許惟勝，「多層次解析影像放大之研究」，國立台灣科技大學纖維及高分子工程研究所碩士論文，1999.
16. 謝東霖，「利用次像素技術於數位影像量測之研究」，國立清華大學動力機械研究所碩士論文，2003.
17. Nayar, S. K. and Nakagawa, Y. “Shape from focus: An effective approach for rough surfaces,” International Conference on Robotics and Automation, pp. 218-225, Cincinnati, OH USA, May, 1990.
18. Walsh D. and Raftery, A. E. “Accurate and efficient curve detection in images: the importance sampling Hough transform,” Pattern Recognition, Vol. 35, pp. 1421-1431, 2002.
19. Chen T. C. and Chung, K. L., “An efficient randomized algorithm for detecting circles,” Computer Vision and Image Understanding, Vol. 83, pp. 172-191, 2001.
20. 鍾國亮　教授著，「影像處理與電腦視覺」，東華書局，2002.
21. 吳成柯、戴善榮、程湘君與雲立實譯，「數位影像處理」，儒林出版社，2001.