本研究針對發光二極體晶粒之表面微影圖樣進行適應性模板建立並檢測其圖樣正確性之研究。由於產業界之自動化檢測為達到產線上高產能，在演算法系統開發上大都採以標準模板進行瑕疵比對以達高檢測速度，而在實際晶粒樣本比對過程中會產生擬合不佳狀況造成檢測系統誤判情況，本研究提出適應性模板方法改善系統擬合效果，使系統誤判情況降低達到高檢測效率。
首先，本研究導入一套晶粒完整度快速篩選流程，透過灰階熵值指標及相關係數指標進行晶粒完整性判斷，於適應性模板建立前預先將嚴重破壞之LED晶粒挑揀出，在快速篩選過程中平均每顆晶粒僅需0.052秒。不僅提升了適應性模板建立系統之可靠性亦提升整體系統之檢測速度。
此外，晶粒圖樣特徵點搜索部分本研究採用尺度不變特徵轉換及Harris-Laplace方法進行尺度空間中特徵點區域搜索及比較，以克服晶粒表面微影圖樣位移及尺度變化，並透過區域模板方法改善定位方式，加快定位速度並建立出適應性模板。本研究並導入接收者操作特徵曲線指標，判斷適應性模板檢測準確率。
準確率驗證部分，本研究以快速相關係數比對法與本研究提出之適應性模板比對法進行比較，結果顯示本研究方法檢測準確率為98.36%，較快速相關係數比對法相高出15.79%之準確性，在時間表現上，由快速篩選到適應性模板建立完成，過程僅需0.192秒即能完成。
最後，使用不同微影圖樣之發光二極體晶粒進行適應性之驗證，本研究提出之適應性模板可依不同微影圖樣建立其專屬模板，足以顯示本研究提出之方法有助於提升發光二極體產業之檢測效率與準確性，且提升該產業之市場競爭力。

This resaearch builds the adaptive template for the surface lithography pattern of light emitting diode (LED) chip and tests the pattern accuracy.
Firstly, a rapid screen process for chip integrity is used in this resaearch, the chip integrity is judged by gray scale entropy and correlation coefficient, the severely damaged LED chips are picked out before the adaptive template is built, so as to enhance the reliability of adaptive template positioning, each chip costs only 0.052 second on average in the rapid screen process. The reliability of adaptive template building method is enhanced, and the overall system inspection is accelerated.
In terms of searching for pattern feature points, this resaearch uses scale invariant feature transform (SIFT) and Harris-Laplace for local search of feature points in the scale space, so as to overcome the changes in the chip surface lithography pattern displacement and scale, and uses feature area pattern to improve the location mode, accelerate location and build the adaptive template. The receiver operating characteristic curve (ROC) is used in this resaearch to judge the inspection accuracy rate of adaptive template.
In terms of the validation of accuracy rate, the fast normalized cross correlation is compared with the adaptive template matching proposed in this resaearch. The result shows that the inspection accuracy rate of this method is 98.36%, higher than the fast normalized cross correlation by 15.79%. In terms of time performance, the process from rapid screen to the completion of adaptive template building only takes 0.192 second.
Finally, the LED chips of different lithography patterns are used to validate the adaptivity. The adaptive template proposed in this resaearch can build special templates for different lithography patterns, meaning the method proposed in this resaearch contributes to increasing the inspection efficiency and accuracy of LED industry, and to enhancing the marketability of the industry.

[1] 田民波、呂輝宗、溫坤禮，白光LED照明技術，五南圖書出版股份有限公司，2011。
[2] 中華民國，經濟部能源局，綠色能源產業資訊網，
http://www.taiwangreenenergy.org.tw/Domain/domain-2.aspx
[3] 柏德葳、蔡亦真、李秋緯，LED技術與應用與發展日趨多元化，TRI產業專題報告-136，拓墣科技股份有限公司，2009。
[4] 菊地正點、王政友，圖解半導體(改訂版)，世茂出版有限公司，2012。
[5] R. Adams, and L. Bischof, “Seeded region growing”, IEEE　Transactions on Pattern Analysis and Machine Intelligence, Vol. 16, No. 6, pp. 641-647, 1994.
[6] C. C. Kang, W. J. Wang, C. H. Kang, “Image segmentation with complicated background by using seeded region growing”, AEU -International Journal of Electronics and Communications, Vol. 66, No. 9, pp. 767-771, 2012.
[7] J. N. Kapur, P. K. Sahoo, and A. K. C. Wong, “A new method for gray-level picture thresholding using the entropy of the histogram”, Computer Vision, Graphics, and Image Processing, Vol. 29, pp. 273-285, 1985.
[8] L. Cao, Z. K. Shi, and E. K. W. Cheng, “Fast automatic multilevel thresholding method”, Electronics Letters, Vol. 38, No. 16, pp. 868-870, 2002.
[9] J. Y. Luo, and D. M. Tsai, “Mean shift-based defect detection in multicrystalline solar wafer surfaces”, IEEE Transactions on Industrial Informatics, Vol. 7, No. 1 , pp. 2206-2220, 2011.
[10] C. Y. Chang, Y. C. Chang, C. H. Li, and M. D. Jeng, “Automatic die inspection for post-sawing LED wafers”, Proceedings of the 2009 IEEE International Conference on Systems, San Antonio, USA, pp. 1615-1620, 2009.
[11] Q. Li , and S. Ren, “A visual detection system for rail surface defects”, IEEE Transactions on Sysetms Man and Cybernetics PART C- Applications and Reviews, Vol. 42, No. 6, pp. 1531-1542, 2012.
[12] P. D. Sathya, and R. Kayalvizhi, “Optimal multilevel thresholding using bacterial foraging algorithm”, Expert Systems with Applications, Vol. 38, No. 12, pp. 15549 - 15564, 2011.
[13] C. Wang, and Z. Ye, “Brightness preserving histogram equalization with entropy : a variatonal perspective”, IEEE Transactions on Consumer Electronics, Vol. 51, No. 4, pp.1326-1334, 2005.
[14] W. K. Pratt, Digital Picture Processing, Wiley-Interscience, U. S. A., 1978.
[15] J. P. Lewis, “Fast normalized cross-correlation”, Vision Interface, pp. 120-123, 1995.
[16] M. Sutton, “Application of an optimized digital correlation methord to palanar deformation analysis”, Image and Vision Computing, Vol. 4, No. 3, pp. 143-150, 1986.
[17] B. Kai, and D. H. Uwe, “Templete matching using fast normalized cross correlation”, Proceedings of SPIE-The International Society for Optical Engineering, Vol. 4387, pp. 95-102, 2001.
[18] D. M. Tsai, C. T. Lin, and J. F. Chen, “The evaluation of normalized cross correlations for defect detection”, Pattern Recognition Letters, Vol. 24, No. 15, pp. 2525-2535, 2003.
[19] C. C. Wang, B. C. Jiang, J. Y. Lin, and C. C. Chu, “Machine vision-based defect detection in IC images using the partial information correlation coefficient”, IEEE Transactions on Semiconductor Manufacturing, Vol. 26, No. 3, pp. 378-384, 2013.
[20] D. M. Gavrila, “A bayesian, exemplar-based approach to hierarchical shape matching”, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 29, No. 8, pp. 1408-1421, 2007.
[21] K. Zhang, L. Jiangbo, G. Lafruit, R. Lauwereins, and V. G. Luc, “Robust stereo matching with fast normalized cross-correlation over shape-adaptive regions”, Proceedings of 16th IEEE International Conference on Image Processing, Cairo, Egypt, pp. 2357-2360, 2009.
[22] D. M. Tsai, and C. H. Yang, “A quantile–quantile plot based pattern matching for defect detection”, Pattern Recognition Letters, Vol. 26, No. 13, pp. 1948-1962, 2005.
[23] S. Kaneko, Y. Satoh, and S. Igarashi, “Using selective correlation coefficient for robust image registration”, Pattern Recognition, Vol. 36, No. 5, pp. 1165-1173, 2003.
[24] D. M. Tsai, I. Y. Chiang, and Y. H. Tsai, “A shift-tolerant dissimilarity measure for surface defect detection”, IEEE Transactions on Industrial Informatics, Vol. 8, No. 1, pp. 128-137, 2012.
[25] C. Harris, and M. Stephens, “A combined corner and edge detector”, Proceedings of the 4th Alvey Vision Conference, United Kingdom, pp. 147-151, 1988.
[26] S. M. Smith, and M. Susan, “Susan-a new approach to low level image processing”, International Journal of Computer Vision, Vol. 59, No. 1, pp.45-78, 2004.
[27] V. A. Anisimov, and N. D. Gorsky, “Fast hierarchical oriented template matching of an arbitrarily”, Pattern Recognition Letters, Vol. 14, No. 2, pp. 95-103, 1993.
[28] D. G. Lowe, “Object recognition from local scale-invariant features”, in ICCV’99:Proceedings of the seventh IEEE International Conference on Computer Vision, Kerkyra, Greece, pp. 1150-1157, 1999.
[29] S. Se, D. G. Lowe, and J. J. Little, “Vision-based global localization and mapping for mobile robots”, IEEE Transactions on Robotics, Vol. 21, No. 3, pp.364-357, 2005.
[30] L. Juan, and O. Gwun, “A comparison of SIFT, PCA-SIFT and SURF”, International Journal of Image Processing, Vol. 3, No. 4, pp. 143-152, 2010.
[31] K. Mikolajczyk, and C. Schmid., “Indexing based on scale invariant interest points”, in ICCV’01:Proceedings of the Eighth IEEE International Conference on Computer Vision, Vol. 1, Vancouver B.C., Canada, pp. 525-531, 2001.
[32] G. Dorkó, and C. Schmid, “Selection of scale-invariant parts for object class recognition”, in ICCV’03:Proceedings of the Ninth IEEE International Conference on Computer Vision, Washington D.C., USA, Vol. 1, pp. 634-639, 2003.
[33] K. Mikolajczyk, and C. Schmid, “Scale & affine invariant interest point detector”, International Journal of Computer Vision, Vol. 60, No. 1, pp. 63-86, 2004.
[34] K. Mikolajczyk, and C. Schmid, “A performance evaluation of local descriptors”, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 27, No. 10, pp. 1615-1630, 2005.
[35] J. Zhang, Q. Chen, Q. Sun, H. Sun , and D. Xia, “A highly repeatable feature detector: improved Harris–Laplace”, Multimedia Tools and Applications, Vol. 52, No. 1, pp. 175-186, 2011.
[36] Z. Cao, Q. Yin, X. Tang, and J. Sun , “Face recognition with learning-based descriptor”, Proc. IEEE Computer Vision and Pattern Recognition, San Francisco, USA, pp. 2707 -2714, 2010.
[37] K. H. Kim, C. Zhang, Z. Zhang, and S. Choi, “Robust part-based face matching with multiple templates”, Proc. IEEE Automatic Face and Gesture Recognition, Shanghai, China, pp. 1-7, 2013.
[38] 繆紹剛，數位影像處理3/e，培升教育出版集團，2005。
[39] 新亞洲儀器技術團隊，機器視覺演算法與應用(初版)，新亞洲儀器，2011。
[40] 繆紹剛，數位影像處理：活用Matlab(第二版)，全華圖書股份有限公司，2011。
[41] T. Fawcett, “An introduction to ROC analysis”, Pattern Recognition Letters, Vol. 27, No. 8, pp. 861-874, 2006.
[42] P. Rober, P. Neuman, and R. Cavanagh，六標準差，美商麥格羅希爾國際出版，2001。
[43] P. Pande, and L. Holpp，樂為良 翻譯，六標準差簡單講，美商麥格羅希爾國際出版，2002。