由於現今半導體製造過程越來越複雜，其相對成本亦越來越高，因此提升產品良率更為重要。工程師可憑藉故障晶圓圖樣追溯發生異常原因的線索，例如異常的製程步驟或是機台發生問題等，進而提升產品的良率並降低成本。以往半導體廠對於故障晶圓圖的判斷，大多以人工目視的方式來進行，但由於人為主觀因素以及對空間圖形辨識能力的差距，經常造成圖形辨識結果並不一致，甚至因此影響問題解決的效率。因此，故障晶圓圖的檢測已成為現代半導體製造的關鍵問題。另一方面，隨著晶圓尺寸及晶片數目的擴大，晶圓圖的維度因此提高，使得故障晶圓圖樣較以往有更多特徵上的變化，例如:圖樣大小、密度、位置與旋轉等。導致傳統的圖形辨識或分類方法較難以準確捕捉到每個維度的變化 。此外，由於晶圓圖樣變化較以往更多，故對於新發現的特殊圖樣，與其相似的晶圓圖數量較為稀少。因此，本研究提出了一種新的概念，以形態學為基礎結合支持向量機(Morphology-based Support Vector Machine, MSVM)的方法解決相似晶圓圖樣自動搜尋的問題，以提升搜尋出同時包含許多特徵變化之相似故障圖樣的效率。利用傳統形態學的概念，例如侵蝕(erosion)、膨脹(dilation)、閉合(closing)與斷開(opening)等，產生許多相似的故障圖樣做為訓練樣本，並依據合作半導體廠的工程師建議，進而提出使晶圓圖樣有密度(density)、位置(shift)與旋轉(rotation)等特徵變化的方法，解決晶圓圖樣本稀少的問題。並利用支持向量機(SVM)方法來做相似度搜尋。本研究利用合作半導體廠的實際資料來驗證方法可行性以及執行效率，並與半導體廠的現行方法進行比較，MSVM確實能夠提升搜尋的準確率，且執行的運算時間也相對較低。本研究成果不僅解決傳統圖形辨識或分類方法無法同時考慮圖樣有多種特徵變化的困難，並提升晶圓圖辨別的一致性以及事故診斷的效率。

Due to the ever greater complexity of processes involved in semiconductor manufacturing, increasingly high inspection costs associated with defective wafers have become a critical concern of modern manufacturers. More importantly, because current high-dimensional wafer bin maps (WBMs) cause many variations in features, it is difficult to capture the variations of each dimension via traditional pattern recognition or classification methods. Therefore, this work proposes a novel similarity searching tool, a morphology-based support vector machine (MSVM) designed for defective wafer detection. Seven kinds of morphology-based training sample generations are presented; the morphological method includes original morphology definitions in addition to our proposed features. The MSVM can categorize practical industrial datasets according to variant degrees of similarities. The experimental results demonstrate the usefulness of our approach in the context of yield improvements in precision, low errors and acceptable computation cost.

1. Quirk˙Serda著，羅文雄、蔡榮輝、鄭柚盈譯，2003，「半導體製造技術」，滄海書局。
2. Tobin, K. W., Gleason, S. S., Lakhani, F., & Bennett, M. H. (1997). Automated analysis for rapid defect sourcing and yield learning. Future Fab International, 1(4), pp. 313. London: Technology Publishing Ltd.
3. Chien, C., Lee, P., Peng, C., 2003. Semiconductor manufacturing data mining for clustering and feature extraction. Journal of Information Management 10 (1), 63–84.
4. Chien, C., Wang, W., Cheng, J., 2007. Data mining for yield enhancement in semiconductor manufacturing and an empirical study. Expert Systems with Applications 33 (1), 1–7.
5. 簡禎富、林鼎浩與劉巧雯，「建構晶圓圖分類之資料挖礦方法及其實證研究」，Journal of the Chinese Institute of Industrial Engineers, Vol. 19, No. 2, pp. 23-38 (2002).
6. Stapper, C. H. (1985). The effects of wafer to wafer defect density variations on integrated circuit defect and fault distributions. IBM Journal of Research Development, 29(1), pp. 87–97.
7. Hsu S.-C., Chien C.-F., (2007). Hybrid data mining approach for pattern extraction from wafer bin map to improve yield in semiconductor manufacturing. International Journal of Production Economics, 107, pp. 88-103.
8. Li T.-S., Huang C.-L., (2009). Defect spatial pattern recognition using a hybrid SOM–SVM approach in semiconductor manufacturing. Expert Systems with Applications, 36, pp. 374-385.
9. Chao, L.-C, Tong L.-I., (2009). Wafer defect pattern recognition by multi-class support vector machines by using a novel defect cluster index. Expert Systems with Applications 36, pp. 10158–10167.
10. Friedman, D. J., Hansen, M. H., Nair, V. N., & James, D. A. (1997). Model-free estimation of defect clustering in integrated circuit fabrication. IEEE Transactions on Semiconductor Manufacturing, 10(3), 344–359.
11. Nieddu, L., & Patrizi, G. (2000). Formal methods in pattern recognition. European Journal of Operation Research, 120(3), pp. 459–495.
12. Bhanu, B., Lee, S., & Ming, J. (1995). Adaptive image segmentation using a genetic algorithm. IEEE Transactions on Systems Man Cybernetics, 25(12), pp. 1543–1567.
13. Chen, C. L., & Chang, M. H. (1998). Optimal design of fuzzy sliding-mode control: A comparative study. Fuzzy Sets and Systems, 93(1), pp. 37–48.
14. Jain, A. K., Duin, R. P. W., & Mao, J. (2000). Statistical pattern recognition: A review. IEEE Transactions on Pattern Analysis and Machine Intelligence, 22(1), pp. 4–37.
15. Fiesler, E. (1994). Comparative bibliography of ontogenic neural networks. In Proceedings of the international conference on artificial neural networks. Sorrento, Italy, pp. 793–796.
16. Cortes, C., & Vapnik, V. (1995). Support vector networks. Machine Learning, 20(3), pp. 273–297.
17. Hsu, C. W., & Lin, C. J. (2002). A comparison of methods for multi-class support vector machines. IEEE Transactions on Neural Networks, 13(2), pp. 415–425
18. Joachims, J. (1998). Text categorization with support vector machines: Learning with many relevant features. In Proceedings of ECML-98, 10th European conference on machine learning, pp. 137–142.
19. V. N. Vapnik, The Nature of Statistical Learning Theory. New York: Springer-Verlag, 1995.
20. Perdisci, R., Gu, G., Lee, W., (2006). Using an ensemble of one-class SVM classifiers to harden payload-based anomaly detection systems. 6th IEEE International Conference on Data Mining Location: Hong Kong, pp. 488-498.
21. Manevitz, L.M., Yousef, M., (2001). One-class SVMs for document classification. Journal of Machine Learning Research 2, pp. 139-154.
22. Georges, M. and Jean, S., History of Mathematical Morphology, 1968.
23. Chih-Chung Chang and Chih-Jen Lin, LIBSVM : a library for support vector machines. ACM Transactions on Intelligent Systems and Technology, 2:27:1--27:27, 2011. Software available at
http://www.csie.ntu.edu.tw/~cjlin/libsvm .
24. J. Principe, N. Euliano, and W. Lefebvre, Neural and Adaptive Systems: Fundamentals Through Simulations, John Wiley & Sons, Inc., 2000.
25. R.R. Yager, D.P. Filev, Approximate clustering via the mountain method, IEEE Trans. Systems Man Cybernet. 24 (1994) 1279–1284.