由於使用機器學習技術在佈局上檢測到的潛在熱點圖像數量過多，因此設計人員
需要花費大量時間來對這些潛在熱點圖像進行分類以利後續修正。 這些熱點圖像
相當多樣並且圖像上的形狀也很複雜。 因此，我們提出了一種基於密度的熱點
圖像分類法，對這些熱點圖像分類，在同時考慮熱點圖像上多邊形的偏移和扭曲
下，抽取熱點圖像的重要圖形特徵。 實驗結果顯示，我們的方法我們的方法可以
比 SIFT 方法更有效地對熱點圖像進行分類，並且分組中的結果相似。

Since the number of hotspot patterns detected on a layout using machine learning technique is very large, it takes designers a lot of time to classify these hotspot patterns for subsequent modification. These hotspot patterns are diverse and complex in shape. Therefore, we propose a density-based hotspot pattern clustering approach to classify these hotspot patterns into groups, which extracts the density feature of hotspot patterns while considering the shifted and distorted polygons on hotspot patterns. Experimental results show that our approach can classify the hotspot patterns more efficiently than SIFT method with similar results in each group.

中文摘要 i
Abstract ii
誌謝辭 iii
Contents iv
List of Tables vi
List of Figures vii
1 Introduction 1
2 Background 4
2.1 Hotspot Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 Density of a Grid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.3 K-means Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.4 Center of Gravity of a Polygon . . . . . . . . . . . . . . . . . . . . . 6
3 Proposed Framework 9
3.1 Density Feature Vector of a Hotspot Pattern . . . . . . . . . . . . . . 9
3.2 Rotational Hotspot Pattern Clustering . . . . . . . . . . . . . . . . . 13
3.3 Hotspot Pattern Clustering . . . . . . . . . . . . . . . . . . . . . . . 15
3.4 Overall Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
4 Experimental Results 17
5 Conclusion 20

[1] Bradski, Gary. “The openCV library.” Dr. Dobb’s Journal: Software Tools for the Professional Programmer 25.11 (2000): 120-123.
[2] Buitinck et al., “API design for machine learning software: experiences from the scikit-learn project,” ECML PKDD Workshop: Languages for Data Mining and Machine Learning, 2013.
[3] Chang, Wei-Chun, et al.,“iClaire: A Fast and General Layout Pattern Classification Algorithm” Proceedings of the 54th Annual Design Automation Conference 2017, 2017.
[4] Jingsong Chen, James Shiely, Evangeline F. Y. Young, “Fast detection of largest repeating layout pattern,” Design-Process-Technology Co-optimization for Manufacturability XIII., vol. 10962. SPIE, 2019.
[5] Charles Elkan, “Using the Triangle Inequality to Accelerate K-Means,” ICML’03: Proceedings of the Twentieth International Conference on International Conference on Machine Learning, pp. 147-153, 2003.
[6] Tianyang Gai, Tong Qu, Xiaojing Su, Shuhan Wang, Lisong Dong, Libin Zhang, Rui Chen, Yajuan Su, Yayi Wei, Tianchun Ye, “Multi-level layout hotspot detection based on multi-classification with deep learning,” Proc. SPIE 11614, Design-Process-Technology Co-optimization XV, 116140W (22 February 2021); https://doi.org/10.1117/12.2583726.
[7] Gonzalez, R.C. and Woods, R.E., Digital Image Processing. Pearson Education. ISBN:9781292223070
[8] J. Jiang et al., “Reducing Systematic Defects using Calibre Wafer Defect Engineering and Machine Learning Solutions,” 2020 International Workshop on Advanced Patterning Solutions (IWAPS), 2020, pp. 1-3, doi:
10.1109/IWAPS51164.2020.9286791.
[9] D. G. Lowe, “Object recognition from local scale-invariant features,” in Proceedings of the Seventh IEEE International Conference on Computer Vision,
1999, pp. 1150-1157 vol.2, doi: 10.1109/ICCV.1999.790410.
[10] Yuansheng Ma, Feng Wang, Qian Xie, Le Hong, Joerg Mellmann, Yuyang Sun, Shao Wen Gao, Sonal Singh, Panneerselvam Venkatachalam, James Word, “Machine learning based wafer defect detection,” Proc. SPIE 10962, DesignProcess-Technology Co-optimization for Manufacturability XIII, 1096208 (20
March 2019)); https://doi.org/10.1117/12.2513232.
[11] Pedregosa et al., “Scikit-learn: Machine Learning in Python,” Journal of Machine Learning Research, vol. 12, pp. 2825-2830, 2011.
[12] H. -C. Shao et al., “From IC Layout to Die Photograph: A CNN-Based DataDriven Approach,” in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, vol. 40, no. 5, pp. 957-970, May 2021.
[13] W. Wen, J. Li, S. Lin, J. Chen and S. Chang, “A fuzzy-matching model with grid reduction for lithography hotspot detection.,” in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, vol. 33, no. 11, pp.
1671-1680, Nov. 2014.
[14] An In-house tool in a semiconductor manufacturing company.