臺灣博碩士論文加值系統

廣告充斥在我們的生活中，我們每天都在接觸各式各樣的廣告。自從人類有市場交易制度以來，廣告就在人類的商業活動中扮演重要的角色。因此，透過研究廣告，我們可以瞭解一些關於歷史與社會科學，甚至是一些關於商業策略的資訊。
在本篇論文中，我們提出一個對報紙影像及網站快照做廣告偵測，切割及分類的架構。首先，我們利用連通元件法(connected components)從影像中擷取廣告候選區塊。接者，我們設計了兩個過濾器過濾非廣告的候選區塊。首先是利用規則法過濾掉明顯不是廣告的候選區塊。在規則法過濾之後，我們提出利用機器學習的方法進一步的分辨出非廣告的候選區塊並加以濾除。對於剩下來的廣告候選區塊，我們利用機器學習透過視覺特徵將他們分類到各個預先定義好的類別。我們統計各種廣告類別在報紙頭版以及網站快照中的分布，發現了一些跟廣告與社會環境相關的有趣資訊。

There are many advertisements in our life, and we read various advertisements every day. Advertisement plays an important role in human commercial activity since human constructed the trade market system. Therefore, with the studies of advertisements, we can realize some issues about historical science and social science, or even the commercial strategy.
In this thesis, we propose an advertisement detection, segmentation and classification framework. First, we extract advertisement candidates based on a connected components method. And then, we design two filters to remove candidates that are not advertisement. The first is a rule-based filter. It filters out the candidates that are obviously not advertisements based on conventions and rules. After the rule-based method, a learning-based filter is designed to further remove the non-advertisement candidates. We classify the remained advertisement candidates into predefine categories using visual features. Finally, we collect the statistics of advertisements in newspaper front pages and website snapshots, respectively, and discover several interesting characteristics.

誌謝 i
摘要 ii
ABSTRACT iii
LIST OF FIGURES v
LIST OF TABLES vi
Chapter 1 INTRODUCTION 1
1.1 Motivation 1
1.2 Contributions 2
1.3 Thesis Organization 3
Chapter 2 RELATED WORKS 4
2.1 Layout Segmentation 4
2.2 Advertisement Classification 6
Chapter 3 ADVERTISEMENT DETECTION, SEGMENTATION, AND CLASSIFICATION 7
3.1 System Overview 7
3.2 Databases 8
3.3 Data Processing 10
3.4 Classification in Heterogeneous Images 21
Chapter 4 STATISTIC RESULTS AND DISCUSSION 23
4.1 Statistics of Newspaper Advertisements 23
4.2 Statistics of Website Advertisements 27
Chapter 5 CONCLUSION AND FUTURE WORKS 31
5.1 Conclusion 31
5.2 Future Work 32
REFERENCES 33
APPENDIX 36

[1] S. Moriarty, N. D. Mitchell, and W. D. Wells, “Advertising & IMC: Principles and Practice,” Pearson, 2012.
[2] A. Vedaldi and K. Lenc, “MatConvNet: Convolutional Neural Networks for Matlab,” Proceedings of ACM International Conference on Multimedia, 2015.
[3] B. Gatos, S. L. Mantzaris, K. V. Chandrinos, A. Tsigris, and S. J. Perantonis, “Integrated algorithms for newspaper page decomposition and article tracking,” In Proceedings of International Conference on Document Analysis and Recognition, 1999.
[4] F. Liu, Y. Luo, M. Yoshikawa, and D. Hu, “A new component based algorithm for newspaper layout analysis,” In Proceedings of International Conference on Document Analysis and Recognition, pp. 1176-1180, 2001.
[5] P. E. Mitchell and H. Yan, “Newspaper document analysis featuring connected line segmentation,” In Proceedings of International Conference on Document Analysis and Recognition, pp. 1181-1185, 2001.
[6] P. E. Mitchell and H. Yan, “Newspaper layout analysis incorporating connected component separation,” Image and Vision Computing, vol. 22, no. 4, pp. 307-317, 2004.
[7] P. E. Mitchell and H. Yan, “Connected pattern segmentation and title grouping in newspaper images,” In Proceedings of International Conference on Pattern Recognition, vol. 1, pp. 397-400, 2004.
[8] K. Hadjar, O. Hitz, and R. Ingold, “Newspaper page decomposition using a split and merge approach,” In Proceedings of International Conference on Document Analysis and Recognition, pp. 1186-1189, 2001.
[9] B. Gatos, S. L. Mantzaris, and A. Antonacopoulos, “First international newspaper segmentation contest,” In Proceedings of International Conference on Document Analysis and Recognition, pp. 1190-1194, 2001.
[10] A. Bansal, S. Chaudhury, S. D. Roy, and J. B. Srivastava, “Newspaper article extraction using hierarchical fixed point model,” In Proceedings of IAPR Workshop on Document Analysis Systems, 2014.
[11] Q. Li, J. Wang, D. Wipf, and Z. Tu, “Fixed-point model for structured labeling,” In Proceedings of International Conference on Machine Learning, vol. 28, pp. 214-221, 2013.
[12] K. Hadjar and R. Ingold, “Arabic newspaper page segmentation,” In Proceedings of International Conference on Document Analysis and Recognition, pp. 895-899, 2003.
[13] R. Garg and A. Bansal, “Text graphic separation in Indian newspapers,” Proceedings of International Workshop on Multilingual OCR, Article no. 13, 2013.
[14] Z. Hua, X.-J. Wang, Q. Liu, H. Lu, “Semantic knowledge extraction and annotation for web images,” In Proceedings of ACM International Conference on Multimedia, pp. 467-470, 2005.
[15] D. Chakrabarti, R. Kumar, and K. Punera, “A graph-theoretic approach to webpage segmentation,” In Proceedings of International Conference on World Wide Web, pp. 377-386, 2008.
[16] F. Fauzi, J.-Lang Hong, and M. Belkhatir, “Webpage segmentation for extracting images and their surrounding contextual information,” In Proceedings of ACM International Conference on Multimedia, pp. 649-652, 2009.
[17] D. Cai, S. Yu, J.-R. Wen, and W.-Y. Ma, “Vips: a vision based page segmentation algorithm,” Technical Report MSR-TR-2003-79, Microsoft Research, 2003.
[18] R. Song, H. Liu, J.-R. Wen, and W.-Y. Ma, “Learning block importance models for web pages,” In Proceedings of International Conference on World Wide Web, pp. 203-211, 2004.
[19] O. Wu, Y. Chen, B. Li, and W. Hu, “Evaluating the visual quality of web pages using a computational aesthetic approach,” In Proceedings of ACM International Conference on Web Search and Data Mining, pp. 337-346, 2011.
[20] R. A. Peleato, J.-C. Chappelier, and M. Rajman, “Using information extraction to classify newspapers advertisements,” In Proceedings of International Conference on the Statistical Analysis of Textual Data, 2000.
[21] X. Yin and W. S. Lee, “Understanding the function of web elements for mobile content delivery using random walk models,” In Proceedings of International Conference on World Wide Web, pp. 1150-1151, 2005.
[22] K. Chatfield, K. Simonyan, A. Vedaldi, and A. Zisserman, “Return of the Devil in the Details: Delving Deep into Convolutional Nets,” Proceedings of British Machine Vision Conference, 2014.
[23] C. C. Chang and C. J. Lin, “Libsvm: A Library for Support Vector Machines,” ACM Transactions on Intelligent Systems and Technology, pp. 1–27, 2011.
[24] http://www.dgbas.gov.tw/ct.asp?xItem=28854&ctNode=3111
[25] Y. G. Jiang, J. Yang, C. W. Ngo, and A. G. Hauptmann, “Representations of Keypoint-Based Semantic Concept Detection: A Comprehensive Study,” IEEE Transactions on Multimedia, vol. 12, issue 1, pp. 42-53, 2010.
[26] J. RR. Uijlings, K. E. A. Ven de Sande, T. Gevers, A. W. M. Smeulders “Selective search for object recognition,” International Journal of Computer Vision, pp. 154-171, 2013.