Real Time Dynamic Threshold Clustering Method for Ancient Chinese Character Identification

Abstract:

Character image clustering can find similar characters which have much reference value to ancient character identification. But the classical clustering method depends much on the threshold. To alter the threshold dynamically and get result in real time, a improved clustering method is proposed. The classical BIRCH CF tree was amended to chain hierarchical structure, and the new tree was built from bottom to top with KNN clustering method. Based on this structure, relative similarity degree was transformed to character number in clustering result. By converting the dynamic threshold clustering problem to finding the different cluster range, this method could get the clustering result in real time.

• Periodical: Applied Mechanics and Materials (Volumes 241-244)
• Pages: 1728-1732
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.241-244.1728
• Online since: December 2012
• Authors: Fang Yang, Li Jiang, Jian Hui Si, Cai Feng Liu, Xue Dong Tian
• Keywords: Chain Hierarchical Structure, Character Identification, Dynamic Threshold, Real Time

References

[1] J.F. Guo, Y.Y. Zhao, W.F. Bian, J. Li: A hierarchical clustering algorithm based on improved cluster cohesion and separation. Journal of Computer Research and Development.Vol45(suppl.) (2008) 202-206.

[2] T. Zhang, R. Ramakrishnan., M. Livny: BIRCH: an efficient data clustering method for very large database. SIGMOD'96.Monteral,Canada,1996.New York: ACM Press(1996)103-114.

DOI: 10.1145/235968.233324

[3] J.T. Zou, F.X. Zhao, H.Y. Wang : Birch clustering algorithm based on the weighted similarity. Mathematics In Practice and Theory. Vol 41(2011) 118-124.

[4] F.J. Shao,B. Zhang, Z.Q. Yu: BIRCH clustering algorithm with mult-threshold. Computer Engineering and Applications. Vol40(2004) 174-176,195.

[5] Y.H. Zhu Y.Z. Jiang: Research of BIRCH clustering algorithm optimization and paralleism. Computer Engineering and Design.Vol28(2007) 4345-4347.

[6] S.Y. Jiang, X. Li: Improved BIRCH clustering algorithm. Journal of Computer Applications.Vol29(2009) 293-296.

[7] L. Jiang: Research on ancient Chinese character clustering based on BIRCH and improved k-medoids algorithm. Baoding: Hebei university(2012).

[8] L. Xu, B.H. Xiao, R.W. Dai, C.H. Wang: A fast classification strategy for large class sets. Journal of computer research and development.Vol45(2008)588-595.

[9] L. Jiang, F. Yang, X.F. Miao, Y. Ha, X.D. Tian: A retrieval method of ancient Chinese character images based on neighbor clustering. ICCSNT'2011, Harbin, China.Vol3(2011):1686-1689.

DOI: 10.1109/iccsnt.2011.6182291

[10] In: Kangxi Zidian(revised by Wang Yinzhi) (In Chinese ),eidted by Y.S. Zhang and J.T. Shanghai ancient books publishing house,1996.