A trend of similarity by changing base number of ICA in character recognition

被引：0

作者：

Ido S. ^{[1
]}

Yamaoka M. ^{[1
]}

机构：

[1] Graduate School of Science and Engineering, Ehime University, 3, Bunkyo-cho, Matsuyama, Ehime

来源：

| 1600年 / Institute of Electrical Engineers of Japan卷 / 136期

关键词：

Base number; Character recognition; FastICA; Image processing; Similarity;

D O I：

10.1541/ieejeiss.136.1106

中图分类号：

学科分类号：

摘要：

An independent component analysis (ICA) has received much attention in the fields of pattern recognition, such as face, fingerprint, figure, character, or number recognition. FastICA is an efficient and popular algorithm for ICA, and an improved algorithm is available for character recognition. In order to increase the rate of recognition, the appropriate setting of parameter or base number is indispensable. We investigate the effects of a change in base number on recognition rate to clarify the index for determining the number. As a result, we confirmed that recognition error rate tends to decrease in case the number is 3, 4, or 5. © 2016 The Institute of Electrical Engineers of Japan.

引用

页码：1106 / 1111

页数：5

共 4 条

[1] Ozawa S., Tsujimoto T., Kotani M., Baba N., Application of independent component analysis to handwritten Japanese character recognition, Proc of International Joint Conf. on Neural Networks (IJCNN99-Washington DC), CD-ROM, (1999)
[2] Hyvarinen A., Oja E., Independent Component Analysis: A tutorial, (1999)
[3] Bartlett M.S., Movellan J.R., Sejnowski T.J., Face recognition by independent component analysis, Neural Networks, IEEE Transactions, 13, 6, pp. 1450-1464, (2002)
[4] Hassaballah M., Kanazawa T., Ido S., Ido S., Efficient eye detection method based on grey intensity variance and independent components analysis, Computer Vision, IET, 4, 4, pp. 261-271, (2010)

← 1 →