全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...

手写字符识别的非线性主动判别函数

DOI: 10.11834/jig.20081009

Keywords: 字符识别非线性主动判别函数Kernel,PCA

Full-Text   Cite this paper   Add to My Lib

Abstract:

提出了用于手写字符识别的非线性主动判别函数,是线性主动判别函数在手写字符非线性变化情况下的推广。该方法利用KernelPCA分析捕捉和表示这种非线性变化。将输入空间非线性映射为特征空间,在特征空间的主子空间中生成最优主动原型模板,其与字符特征向量在特征空间主子空间的投影之间的距离即为非线性主动判别函数;同时,基于最小分类错误准则对该函数进行了优化。实验结果表明,非线性主动判别函数获得了比线性主动判别函数更高的识别率。

References

[1]  Jain A K , Zongker D. Representation and recognition of handwritten digits using deformable templates [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1997, 19 (12) : 1386 - 1391.
[2]  Seiichi Uchida, Hiroaki Sakoe. Eigen-deformations for Elastic Matching based Handwritten Character Recognition [J]. Pattern Recognition, 2003, 36(9) :2031 - 2040.
[3]  Shi Da-ming, Gunn S R, Damper R I. A radical approach to handwritten chinese character recognition using active handwriting models [ A]. In: Proceedings IEEE Conference on Computer Vision and Pattern Recognition[C]. Kauai, Hawaii, USA,2001:670-675.
[4]  Shi Da-ming, Gunn S R, Damper R I. Handwritten chinese radical recognition using nonlinear active shape models [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2003, 25(2): 277-280
[5]  Sun Guang-ling, Tang Xiang-long, Liu Jia-feng. Active discriminant functions for handwriting recognition[J]. Journal of Software,2005, 16 (4) :166-176.
[6]  Scholkopf B, Smola A J, Muller K. Kernel principal component analysis [A]. In: Scholkopf B, Burges C J C, Smola AJ (Eds.), Advances in Kernel Methods[C], Cambridge, MA, USA : MIT Press, 1998. 327 - 352.
[7]  孙广玲 唐降龙.基于梯度归一化的模糊梯度特征提取方法[J].哈尔滨工业大学学报,2006,38(12):2030-2035.

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133