基于DWT的二次特征提取脱机中文签名鉴定

山东大学学报(工学版)

基于DWT的二次特征提取脱机中文签名鉴定

田伟, 乔谊正, 马志强

山东大学控制科学与工程学院,山东济南250061

收稿日期:2007-01-16 修回日期:1900-01-01 出版日期:2007-06-24 发布日期:2007-06-24
通讯作者: 田伟

Offline Chinese signature verification based on the second feature extraction by DWT

TIAN Wei,QIAO Yi-zheng,MA Zhi-qiang

School of Control Science and Engineering, Shandong University, Jinan Shandong 250061, China

Received:2007-01-16 Revised:1900-01-01 Online:2007-06-24 Published:2007-06-24
Contact: TIAN Wei

摘要/Abstract

摘要： 提出了一种基于投影特征和离散小波变换（DWT）的脱机中文签名鉴定的新方案.该方案利用小波分解的高频系数具有较好时间分辨率的特点，对调整后的投影特征矢量选用各层高频系数进行单支重构提取二次特征，此二次特征在增强了签名者的个人特性的同时也增加了签名特征隶属于真签名的不确定性，由此引入模糊网络来表征这种不确定性进行鉴定.实验中对中文签名数据库取得了大约11%的平均错误率，较现有文献的实验结果减少了大约6%.

关键词: 脱机签名鉴定, 投影特征, 离散小波变换, 模糊网络

Abstract: A novel approach to offline Chinese signature verification is proposed based on projection profiles and discrete wavelet transform (DWT). Since the high frequency coefficients of wavelet decomposition represent the better timeresolution features, the reconstructed signals are the second features of the adjusted projection profiles. These features can enhance the signature characteristics, while with an increase of uncertainty to the true signatures. Fuzzy nets are introduced to describe the variation and to employ verification. The average error rate is about 11% obtained from a Chinese signatures database, which is about 6% less than that of the existing research.

Key words: projection profiles, DWT, fuzzy net , off-line signature verification

中图分类号:

TP391.43

田伟,乔谊正,马志强 . 基于DWT的二次特征提取脱机中文签名鉴定[J]. 山东大学学报(工学版), 2007, 37(3): 55-59 .

TIAN Wei,QIAO Yi-zheng,MA Zhi-qiang . Offline Chinese signature verification based on the second feature extraction by DWT[J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2007, 37(3): 55-59 .

参考文献

多维度评价

Viewed

Full text

Abstract

Cited

Shared

Discussed

[1]	赵彦霞, 王熙照. 基于SVD和DCNN的彩色图像多功能零水印算法[J]. 山东大学学报(工学版), 2018, 48(3): 25-33.
[2]	李岳阳,王士同. 基于鲁棒性神经模糊网络的脉冲噪声滤波算法[J]. 山东大学学报(工学版), 2010, 40(5): 164-170.