Applying FFT features for fingerprint matching

被引：0

作者：

Tachaphetpiboon, S. ^{[1
]}

Amornraksa, T. ^{[2
]}

机构：

[1] Phetchaburi Rajabhat Univ, Dept Comp Sci, 38 Hadchaosumran Rd, Muang 76000, Phetchaburi, Thailand

[2] King Mongkut Univ Technol Thonburi, Dept Comp Engn, Bangkok 10140, Thailand

来源：

INTERNATIONAL SYMPOSIUM ON WIRELESS PERVASIVE COMPUTING 2006, CONFERENCE PROGRAM | 2006年

关键词：

signal processing; image processing; fingerprint matching; FFT;

D O I：

暂无

中图分类号：

TP301 [理论、方法];

学科分类号：

081202 ;

摘要：

In this paper, a feature extraction method based on the fast Fourier transform (FFT) is proposed for the use of fingerprint matching. The performance of our approach is evaluated by the amount of time required in the features extracting and feature matching processes. The recognition rate obtained from the proposed method is also evaluated by the kNN classifier. The experimental results show that, by using our approach, the processing time in the features extraction and searching & matching processes is reduced by up to 34.95 times, and decreased by up to 0.39 milliseconds, respectively, compared to the method proposed in [6]. Furthermore, with our proposed method, the fingerprint recognition rate can be improved by 1.1% to 13.19%.

引用

页码：366 / +

页数：2