傅里叶变换在指纹图像增强中的应用  

Fourier transform applied in fingerprint image enhancement

作　　者：谢新辉[1] 唐立军[1] 宋海吒[1] 唐俊龙[1] 

机构地区：[1]长沙理工大学,长沙410114

出　　处：《计算机工程与应用》2012年第8期197-199,共3页Computer Engineering and Applications

基　　金：湖南省自然科学基金(No.07JJ3112);长沙市科技局科技项目(No.K0803081-11);湖南省教育厅科研基金资助项目(No.08C111);长沙理工大学重点学科建设项目

摘　　要：提出一种基于傅里叶变换的指纹图像增强方法,通过对图像分块进行傅里叶变换,用求期望值的方法获取指纹图像的频率、方向以及能量分布,在此基础上根据指纹纹线的分布在脊线方向变化剧烈的特殊性,采用在径向滤波器滤波后增加角方向滤波器的滤波方法,对图像进行滤波增强。实验结果表明,该方法对低质量的指纹图像增强效果显著,同时其计算量较少,实现时间短,对实时指纹识别系统的应用具有重要的意义。A new method based on Fourier transform is applied in fingerprint image enhancement. With the evaluation of expectations, the original image is segmented into blocks to calculate Fourier transform from which its frequency, direction and energy distribution can be obtained simultaneously. Considering the directional particularity of the fingerprint ridges, both a radial filter and an angular ill-ter are utilized for enhancement. Experimental results show that this method can enhance the low quality fingerprint images obviously with less computation workload, so it has significance for the application of real-time fingerprint identification system.

关 键 词：傅里叶变换 径向滤波 角方向滤波 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]