计算机技术与发展2018，Vol.28Issue(5)：164-167,173,5.DOI:10.3969/j.issn.1673-629X.2018.05.037
基于Curvelet变换的指纹图像去噪
Fingerprint Image Denoising Based on Curvelet Transform
摘要
Abstract
Fingerprint image plays an important role in biometrics because of its lifetime invariance,uniqueness and convenience.Howev-er,in the collection and transmission the fingerprint will inevitably be polluted by the noise from outside,thus affecting the accuracy of fingerprint identification system.In this paper,we use the multi-scale geometric denoising method based on Curvelet transform and pro-pose the adaptive threshold selection for each Curvelet sub-band after multi-scale transformation.Then we adopt an improved threshold function to overcome the shortcomings of the traditional soft and hard threshold function.The experiment shows that the proposed method can improve the edge,straight line and curve features in the image,and the PSNR after denoising is higher,with better effect than tradi-tional threshold processing method.关键词
Curvelet变换/阈值函数/Wrapping算法/自适应阈值/阈值去噪Key words
Curvelet transform/threshold function/Wrapping algorithm/adaptive threshold/threshold denoising分类
信息技术与安全科学引用本文复制引用
张建明,邱晓晖..基于Curvelet变换的指纹图像去噪[J].计算机技术与发展,2018,28(5):164-167,173,5.基金项目
江苏省自然科学基金(BK2011789) （BK2011789）
东南大学毫米波国家重点实验室开放课题(K201318) （K201318）
