一种新的基于相位保持的指纹图像增强算法  

Novel fingerprint image enhancement algorithm using phase preserving

作　　者：汪剑鸣[1] 沈虹[1] 丁润涛[2] 

机构地区：[1]天津工业大学信息与通信工程学院,天津300160 [2]天津大学电子信息工程学院,天津300072

出　　处：《计算机应用》2006年第1期140-142,共3页journal of Computer Applications

基　　金：天津市高等学校科技发展基金资助项目(20030502)

摘　　要：图像的相位信息在人眼感知图像的过程中起着十分重要的作用。如果指纹图像增强算法能够保证图像的相位信息不被破坏,将有助于保持指纹图像中纹理特征的不变。首先通过实验表明了指纹图像的纹理特征主要是由图像的相位信息来决定的,然后把相位保持原理引入到指纹图像增强算法中,提出一种新的基于相位保持的指纹图像增强算法。该算法采用复Gabor小波来提取指纹图像的相位信息,并利用指纹图像的局部方向信息在相位保持的原则下来进行图像增强。为了验证该算法的有效性,选取了两种低质量的指纹图像来进行处理,实验结果表明增强后的图像其图像质量有了显著的提高。Phase information of an image plays a very important role during perceiving image by human eyes, and it will help to keep fingerprint texture characteristics unchanged if phase corruption of image can be avoided in a fingerprint image enhancement algorithm. Firstly, an experiment was designed to show that a significant amount of the defining information of fingerprint image texture was contained in phase information. Then a novel fingerprint image enhancement algorithm based on phase preserving was proposed. In the algorithm, complex Gabor wavelets were adopted to extract phase information in fingerprint image, and local orientation information was utilized to enhance fingerprint image under the principle of phase preserving. To evaluate the performance of the algorithm, two kinds of low quality fingerprint images were enhanced. Experiment results show that the quality of enhanced images can be improved significantly.

关 键 词：指纹识别 图像增强 相位保持 GABOR小波 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]