摘要
根据指纹的特点,提出了一套有效的指纹识别预处理算法。包括前背景分离、基于空域和频域的图像增强、方向滤波、二值化、细化以及细化去噪处理。由于常用算法在指纹处理中存在缺陷,本文改进了增强算法和细化算法。实验证明,经过本算法处理后,减少了指纹图像中的伪结点和原有特征的丢失,有利于指纹特征提取。
This paper proposes a set of effective preprocessing algorithm in fingerprint recognition constituted with the features of fingerprint. Foreground /Background segmentation, fingerprint enhancement based on special domain and frequency domain, filtration in direction, binarization,thinning and noises elimination are all discussed.Because of the limitation of traditional methods in fingerprint image processing,we make improvement in enhancement and thinning.Experiments show that the algorithm can reduce pseudo crunodes and protect true minutias,it is helpful for the feature extraction.
出处
《微计算机信息》
北大核心
2007年第21期301-302,275,共3页
Control & Automation
关键词
指纹识别
图像预处理
图像增强
方向滤波
Fingerprint Recognition, Image Preprocessing, Fingerprint Enhancement, Directional Filter