基于主线特征的双向匹配的掌纹识别新方法

掌纹识别是利用人的手掌掌纹图像对其身份进行认证的一种生物特征识别技术，目前的掌纹研究主要集中在掌纹特征线的提取算法上，而对特征线的筛选和匹配的问题讨论较少，掌纹上的纹线比较复杂，深浅粗细长短不一，实施任何一种边缘提取算法都要考虑纹线的取舍问题，首先介绍了提出的应用最大内切圆对掌纹有效区域进行分割和对准的方法，较好地解决了掌纹的定位问题，然后提出了掌纹特征线族的概念，用以刻画掌纹上的主要特征，从而将掌纹纹线特征分为主要特征和次要特征．通过对主要特征与全部特征的双向匹配，给出最终的识别结果，将该方法与之前提出的基于傅里叶变换的方法在自行研制的掌纹采样设备所采集的掌纹库(90人450幅)上进行了比较实验，实验结果证明新方法可以处理原方法无法定位的掌纹图像，同时识别率也有明显提高。

A New Palmprint Identification Method Using Bi-Directional Matching Based on Major Line Features

In this paper, a novel online palmprint identification method is presented, which includes the inscribed circle based segmentation, line feature extraction and bi directional line feature matching method Unlike the traditional rectangle based segmentation approach, a circle based positioning and segmentation method is proposed, which can deal with more palmprint distortions The bi directional matching method is to classify the lines into major and minor features first Then the two maps are matched in two directions and the matching scores are averaged By doing the bi directional matching, small noises and distortions that are introduced by palmprint capturing and crease extraction are compromised The experimental results verified the effectiveness of the proposed identification method