基于结构的指纹特征抽取方法

为了弥补传统的提取细节特征点的方法在小面积指纹图像识别上的不足，针对指纹脊线纹理结构进行了详细研究，分析了Gabor滤波器应用在纹理上的原理，建立了用于指纹识别的Gabor滤波器的物理模型，在此基础上设计了基于Gabor滤波器获取指纹特征的算法.首先提出了一种快速和简便的自动定位指纹图像中心点方法，以中心点为参考点在其周围确定特征提取区，根据Gabor滤波器的性质对其进行8方向滤波提取特征.通过在标准指纹数据库上的对比实验，显示了该方法的有效性.

Method on Structure-Based Fingerprint Feature Extraction

The algorithm of fingerprint feature extraction using Gabor filterbank was proposed in order to overcome the drawback of traditional minutiae feature extraction.Through studying texture structure information of fingerprint ridges and applying Gabor fiherbank to texture,a physical model of Gabor filtering was established.A quick and simple method of center point searching was presented,according to center point's position round grid feature area was loca- ted,filtering by eight directions of Gabor filterbank and feature extraction.The result of com- parison experiment showed validity of Gabor-based.