摘要
指纹的旋转和平移是指纹匹配中必须解决的问题,针对这一问题提出了一种新的指纹局部特征描述方法,主要是利用特征点之间的距离、特征点类型来构建新的局部特征向量,利用这些局部结构特征来实现对匹配参考点的确定,将存在平移和旋转的两个指纹特征点集转换为以匹配参考点为极点的极坐标下进行全局匹配。实验结果表明该算法对于指纹图像的旋转和平移具有较强的适应性,可以满足实际应用的需要。
The problem of the translation and rotation of fingerprint must be solved when matching fingerprint,so we will introduce a new description method of fingerprint local characteristic.We have designed a new local characteristic vector including the distance between neighboring minutiae and the type of minutiae.Using the local characteristic vector we have ascertained the matching reference minutiae,We have transformed two fingerprint minutiae sets which have translation and rotation difference into the polar coordinates whose poles are the matching reference minutiae.Then we have done global matching.Experimental results indicate that our method adapt well to the translation and rotation of fingerprint and can meet the need of application.
出处
《计算机工程与应用》
CSCD
北大核心
2006年第5期74-76,共3页
Computer Engineering and Applications
基金
国家自然科学基金资助项目(编号:60173041)
湖南省自然科学基金资助项目(编号:02JJY2094)
关键词
指纹识别
点模式匹配
局部向量
多边彤
匹配参考点
identification of fingerprint,point pattern matching,local vector,polygon,matching reference minutiae
