一种滤除细化指纹图像中伪特征点的新方法  

New Approach to Eliminate False Minutiae in Thinned Fingerprint Image

作　　者：刘灵丽[1] 李丽娟[1] 

机构地区：[1]湖南大学计算机与通信学院

出　　处：《计算机应用研究》2006年第11期239-240,243,共3页Application Research of Computers

基　　金：国家自然科学基金资助项目(60373062);教育部科研重点资助项目(03092);湖南省杰出中青年基金资助项目(02JJYB012)

摘　　要：提出了一种新的滤除指纹细节伪特征方法,它首先对非边界端点采用脊线跟踪法滤除毛刺和短脊产生的伪特征点,然后用窗口法滤除断脊、假桥、小孔和岛屿产生的伪特征点。结果表明,该方法简单、处理速度快,滤除精度基本可以满足应用的需要。A new approach to eliminate the false minutiae in the thinned fingerprint image is presented, First, the false minutiae of spike or spur are eliminated by following the ridge-line from non-boundary endings. Second, the left false minutiae of ridge disconnection, bridge, hole or island are eliminated by a neighborhood. It is found that the approach is simple and the speed is high, Moreover, the accuracy of minutiae elimination can fill the demand of application.

关 键 词：细化指纹图像 细节特征 脊线跟踪 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]