Abstract: Objective To study the stability of pore features with AFIS fingerprint database. Methods By using of AFIS function of duplicate checking, the fingerprint images collected from the same person at different times are searched, and the distances and angles between pore feature and detail features in the collected fingerprint images are observed and measured. Besides，the statistics above are analyzed with the coefficient of variation and degree of deviation. Results The shape and size of the pore features vary greatly, the stability are poor, and the position of the pore features change less. The coefficient of variation of distance and angle between pore features and detail features is less than 8%, and most of the deviation is less than 9%. The position variation is in the acceptable range. Conclusion The physiological changes of pore features are far less obvious than the changes causing by different methods and pressure when collecting the fingerprint. The distribution of the pore features, which are measured referred to nearby the detail feature, remain stable at a certain physiological stage. The physiological stability of pore features provides a theoretical basis for fingerprint identification assisted by pore features.

Key words: automatic fingerprint identification system, pore feature, stability, coefficient of variation, degree of deviation

摘要： 目的 研究AFIS十指指纹数据库中汗孔特征的稳定性。方法 通过AFIS查重功能，检索同一人不同时间被采集的指纹图像，观察汗孔特征在20余年间位置、形态、大小的变化，测量指纹图像中汗孔特征与细节特征构成的距离与角度，用变异系数和偏离度进行统计分析。结果 汗孔特征的形态、大小变化大，稳定性差，汗孔特征的位置变化较小。汗孔特征与细节特征间距离与角度的变异系数小于8 ％，偏离度多数小于9 ％，其位置变化均在可接受范围内。结论 汗孔特征的生理变化远没有捺印压力、捺印方式变化引起的变化大。以细节特征为参考定点考察汗孔特征，其位置在一定生理阶段内保持稳定，汗孔位置的生理稳定性为利用汗孔特征辅助指纹同一认定提供了理论依据。

关键词: 指纹自动识别系统, 汗孔特征, 稳定性, 变异系数, 偏离度