伪造指印识别方法研究进展

Review into Progressing Researches on Recognition of Forged Fingerprints

随着近年来伪造指纹或声称指纹/指印被伪造的案件不断增多,相应检验方法的缺失已成为指纹检验领域的痛点之一。本文首先阐述了伪造指印识别的研究背景,分别按照伪造方式和是否可见两个标准归纳了伪造指印的分类,然后重点对伪造指印识别方法(主要是形态学比对方法和机器学习方法)的研究进展进行了综述,最后结合形态学比对方法的研究依据及机器学习方法的发展趋势,对伪造指印识别方法研究的发展趋势做了简要展望。

Fingerprint forgery has been becoming easier when the available materials can be used to produce fi ngermarks onto any matrix of interest.However,there are relatively rare cases about fingerprint forgery though sparse reports were mentioned it even at the start of 20th century,resulting in few practical re cognition methods for law enforcement practice to identify forged fingerprints presently.Nevertheless,the lack of corresponding examination methods would cause embarrassment for fi ngerprint detection as the increasing numbers of cases are recently looming about forged fi ngerprints or claimed fi ngerprint forgeries.Therefore,this article tries to focus on the issue of forged fi ngerprint recogni tion,beginning with illustrating the relevant research background and successively summarizing the classifi cation of forged fi ngerprints according to two different criteria,i.e.,whether the forgery techniques require the donor’s cooperation or not and the print is visible.The following review is emphatically placed on the research advances of technologies about forged fi ngerprint recognition that mainly involves with morphological comparison approaches and machine learning convolution(especially deep learning).Finally,a brief analysis is given on the future developing trend of forged fingerpr int recognition technology based on the previous expatiation.