Fingerprint pore matching using deep features

作者：

Highlights：

• The learning model solved inter-class difference and intra-class similarity of pores.

• The proposed DeepPoreID is very effective to represent the local pore feature.

• Better recognition accuracy is achieved by the proposed method in EER and FMR1000.

• The proposed method deals with partial fingerprint matching problem well.

摘要

•The learning model solved inter-class difference and intra-class similarity of pores.•The proposed DeepPoreID is very effective to represent the local pore feature.•Better recognition accuracy is achieved by the proposed method in EER and FMR1000.•The proposed method deals with partial fingerprint matching problem well.

论文关键词：Fingerprint recognition,Pore representation,Direct pore matching,Convolutional neural networks

论文评审过程：Received 13 April 2019, Revised 18 December 2019, Accepted 16 January 2020, Available online 16 January 2020, Version of Record 22 January 2020.

论文官网地址：https://doi.org/10.1016/j.patcog.2020.107208