Sparse coding based orientation estimation for latent fingerprints

作者：

Highlights：

• A sparse coding based orientation estimation algorithm is proposed for fingerprints.

• Dictionaries are learned to capture the prior knowledge of various orientations.

• Multi-scale sparse coding is iteratively applied to correct corrupted orientations.

• Experiments on NIST SD27 have shown the effectiveness of the proposed algorithm.

摘要

•A sparse coding based orientation estimation algorithm is proposed for fingerprints.•Dictionaries are learned to capture the prior knowledge of various orientations.•Multi-scale sparse coding is iteratively applied to correct corrupted orientations.•Experiments on NIST SD27 have shown the effectiveness of the proposed algorithm.

论文关键词：Orientation estimation,Sparse coding,Orientation field,Latent fingerprint identification,Biometrics

论文评审过程：Received 28 January 2016, Revised 25 December 2016, Accepted 3 February 2017, Available online 9 February 2017, Version of Record 17 February 2017.

论文官网地址：https://doi.org/10.1016/j.patcog.2017.02.012