Machine learning-based offline signature verification systems: A systematic review

作者：

Highlights：

• A systematic review of ML-based offline signature verification systems.

• Concentrates on datasets, preprocessing techniques and feature extraction methods.

• Reviewed ML-based forgery detection models and performance evaluation metrics.

• Consolidates the state-of-the-art OfSV systems performances on five public datasets.

• Fifteen open research issues are also identified.

摘要

•A systematic review of ML-based offline signature verification systems.•Concentrates on datasets, preprocessing techniques and feature extraction methods.•Reviewed ML-based forgery detection models and performance evaluation metrics.•Consolidates the state-of-the-art OfSV systems performances on five public datasets.•Fifteen open research issues are also identified.

论文关键词：Offline signature verification,Feature extraction,Writer identification,Deep convolutional neural network,Handwriting recognition,Signature forgery detection

论文评审过程：Received 24 April 2020, Revised 29 October 2020, Accepted 6 January 2021, Available online 13 January 2021, Version of Record 18 January 2021.

论文官网地址：https://doi.org/10.1016/j.image.2021.116139