
Offline Signature Verification based on Centerline Similarities
By: Erika Griechisch and Gábor Németh
Abstract
Digital signatures are becoming increasingly common in author identification. However, handwritten signatures play an important role in different aspects of life like business and the banking sector. Offline signature verification methods analyze the images and shapes of the signatures. Several methods use skeleton that is a frequently used shape descriptor that summarizes the general form of objects.
Here, we present an offline signature verification method which is based on similarity measures designed for a comparison of 2D skeleton-like shape features. The proposed method was evaluated on the publicly available SigComp2011 and SigWiComp2013 databases. The results that we obtained are competitive with the systems submitted to two recent signature verification competitions.
Here, we present an offline signature verification method which is based on similarity measures designed for a comparison of 2D skeleton-like shape features. The proposed method was evaluated on the publicly available SigComp2011 and SigWiComp2013 databases. The results that we obtained are competitive with the systems submitted to two recent signature verification competitions.
DOI: https://doi.org/10.69525/jasqde.219 | Journal eISSN: 1524-7287
Language: English
Page range: 17 - 27
Published on: Dec 1, 2015
Published by: American Society of Questioned Document Examiners
In partnership with: Paradigm Publishing Services
© 2015 Erika Griechisch, Gábor Németh, published by American Society of Questioned Document Examiners
This work is licensed under the Creative Commons Attribution 4.0 License.