Abstract
This paper introduces a new system to identify handwritten signatures. For feature generation, we propose the Histogram of templates, while the Artificial Immune recognition System (AIRS) is used to achieve the identification task. A writer-independent strategy is proposed to train the AIRS to get an open system that can identify any new writer. Experiments are conducted on a benchmark dataset composed of 55 writers. The results obtained are satisfactory and confirm the usefulness of the proposed system.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have