Offline Handwritten Signature Recognition based on Discrete Cosine Transform and Artificial Neural Network
Jumana Eltrabelsi, Ahmed O. Lawgali · 2021
Abstract- Signature recognition is of great importance in our daily life as well as in automatic identification systems. It is an acceptable biometric method used to identify a person, because each person is distinguished by its signature. This paper presents a technique based on Discrete Cosine Transform to capture features to distinguish signatures handwritten. Discrete Cosine Transform parameters are extracted from signature image and these parameters are entered for the Artificial Neural Network for the purpose of classification. This work was tested with sigcomp2011 database and a comparison made against some existing techniques, and promising results have been obtained.