The Parisian Academic Journal_Applied and Natural Sciences

ISSN:2682-0897

Paris, France

Volume. 2024-1
Issue. 3
First Page:7 Last Page:19

Verification of Arabic Handwritten Signature Based on Artificial Neural Network

Hakim A. Abdo , Ahmed A. Abdu , Ahmed A. Hamoud

Abstract: The main aim of our paper is to verification of Arabic handwritten signature using Artificial Neural Network (ANN) technique. In this work, the features that are used are total area, convex area, equivalent diameter, Euler number, extent, major axis length, mean orientation, solidity and number of objects. Before extracting the features, preprocessing of a scanned image is necessary to isolate the signature part and to remove any spurious noise present. To improve the accuracy of the extracted features of the investigation signature are trained by using an artificial neural network. The same procedure is repeated for the test signature. The system is initially trained using a database of signatures obtained from those individuals whose signatures have to be authenticated by the system. The details of preprocessing as well as the features depicted above are described throughout the discussion. Then ANN technique was used to verify and classify the signatures: exact or forged, the correct classification rate is 86.11 % for generalization and 92.1 % for recall. The implementation details and simulation results are discussed in this work. The signature of Arabic signature based on artificial neural network is implemented using MATLAB R2017a. This work has been tested and found suitable for its purpose.

Keywords:

Signature
verification
artificial
neural
network
features
extraction

Free Full Text Available: PDF