Off-line Signature Recognition Using Weightless Neural Network and Feature Extraction Ali Al-Saegh Computer Engineering Department University of Mosul Mosul, Iraq ali.alsaegh@uomosul.edu.iq I. INTRODUCTION The signature is a conventional biometric way for authentication and authorization purposes in financial and legal transactions. Handling bank accounts, employee attendance and other such situations require systems for automatic signature recognition. Identifying the owner of a signature using an automatic recognition system involves comparing the provided signature with those stored in a database. While the verification process tests the provided signature and gives an output of either accepting it as a genuine signature or rejecting it as a forged signature [1]. Hence, due to its lively role in our daily life, the field of signature recognition and verification has been widely investigated. Two methods for automatic signature recognition are available, namely the on-line and off-line methods. The former method also called the dynamic method utilizes time, stroke, speed, and pressure of writing. So, signatures which are recognized using this method are captured using digitizing tablets or pressure sensitive pens. The latter method also called the static method utilizes scanned or photographed images of signatures for recognition. Off-line recognition is cheaper than the on-line one [2], while it is more complex because of the lack of available information about the signature [3]. In this paper, a novel use of the weightless neural network (WNN) and feature extraction is presented. The WNN is flexible, easy to design and implement, and capable of training with few samples. Besides these characteristics, the proposed feature vector is easy to compute, has relatively reasonable size and differentiates among samples excellently. Therefore, the proposed system is considered as a new way for handwritten signature recognition in an off-line manner and can be generalized for solving a variety of problems. The rest of this paper is organized as follows: Section II introduces some previous works, Section III explains briefly the weightless neural network, Section IV describes the overall proposed system, Section V gives an explanation about the used data set and the designed weightless neural network, Section VI presents the obtained results and a comparison with other studies, and Section VII concludes this paper. II. LITERATURE REVIEW Vast number of researches considered the واﻻﻟﻜﺘﺮوﻧﻴﺔ اﻟﻜﻬﺮﺑﺎﺋﻴﺔ ﻟﻠﻬﻨﺪﺳﺔ اﻟﻌﺮاﻗﻴﺔ اﻟﻤﺠﻠﺔIraq J. Electrical and Electronic Engineering ﻡﺠﻠﺪ11 اﻟﻌﺪد ، 1 ، 2015 Vol.11 No.1 , 2015 Abstract : The problem of automatic signature recognition and verification has been extensively investigated due to the vitality of this field of research. Handwritten signatures are broadly used in daily life as a secure way for personal identification. In this paper a novel approach is proposed for handwritten signature recognition in an off-line environment based on Weightless Neural Network (WNN) and feature extraction. This type of neural networks (NN) is characterized by its simplicity in design and implementation. Whereas no weights, transfer functions and multipliers are required. Implementing the WNN needs only Random Access Memory (RAM) slices. Moreover, the whole process of training can be accomplished with few numbers of training samples and by presenting them once to the neural network. Employing the proposed approach in signature recognition area yields promising results with rates of 99.67% and 99.55% for recognition of signatures that the network has trained on and rejection of signatures that the network .has not trained on, respectively Index Terms—Feature extraction, off-line signature recognition, RAM-based neural network, weightless neural network (WNN) 124