Hybrid Off-Line Handwritten Signature Verification Based on Artificial Immune Systems and Support Vector Machines
This paper proposes a new handwritten signature verification method based on a combination of an artificial immune algorithm with SVM. In a first step, the Artificial Immune Recognition System (AIRS) is trained to develop a set of representative data (mem
- PDF / 382,953 Bytes
- 8 Pages / 439.37 x 666.142 pts Page_size
- 6 Downloads / 171 Views
t. This paper proposes a new handwritten signature verification method based on a combination of an artificial immune algorithm with SVM. In a first step, the Artificial Immune Recognition System (AIRS) is trained to develop a set of representative data (memory cells) of both genuine and forged signature classes. Usually, to classify a questioned signature, dissimilarities are calculated with respect to all memory cells and handled according to the k Nearest Neighbor rule. Presently, we propose the training of these dissimilarities by a Support Vector Machine (SVM) classifier to get a more discriminating decision. Histogram of oriented gradients is used for feature generation. Experiments conducted on two standard datasets reveal that the proposed system provides a significant accuracy improvement compared to the conventional AIRS. Keywords: Artificial immune recognition system ture verification · HOG · SVM
1
· Handwritten signa-
Introduction
The authentication of individuals based on their handwritten signatures is required in various life domains like official contracts, banking or financial transactions. Thereby, Signature Verification Systems (SVS) are developed to allow an automatic and robust identity verification. Roughly, there are two approaches for developing an SVS. The first is based on on-line dynamic signatures while the second approach is dedicated for off-line signatures that are previously affixed on paper documents. Also, the verification process can be addressed either as writer-dependent or writer-independent [1]. In the case of writer-dependent, a specific verification system is developed for each person to authenticate its genuine signatures. Furthermore, the writer-independent verification develops one generic system for all persons, which aims to detect genuine signatures without giving information about the writer identity. During the past years, various signature verification systems were developed based on Dynamic Time Warping, neural networks, Hidden Markov Models and c Springer International Publishing Switzerland 2016 A. Campilho and F. Karray (Eds.): ICIAR 2016, LNCS 9730, pp. 558–565, 2016. DOI: 10.1007/978-3-319-41501-7 62
Hybrid Off-Line Handwritten Signature Verification
559
SVM [2]. Currently, SVM is the most commonly used classifier since it can significantly outperform the others [3]. However, verification scores reported in literature still require improvement, which lets the development of efficient verification methods an ongoing research topic. Recently, inspired from the human immune system, the Artificial Immune Recognition System (AIRS) introduced by Watkins [4] provided interesting performance in various pattern recognition tasks, such as thyroid diagnosis [5] and fault detection [6]. In earlier works, we have successfully employed the AIRS for writer-dependent off-line signature verification [7–10]. Basically, AIRS imitates some functions of the natural immune system. It adopts a supervised learning process to create new data, called memory cells that represent all classes. Then, k Nearest Nei
Data Loading...
