The following information was submitted:
Transactions: WSEAS TRANSACTIONS ON SIGNAL PROCESSING
Transactions ID Number: 32-764
Full Name: Ashraf Zaher
Position: Associate Professor
Age: ON
Sex: Male
Address: Kuwait University, Science College, Physics Department, P. O. Box 5969, Safat 13060, Kuwait
Country: KUWAIT
Tel: +965 99657906
Tel prefix:
Fax:
E-mail address: zaher_ashraf@yahoo.com
Other E-mails: ashraf.zaher@ku.edu.kw
Title of the Paper: A Robust Hybrid Technique for Signature Verification using Intelligent Encoding of Spatiotemporal Data
Authors as they appear in the Paper: Ashraf A. Zaher and Abdulnasser Abu-Rezq
Email addresses of all the authors: zaher_ashraf@yahoo.com
Number of paper pages: 25
Abstract: This paper combines both offline and online signal processing techniques to construct a hybrid signature verification system that has a robust performance. The proposed system is real-time and is aimed at verifying handwritten signatures collected using a digitizing tablet. Consistency checking is performed prior to allowing signatures to be enrolled in the database via collecting 10 signatures from each signer and measuring the deviation in both the total signing time and the binary pattern of the pen movements from the average of the best six signatures. The system has a flexible architecture that can be easily modified to accommodate different degrees of security and consists of three consecutive phases. The first (introductory) phase is an online approach that is quite similar to the enrollment phase and represents an initial bottle neck for the verification process so that simple forgeries are quickly filtered out. The second (main) phase uses a combination of!
neural networks and linear predictive coding to construct a majority voting committee in a pattern recognition context to decide on the authenticity of the signatures that passed the first phase of the verification process. The third (optional) phase of the system encapsulates real-time data features into a set of stationary image frames suitable for offline processing. This phase is only required if the second phase fails to give a conclusive verdict whether a questionable signature is authentic or forged. A total of eight features were used in the implementation of the system that can be easily constructed using commercially-available software and hardware resources. The results proved to be promising achieving a 2.9% for the false acceptance rate and 8.8% for the false rejection rate. The paper concludes with a detailed discussion regarding the advantages and limitations of the proposed system and suggests feasible improvements as well as future extensions.
Keywords: Signature Verification, Pattern Recognition, Hybrid Systems, Neural Networks.
EXTENSION of the file: .pdf
Special (Invited) Session:
Organizer of the Session:
How Did you learn about congress:
IP ADDRESS: 62.215.90.54
