| Offline Signature verification using deep learning method |
| Paper ID : 1087-ISCBAS |
| Authors |
|
Nehal Hamdy Al-Banhawy *1, Heba Mohsen2, Neveen Ibrahim Ghali3 1Math and computer science department, Faculty of Science, Al-Azhar university, Cairo, Egypt 2Lecturer, Computer Science Department, Faculty of Computers and Information Technology, Future university in Egypt, New Cairo, Egypt 3Head of Digital Media Technology Department, Faculty of Computers and Information Technology, Future university in Egypt, New Cairo, Egypt |
| Abstract |
| One of the biggest challenged biometric authentication problems in recent years that we are often exposed to in our daily lives is signature verification. There are two main types of signature verification systems: Offline (static) systems and Online (dynamic) systems. The offline signature verification systems are more difficult than online systems because it have additional information such as velocity of writing, pen up and down which allow to extract more features. This paper presents a deep learning method based on convolutional neural network (CNN) architecture for solving offline signature verification problem to prevent signature faking by thieves. The CNN was used for extracting features and classifying the signatures as genuine or forged. The proposed method succeeded in achieving an accuracy of 94.73 % on Cedar dataset by using a genuine signatures and forged signatures for testing, which indicate that the method was effective and can be supported by more feature extractors to get better results. |
| Keywords |
| signature verification; deep learning; convolutional neural network |
| Status: Abstract Accepted |