Neural network technology for pattern recognition, stock prediction and market forecasting

Neural Networks Based Signature Recognition

Download now Matlab source code
Requirements: Matlab, Matlab Image Processing Toolbox, Matlab Neural Network Toolbox and Matlab Signal Processing Toolbox.

Signature verification technology utilizes the distinctive aspects of the signature to verify the identity of individuals. The technology examines the behavioral components of the signature, such as stroke order, speed and pressure, as opposed to comparing visual images of signatures. Unlike traditional signature comparison technologies, signature verification measures the physical activity of signing. While a system may also leverage a comparison of the visual appearance of a signature, or "static signature," the primary components of signature verification are behavioral. In the last few decades, many approaches have been developed in the pattern recognition area, which approached the off-line signature verification problem. There are two main approaches for off-line signature verification: static approaches and pseudodynamic approaches. The first one involves perceptive characteristics, therefore easy to imitate. The second involves imperceptive characteristics, therefore difficult to imitate. As for the verification process, there are many approaches that are used nowadays, for example, Hidden Markov Models, the Euclidean Distance Classifiers, Elastic Image Matching and others. Neural Networks have, in the last decade, attracted the attention of many researchers in the pattern recognition area, for example the recognition of handwritten text, speech recognition and recently the verification of on-line signatures. These models have the capacity to absorb the variability between patterns and their similarity.

Code has been tested using Off line signature database, Grupo de Procesado Digital de Señales, available at http://www.gpds.ulpgc.es/download/index.htm.

Index Terms: Matlab, source, code, signature, on-line, off-line, verification, matching, ann, nn, neural, network, networks.

Release 1.0 Date 2008.12.15
Major features:

• Neural networks based off-line signature recognition
• Fast and optimized implementation
• Customizable code
• Demo code (protected P-files) available for performance evaluation