Biometric signature verification system

• Main Authors: Badenhop, Henry Peter S., Cheng, Melissa C., Dizon, Ma. Arlette F., Go, Joel T.
• Format: text
• Language: English
• Published: Animo Repository 1995
• Subjects: Digital electronics; Signature writing machines; Signs and symbols > Data processing; Data transmission systems; Electronic data processing > Distributed processing; Analog-to-digital converters; Programming (Electronic computers); Engineering
• Online Access: https://animorepository.dlsu.edu.ph/etd_bachelors/14009
• Institution: De La Salle University

Signature Verification involves the checking for the authenticity of a signature. It is composed of a hardware system that acquires the data and a software algorithm that verifies the signature. There are many techniques in signature verification but such approaches rely on different basis. This thesis will present a technique in signature verification based on the dynamic characteristics of the signature, namely pressure and acceleration. A concept of acceleration in an invariant system will be serve as basis for the necessary acceleration data. A neural network will be used to find out the most likely set of pressure inputs and acceleration inputs for a particular signature. The personal computer will then disclose the authenticity of the signature.