Mobile based palmprint recognition system

• Main Author: Munjal, Neera
• Other Authors: Li Fang
• Format: Final Year Project
• Language: English
• Published: 2015
• Subjects: DRNTU::Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision
• Online Access: http://hdl.handle.net/10356/62813
• Institution: Nanyang Technological University

Personal identification and verification play an important role in the society. Traditional authentication methods like password and smart card, often cannot meet today’s security as they can be easily forgotten, lost or stolen. Biometric technology is a new solution to these problems. Among all the variable biometric technologies, automatic palmprint verification is an important complement to biometric authentication. Biometric is a way to identify humans by their physiological characteristics or traits like voice, DNA, hand print or behaviour with the use of Digital Image Processing and Pattern Recognition technologies. Generally, biometrics involves comparing a person’s physical features to the image of those features stored on a computer system in order to verify or determine the identity. Recently mobile devices have become a vital part of our daily lives. We use mobile devices to do mobile banking etc. which has increased the need to secure smartphones and hence there is a need to implement a mobile based verification system. This project comprises of 4 main modules which are Image Acquisition, Palm Positioning, Feature Extraction and matching. The biometric-based person identification system works on an 8-bit gray scale palmprint TIFF image. Region of Interest (ROI) extraction is then performed so that the palms are properly aligned and normalized. After extracting the clip Region of Interest (ROI), the system generates the Line Edge Map (LEM) representation of the palmprint features, and performs features matching based on Line Segment Hausdorff Distance. To conclude, in this project a mobile based palmprint verification has been implemented and with the already provided database the system achieves the FAR and FRR at 4.968% and 4.3% respectively at a threshold of 1.6