Date of Graduation

12-2018

Document Type

Thesis

Degree Name

Bachelor of Science in Computer Engineering

Degree Level

Undergraduate

Department

Computer Science and Computer Engineering

Advisor

Parkerson, Pat

Reader

Nelson, Alexander

Second Reader

Gauch, John

Abstract

Fingerprint identification is one of the most accurate sources of identification, yet it is not widely used in public facilities for security concerns. Moreover, the cost of fingerprint system is inaccessible for small-budget business because of their high cost. Therefore, this study created an open-source solution to secure fingerprint samples in the database while using low-cost hardware components. Locality Sensitive Hashing Algorithms such as ORB and Image hash were compared in this study as a potential alternative to SURF. To test the design, fifteen samples were collected and stored in a database without verifying the quality of the samples. Then, thirteen other samples were read from the sensor and forty-five permutations were created from the first fifteen samples. The results showed that a low-cost system can secure fingerprint sample in a database using Open-source technologies, but the identification process needs some improvement. Also, the study showed that image hash is a good alternative to SURF when the sensors readings are a force to one position.

Keywords

Fingerprint, security, database, feature extractor

Share

COinS