Palmprint biometric data acquisition: extracting a consistent Region of Interest (ROI) for method evaluation
Cox, Andrew (2014)
Cox, Andrew
2014
MDP in Software Development
Informaatiotieteiden yksikkö - School of Information Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2014-12-19
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:uta-201412312554
https://urn.fi/URN:NBN:fi:uta-201412312554
Tiivistelmä
Traditionally personal identification was based on possessions. This could be in the form of a physical key, ID card, passport, or some kind of knowledge based entry system such as a password. All of these are prone to attack where impersonation of your identity for some kind of immediate financial gain, or the more serious identity theft, is possible simply by being in physical possession of an identity device or knowledge of a password.
In contrast biometric identification attempts to identify who you are. Iris or retina patterns, palmprint, fingerprint, face and voice recognition are well known examples of biometric attributes. Some biometrics such as fingerprints were established in the latter 19th century well before computers were commonplace. Others such as face, iris and voice recognition have emerged as computer technology and methodologies have developed. More recent research has also devoted attention to internal physiological biometrics based on brain (electroencephalogram), heart activity (electrocardiogram) and palm vein patterns. Even your personal gait based on how you walk has been investigated. Both security and forensic applications compete to find the best identification method trading off accuracy for performance depending on the intended application.
This thesis is a continuation of previous research to develop a tool for distributed palmprint image data gathering. This would enable researchers to concentrate on method evaluation whilst not losing valuable time in data validation. This simple tool will enable palmprint biometric diversity across continents to be gathered. This thesis continues by establishing how to extract a consistent region of interest in the acquired palmprint images from a mobile phone ,or statically mounted digital, camera. The importance of establishing a consistent region of interest is considered by studying a simple existing identification method applied to a known palmprint database. In the discussions and conclusions the usefulness of this method is established and the final research outlined that is needed to finalize the palmprint acquisition tool for academic research.
In contrast biometric identification attempts to identify who you are. Iris or retina patterns, palmprint, fingerprint, face and voice recognition are well known examples of biometric attributes. Some biometrics such as fingerprints were established in the latter 19th century well before computers were commonplace. Others such as face, iris and voice recognition have emerged as computer technology and methodologies have developed. More recent research has also devoted attention to internal physiological biometrics based on brain (electroencephalogram), heart activity (electrocardiogram) and palm vein patterns. Even your personal gait based on how you walk has been investigated. Both security and forensic applications compete to find the best identification method trading off accuracy for performance depending on the intended application.
This thesis is a continuation of previous research to develop a tool for distributed palmprint image data gathering. This would enable researchers to concentrate on method evaluation whilst not losing valuable time in data validation. This simple tool will enable palmprint biometric diversity across continents to be gathered. This thesis continues by establishing how to extract a consistent region of interest in the acquired palmprint images from a mobile phone ,or statically mounted digital, camera. The importance of establishing a consistent region of interest is considered by studying a simple existing identification method applied to a known palmprint database. In the discussions and conclusions the usefulness of this method is established and the final research outlined that is needed to finalize the palmprint acquisition tool for academic research.