Handling small rotations in Finger vein recognition
April 12, 2018
Biometrics is about recognizing persons based on their physical properties, behavior or traces they leave. Examples of biometric modalities are face, fingerprint, iris, voice etc. Many of these modalities have the disadvantage that they can easily be copied or obtained, e.g. it is very easy to take a photograph of somebodies face or to obtain his fingerprints from a smooth surface like a wine glass. Using the copied biometric, access can be obtained to mobile phones, secure areas and information etc. The vein pattern inside the finger can also be used as a biometric, because it is unique for every person, but it is very hard to copy this without the cooperation of the person involved. Therefore at DMB, we designed a finger vein scanner and investigate finger vein recognition. Around this setup a number of student assignments have been defined.
Figure 1: Left/middle: Finger vein scanner, Right: nger vein patterns
The finger vein acquisition system as shown in the figure above is our 2nd version and the hardware has been developed by a bachelor student. It includes 3 cameras for future research on pose independent and 3d finger vein recognition. From a research perspective it is interesting to investigate the influence of small rotation of the finger with respect to non rotated fingers on sensor image and system performance. There are two principled manners to obtain such images: rotating the hand by a slab construction or rotating the sensor(s) on a ring. This assignment addresses the design and implementation of an update to the finger vein scanner to include the small rotations and investigation of the inuence on the images and recognition performance.