A low cost finger vein authentication system, using maximum curvature points

Finger vein patterns have been proposed as a suitable biometric feature for authentication applications. Systems using this feature are generally low cost, accurate and easy to use. Such a system is described in this work. Infrared light is used to capture an image of a finger and a pattern recognition algorithm extracts the vein patterns. For robust and precise extraction of the depicted veins, a method of calculating local maximum curvatures in cross-sectional profiles of a vein image was used. The authentication algorithm has been evaluated with images acquired from the device built, as well as with images from a 400 fingers database. The false acceptance and false rejection rates obtained were promising.