Contactless finger-vein verification based on oriented elements feature

Abstract

Finger vein pattern has emerged as one of the most popular and promising biometric features for personal identity verification due to its uniqueness, stability, high distinguishability and non-invasive procedure. Many finger vein recognition algorithms have been proposed and successfully applied for identity authentication in the past decades. However, most of the previous methods usually only use the images captured by using a special fixed setting to avoid translation and rotation resulting from finger moving during image acquisition. Contactless finger vein recognition has recently begun to draw attention of researchers because of its user friendliness. Contactless vein images are captured under free conditions and usually have significant variations on rotations and translations. Conventional powerful recognition researches are not very effective for contactless vein images. Therefore, this paper proposed a new method based on finger vein patterns by employing region of interest extraction and oriented elements feature extraction scheme to effectively deal with non-ideal scenarios such as variations. A region of interest based on rotation rectified is employed to reduce the effects of rotation and translation during finger vein image capture. Then, the direction feature is drawn from the texture and steady orientation characteristics of finger vein images by using the magnitude and orientation of the gradient of points on finger vein lines in fuzzy division region. Using the extracted feature can effectively solve the problem of the geometric deformation. High accuracy has been obtained by the proposed scheme. Experimental results show that this new algorithm is effective and feasible in finger vein recognition system.