A novel Approach to improve biometric authentication using Steerable-Locality Sensitive Discriminant Analysis

Abstract

the increasing needs in security systems, palm vein authentication is one of the important and reliable solutions for identity security for biometrics based authentication systems. Palm vein, as a biological characteristic of an individual, has been increasingly utilized for personal authentication in advanced security applications. Palm vein patterns are a unique attribute of everyone and can therefore be used as a biometric characteristic. The human palm vein pattern is extremely complex and it shows a huge number of vessels. The biometric information is located inside the human body, and therefore it is protected against forgery and manipulation. In the proposed method, the Multilevel Gaussian-Second-Derivative (MGSD) is proposed for enhancement the palm vein images. Secondly, a new feature extraction method based on Steerable filter and Locality Sensitive Discriminant Analysis is proposed called Steerable - Locality Sensitive Discriminant Analysis (SLSDA). Finally, the Correlation Distance method is proposed for verify the tested palm vein. The EER to the proposed authentication system is 0.1087%.