Palmvein Recognition using Apparent and Latent Direction Code

Abstract

Palm vein trait is used for personal authentication. The existing directional approach extracts only the surface direction features from the raw palm vein images and lacks the sensitive latent direction function of the palm vein image. The considered, Double-layer extraction method for identification of palm vein extracts an insightful latent direction feature. It first eliminates the evident path from a palm vein surface layer. Then latent directional characteristics are extracted from the palm vein energy map code. The apparent and latent direction characteristics are pooled as the characteristics of the histogram by multiplication schemes. For feature matching, correlation coefficients are used.