Joint Line and Orientation Features in 3D Palmprint

Abstract

Two dimensional palmprint has been recognized as an effective biometrics identifier in the past decade. Recently, three-dimensional palmprint recognition was proposed to further improve the performance of palmprint systems. This chapter presents a simple yet efficient scheme for three-dimensional palmprint recognition. After calculating and enhancing the Mean: Curvature Image of the three dimensional palmprint data, we extract both line and orientation features from it. The two types of features are then fused at either score level or feature level for the final three-dimensional palmprint recognition. The experiments on the HKPU three-dimensional palmprint database which contains 8,000 samples from 400 palms show that the proposed feature extraction and fusion methods lead to promising performance.