Efficient Point based Feature Extraction Methods for Palmprint Biometric System

Abstract

Biometrics is a method of evaluating human characteristics in order to authenticate or classify an individual’s identity. Palmprint is one of the human physiological characteristics that attracts the attention of researchers as a means of protection.In this work, the Chinese Academy of Sciences Institute of Automation (CASIA) used for investigations. It is understood that corners are areas where the force varies in all directions. Palmprint features can be in the form of corner positions. Efficient point-based feature extraction methods used for various object recognition are selected and discussed. Those methods were implemented on palmprint and their efficiency is analyzed. The palmprint Point features are extracted using SUSAN Operator, Wavelet-based Salient Point Detection, Trajkovic and Hedley Corner Detector and Forstner Operator. In which Forstner Operator performs good and achieves 95.35% accuracy for 1% of FAR and 95.7% accuracy for 2% of FAR.