Personal Identification via Hand Feature Extraction Algorithm

摘要

Identification of people using biometric traits from hand components like fingerprint, hand-vein, palm-print and hand geometry is more secure compared to Personal Identification Number (PIN). Extracting robust features from these biometric traits is the most difficult challenge that needs to be tackled for accurate identification of people. In this work, extraction of robust features from hand image is proposed. This work involves preprocessing of hand images to obtain edge detected hand images, generation of feature vector suitable for identification of people using Euclidean distance. Experiment was carried out to compare hand feature vectors extracted from different hand images. The results obtained shows high inter-personal variation and very low intra-personal variation.