నైరూప్య

Person Identification Based on Recognition of Palmprint Using Orientation Features

Basanti B. Sawant, M.Talib,Sagar S. Jondhale , Pradeep M.Patil

In this paper orientation features of palmprint images have been efficiently used for person identification. Performance of the proposed algorithm has been evaluated using standard database of PolyU available online. The region of interest (ROI) is extracted from the palmprint images and the variance feature vector of orientation field has been computed and is treated as the feature vector (template) of that palmprint for the person and stored in the database during enrolment. When a query palmprint image is applied, the ROI and variance feature vector of orientation field is computed and matched with that of the templates in the database using Euclidian distance to find best match. The algorithm performs well for various thresholds on the PolyU database.