IFS04-Towards More Accurate Matching of Contactless Palmprint Images under Less Constrained Environments

By | July 30, 2018

ABSTRACT:

Contactless personal identification utilizing biometrics qualities brings multifaceted advantages with enhanced cleanliness, client security, and the convenience. Such imaging likewise produces deformation free palmprint images which can prompt higher matching exactness as the ground truth data is better protected as contrasted and those from contact-based imaging. Advancement of palmprint identification technologies for new domains requires research about utilizing larger palmprint databases that are gained from more sensible populaces, under contactless, surrounding, and indoor and outdoor environments. This project shows such another contactless palmprint database procured from 600 different subjects, which is the larger to date and is likewise made available in the general public domain. Not at all like contactless fingerprints, contactless palmprint images often illustrate pose deformation along the optical axis of the camera, which likewise degrades the matching accuracy. This project additionally presents another approach for coordinating contactless palmprint images utilizing accurate deformation alignment and matching. The experimental results are approved on three freely accessible contactless palmprint databases. Comparative results introduced in this project demonstrate consistently outperforming over competing techniques in the literature and validate the effectiveness of the investigated approach. These results additionally fill in as baseline performance to advance much necessary further research utilizing the most difficult and larger database presented from this project.

Leave a Reply

Your email address will not be published. Required fields are marked *