A Portable Biometrics System Based on Forehead Subcutaneous Vein Pattern and Periocular Biometric Pattern

Abstract

In this paper, we described the development of a contact-less and real-time system for entry control based on the forehead subcutaneous vein pattern and periocular biometric pattern. The system is developed with a single-board computer along with some accessories. This makes the system portable and low-cost. We named it Forehead vein and Periocular Pattern-based Biometric System (FPPBS). In the FPPBS, the camera acquires the pattern of vascular structure present in the subcutaneous layer of the forehead region and the edge patterns of the periocular region. The images captured are processed to get the vein and periocular patterns highlighted before feature extraction and matching. The matching is done with a newly proposed deep learning-based algorithm named Vein and Periocular Pattern-based Convolutional Neural Network (VP-CNN). In this paper, we also have presented a database namely the FSVP-PBP database. The proposed FPPBS is also compared with the contemporary biometric systems. It is found to be both beneficial and better performing than the systems in use for entry control.