A Finger Vein Image Recognition based on DLT
B. Yamini, R. Kennady, M. Nalini, R. Sıva Subramanıan, M. Anuradha, D. Chitra Devi
Abstract
Since existing finger vein segmentation networks are overly complex and not appropriate for implementation in portable devices, this article proposed an inexpensive real-time finger vein classification method that uses the deep learning technique. Decreasing the parameters of the compact network leads to a deduction of the segmentation index and the large-running time required for deep learning networks on hardware systems. Finger vein recognition is a biometric identification technology that relies on pattern recognition algorithms based on pictures of human finger vein patterns under the skin’s surface. Finger Vein ID is a biometric identification method that compares an individual’s vascular pattern to data already collected. Since finger veins are different even in twins who are identical, reside under the skin, and stay with an individual for a lifetime, finger vein identification eliminates the risk of fraud. The recognition of finger vein patterns has substantially improved thanks to a number of deep learning approaches.