| Autor: |
Abdullahi, Sunusi Bala, Bature, Zakariyya Abdullahi, Muhammad, Auwal |
| Předmět: |
|
| Zdroj: |
International Journal of Design, Analysis & Tools for Integrated Circuits & Systems; Dec2022, Vol. 11 Issue 2, p43-47, 5p |
| Abstrakt: |
The existing works on multimodal biometric cues (fingerprint and veins) were found effective for biometric recognition. However, small images size and failure to extract the spatial and temporal information of the existing methods lead to low recognition accuracy. This work proposed spatial and temporal multimodal fingerprint and finger veins network (STMFPFV-Net). In this method, effective image augmentation techniques and large image sizes were utilized as a proxy in the deep convolutional neural network fusion model to improve image variabilities. Recognition was performed, using concatenated layers of a sequential dense network. To evaluate the performance of the proposed method, the NUPT-FPV dataset was used, which contained 16,800 images for fingerprints and finger veins each, respectively. These images were extracted from 6 fingers of 140 participants in both left and right hands, repeated 10 times each. The proposed method when evaluated using standard protocols outperform existing algorithms. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
