View article

Because of their high level of precision, biometric systems continue to attract the attention of several researchers. Different biometric traits have been investigated for use in security systems, such as fingerprints, faces, irises, palmprints, and knuckle prints. In most cases, bi-dimensional information is utilized. To achieve this aim, we have examined the performance of biometric identification systems based on a 3D-FKP database through five pre-trained networks such as AlexNet, VGG19, GoogleNet, ResNet50, and DenseNet201. The obtained experimental results illustrate the effectiveness of the suggested approach, with a high recognition rate and accuracy.