Contactless Palmprint Recognition using Deep Learning Technology

Abstract

The interest of researchers in the subject of biometrics has increased, which has opened new horizons in people identification systems, and one of these measurements is the contactless palm print. Identification of people through a contactless palm print is very important in the process of identifying terrorists and criminals whose faces are often covered. Therefore, a contactless palm print recognition system has been proposed through two pre-processing methods using a neural network. A Contrast Limited Adaptive Histogram Equalization (CLAHE) filter was used in the first processing of the data stage as well as the normalization method. The system was applied to several databases, including the standard Indian Institute of Technology Delhi (IITD) ones, and those collected from the Computer Department, College of Science, University of Diyala, After the process of data division, training and testing, the proposed system reached satisfactory results compared to previous work, and the accuracy was 99.95.