Palm Print Recognition Based on a Fusion of Feature Selection Techniques

Automatic recognition of biometrics has been considered the best way for identifying and authenticating individuals. Palm-print recognition is one of the newest techniques in this field. This paper presents an approach to the automatic recognition of Palm-prints based on 2D images. The proposed approach adopts a combination of three different techniques of feature selection: histograms of oriented gradients, local binary patterns, and principal component analysis to increase the classification performance. For the classification, linear discriminant analysis is applied. The proposed approach was tested using palm-print datasets from PolyU-II and IIT-Delhi. Both datasets produced results that were more than 99% accurate. These findings outperform those of previous studies.