Abstract: Abstract: Palmprint recognition is one of the most popular biometrics identification technologies. Among the various methods of palmprint recognition, the method based on directional features has achieved good results. In order to further improve the recognition accuracy, a palmprint recognition algorithm combining global and local directional features is proposed, which mainly combines the method based on directional coding, the method based on local descriptor of directional features and the method of combining directional features and correlation filters. The first two methods belong to the spatial domain method, which can extract the local directional features of the palmprint. The third method belongs to the frequency domain method, which can effectively extract the global directional features. The recognition results of the three methods are merged at the score level. The algorithm is validated on two palmprint databases. The experimental results show that the recognition performance of this method is better than other palmprint recognition methods.

Key words: Keywords: biometrics, palmprint recognition, directional representations, information fusion, score level fusion