A Comparative Study of Palmprint Recognition Algorithms

Abstract

Palmprint images contain rich unique features for reliable human identification, which makes it a very competitive topic in biometric research. A great many different low resolution palmprint recognition algorithms have been developed, which can be roughly grouped into three categories: holistic-based, feature-based, and hybrid methods. The purpose of this article is to provide an updated survey of palmprint recognition methods, and present a comparative study to evaluate the performance of the state-of-the-art palmprint recognition methods. Using the Hong Kong Polytechnic University (HKPU) palmprint database (version 2), we compare the recognition performance of a number of holistic-based (Fisherpalms and DCT+LDA) and local feature-based (competitive code, ordinal code, robust line orientation code, derivative of Gaussian code, and wide line detector) methods, and then investigate the error correlation and score-level fusion performance of different algorithms. After discussing the achievements and limitations of current palmprint recognition algorithms, we conclude with providing several potential research directions for the future.