Palmprint recognition using polynomial neural network

Abstract

In this paper, we propose a robust palmprint recognition approach. Firstly, a salient-point based method is applied to segment as well as align the region of interest (ROI) from the palmprint image. Then, a subspace projection technique, namely, independent component analysis (ICA) is performed on the ROI to extract features. Finally, a polynomial neural network (PNN) is used to make classification on reduced feature subspace. The effectiveness of the proposed method has been demonstrated in experiments. © 2010 Springer-Verlag.