Multispectral palmprint recognition using Image-Based Linear Discriminant Analysis

In order to improve the accuracy of palmprint recognition, multispectral imaging has been employed to acquire more discriminative information. But how to use the multispectral information sufficiently is a problem to be solved. In this paper we using the Image-Based Linear Discriminant Analysis (IBLDA) and fusing the four bands palmprint images. We divide the four bands palmprint images into two groups and then construct a complex matrix using the two bands of every group and did the feature level fusion for each group respectively. The experiment shows that a higher recognition rate can be achieved when we used IBLDA method.