Wavelet based independent component analysis for palmprint identification

This work presents a multi-resolution analysis based independent component analysis (ICA) method for automatic palmprint identification. The ICA is well known by its feature representation ability recently, in which the desired representation is the one that minimizes the statistical independence of the components of the representation. Such a representation can capture the essential feature and the structure of the palmprint images. At the same time, the palmprints have a great deal of different features, such as principal lines, wrinkles, ridges, minutiae points and texture, which can be regarded as multi-scale features. Then, it is reasonable for us to integrate the multi-resolution analysis method and ICA to represent the palmprint features. The experiment results show that the integrated method is more efficient than ICA algorithm.