Multiple Hand Based Biometrics Using Multiresolution Framework

Development of Hand-based Biometrics system using palmprint, fingerprint and finger-knuckle print (FKP). The appropriate orientation, multi-scale and multi-resolution transforms along with subspace analysis techniques for extracting the local and global features of fingerprint, palmprint and FKP are indentified. For this purpose various methods like Gabor filter, Radon transform, Wavelet transform along with PCA and LDA and imprint identification using 2D Gabor transform extracts local and global textural features from fingerprint, palmprint and FKP images. This method is invariant to rotation and translation variation of the imprint images. Selection of the performance parameters plays a significant role. Higher values of dx and dy makes the system more robust to noise but will not able to capture edge information at fine level i.e. it is more likely to smooth the image to the extent that the ridge and furrow details in the imprint are lost.