A COMPARATIVE ANALYSIS FOR IMAGE FUSION APPROACH TO MULTISPECTRAL PAMPRINT RECOGNITION

Enhancing the reliability of palmprint recognition system lies in its robustness and antispoofing capabilities. Multispectral imaging offers new techniques to obtain more discriminative features from images collected under different wavelengths. This work presents the fusion of palmprint images taken under different light illumination Red, Green, Blue and Near Infra-red (NIR) concatenated to form longer feature vector at Image level. The experiment involves the fusion of the individual spectra as triple (R, B, NIR and G, B, NIR) and quadruple (R, G, B, NIR) followed by verification of system performance by a number of classifiers. The method demonstrates an increase in recognition performance among the concatenated spectra and single spectral whe compared to previous works.