Face and palmprint multimodal biometric systems using Gabor-Wigner transform as feature extraction

This paper explores different multimodal biometric systems based on Gabor-Wigner transform (GWT) for subject recognition. This transform provides a simultaneous analysis of space and frequency components of a biometric image. GWT was initially proposed in the literature for signal analysis. In this technique, the GWT is utilized for extraction of feature vectors from different biometric modalities. An optimization technique, particle swarm optimization, is then used to select the dominant features from the feature vectors. This technique not only improves the performance of the system but also reduces the dimension of the obtained feature vectors. A detailed study has been carried out to investigate the fusion of face and palmprint images at different levels. The receiver operating characteristic curve and the equal error rate are used to evaluate the performance of the technique.