Face localization based on the morphological multiscale fingerprints

Face localization is a previous step for face recognition systems. We propo se morphological multiscale fingerprints (MMF) as a holistic feature extrac tion technique for face localization. The MMF is computed as the local maxi ma and minima preserved up to a certain scale in a multiscale analysis base d on morphological erosion and dilation. We do not consider any geometrical model of the face features. As an initial validation of our approach, we h ave compared it with principal component analysis (PCA) over a small custom image database. Afterwards, we have tested it over the CMU image database, with good results. Finally, we present the receiving operator curve (ROC) results as well as some instances of face detection. (C) 2001 Elsevier Scie nce B.V. All rights reserved.