Photorefractive correlator using spatial eigenimage filters for real-time human face recognition

In this paper, techniques of volume holographic associative storage and statistical pattern recognition method are combined to develop a photorefractive correlator for real-time human face recognition. A set of eigenimages extracted from a large number of training images by K-L transform are stored in a photorefractive crystal by using the two-wave mixing volume holographic storage technique and used as the spatial filters. When any new image inputs the correlator, spatially separated beams with different light intensities are obtained in parallel, which represent the optical correlation results between the input and the set of eigenimages. Treating them as feature vectors, the statistical discriminant analysis will be applied to these features for classification. Since the face images in the training set can be selected to representing typical distortions, the system can deal with various distortions to a large extent.