We use the face database from Olivetti Research Lab. This database has been widely used for testing face recognition systems, and the best reported result to our knowledge is 96% correct classification [6], when the database is divided into two equally large sets (for training and testing). The database contains 10 images of 40 subjects for a total of 400 images. The images are gray-level of size . We have manually located the center of the eyes in all the images, and then extracted a window at this region (for both eyes). If this subwindow includes an area outside the original face image, we pad the image with the mean of the gray-level values in the full face image. The extracted eye-images show considerable variation with respect to open/closed eyes, expression, glasses (reflection in glasses) and gaze, making this a challenging dataset. Some example eye-images are shown in figure 1.
Before feature extraction, all images were histogram equalized. As an additional preprocessor for the eigenfeature approach, the mean image was computed from the training set and subtracted from all the training and testing images.