Face and Gender Recognition Using Principal Component Analysis

Face recognition is a biometric analysis tool that has enabled surveillance systems to detect humans and recognize humans without their co-operation. In this paper we evaluate the basics of the Principal Component Analysis (PCA) and verify the results of this algorithm on a training database of images. The same principle is in effect used to recognise the gender of the test image by evaluating the Euclidian distance of the test image from the images in the database. The proposed gender and face recognition technique using PCA is verified for both test images of a man and a woman. It was observed that if the number of images of a particular subject was more in the database, the gender recognition becomes even better. The effect of salt and pepper noise and image cropping was also observed and the results hold true for noise up to 40 percent of the image pixels. KeywordsFace Recognition, Gender Recognition, Principal Component Analysis, Eigenfaces, noise, cropping.