I have been trying in vain for a few weeks to implement the Eigenface algorithm for Face Recognition. It is the technique based on Turk and Pentland's paper. I am working in VC++6. Although I know the steps involved in this algorithm, I am finding it difficult to realise this theory.
Do I have to use some library to convert a raw image in to an array of numbers representing pixel intensities? If so, how does one extract the eigenface from each image in that new format?
I would be extremely grateful if someone has been successful in implementing the eigenface algorithm. I would appreciate any help,