When I researched facial recognition I found several algorithms on the web. They were complex and above the level of what this project was supposed to do. They could be used to extend on this project in the future.
Two of the algorithms I read about were eigenface and hidden Markov model. The basic idea of Eigenface is that an average of all the faces is taken make the average face over the facespace. Then pictures are compared to the average to see where they differ and to identify them as a possible match. A more detailed explanation can be found here.
Hidden Markov Method involves dividing the face into blocks and using probability formulas to find where features will lie in the blocks.
A detailed abstract pdf about Hidden Markov Method
More Algorithms
Home