Hello, I carefully read your project proposal. For years I have developed artificial intelligence techniques including face recognition, speaker verification. For face recognition I have FERET database-UOS(color and grey) and our database(1250 persons) built ourselves. In order to train model,I used Adaboost method using subspace analysis and neural network analysis and optimized Gabor feature parameters. Recognition performance is 97% in front, 30 degrees of right-left, 15 degrees of rotate for face state, and in the task of more than 100 persons. Currently I have the project(.Net) similar to this project. It do capture the photo of a person using camera, detect face and idetify person by face and resister details for the person into database. If you have interest, I will show the example similar to your project. I will wait your reply. Sincerely!