Our objective is the design of a face recognition system for its use in controlled environments. For that reason, different tasks will be performed, that chronologically can be stated as: The study of the basis of pattern recognition, and especially the statistical pattern recognition. The study of the main feature selection techniques, focusing in principal component analysis (PCA), and the basic classifiers, as the nearest neighbour and Parzen classifiers. The study of the main combinations techniques, as we will try to take advantage of the availability of sets of faces to make better-founded decisions The evaluation of the main problems, like pose variation or light changes. The choice and posterior adaptation of these different general tools to our needs, like the use of modifications of the Parzen classifier. The design of the main algorithm, focusing on the differences between the two working modes, identifying (when we try to recognize a person among the different people in our Data collection Cut out faces Normaliza-tionFeature selection PCA / LDA projectionModel selection Represen-tation of the classes UpdatingClassifier selection k-NN ParzenTraining Parameter optimiza-tion Testing Identifica-tion / verifica-tion of groups of faces 14 system) and verifying (when we know a priori the id of the person and we try to assure its identity), and the use of single images or sets of faces. The implementation of the algorithms in C, trying to become familiar with the development environment and considering the possible optimization but maintaining the clarity of the code. The testing of the whole application using faces from different databases and from real data; that is, captured in real scenarios and in non optimal conditions