Reconocimiento y creación del modelo facial 3D mediante sistema de video aplicado a la seguridad usando inteligencia artificial.

Abstract

With the present work designed and implemented a system of recognition and creation of the 3D facial model that simulates any type of input to a certain place. The system involves two stages: the user registration stage, and the execution stage of the main program in which the recognition and creation of the model is carried out. It used the detection method using a generic model of the face, besides the method of artificial intelligence was fuzzy logic, whose fuzzy controller is the Mamdani type and with a defuzzifier that was used to validate the similarity between the active user, and the information stored in the database. For the production of characteristic vectors of the users, applied the Gabor filters and for comparison used the Euler distance. In order to have more precision when recognizing a person used as input eight images such as: full face, eyes, mouth, and nose in profile and front view. It was obtained that the system eliminated the restriction for the input image has light, background, and exact position that other systems require. It is concluded that, by using an artificial intelligence algorithm as a means to select the user to whom access is allowed. On the other hand, it ensures to avoid and identification error because many times a particular trait of a person is usually identical or almost similar to another, but not his total face. If any other recognition algorithm were developed, it is advisable to take into account the response time of the algorithm; since when dealing with automated security, the important aspects are reliability and response time.