THE MODELING OF FACIAL RECOGNITION PROCESS IN PROSPECTIVE OF SIMULATION TECHNIQUES (A methodical elaboration through the built-in modules of Matlab)

Abstract

Biometrics is a field of study and dissemination that brings a host of contributing domains in an integrated way to generate products that find multiple applications particularly in creating an individual's identity. Modeling biometric processes reveals a series of discussions in various fields, mainly in medicine and in the digital image processing field. These discussions range from the simulation levels of these models to those of application and concrete products that use biometric systems for identifying and recognizing individuals. Indeed, over many biological parameters of an individual to be identifiable, there is the face that generates a lot of information about the physical data of the individual. According to specific studies, facial components combined according to different methods can be used to design simulation models that come in aid of generating closed-source and dedicated programs of biometric recognition (Das, 2018). This paper deals with a simulation model, through the Matlab test-bench, of the facial recognition process. The model is created at the proposal level for the subject field and gives the limits and recommendations for improvements and further developments of this simulation method. The modeling methodology includes two parts; the first part is the Matlab program that has been coded in such a way that it takes some images that represent the faces of different people published in the opensource database on the Internet; the second part is the creation of a database that retains these images to enable face identification when entering a new image program. The facial search and retrieval process are built through some built-in functions that the latest versions of Matlab offer. Finally, the simulation model provides a new facial identification model that leaves room for further discussion and limits to the relatively small number of face images that can support the database in Matlab.