Pseudo 2D Hidden Markov Model and Neural Network Coefficients in Face Recognition

Abstract

For thousands of years, humans have instinctively used some physical characteristics (such as face, voice, posture, etc.) to recognize each other. About half the 800, A. Bertillon, chief of criminal identification section of the Paris police, plans to use some measures of the human body (height, length of arms, feet, fingers, etc.) to identify those responsible crimes. Towards the end of the nineteenth century, this original idea was further developed through the discovery (due studies F. Galton and E. Henry) the distinctiveness of fingerprints: they uniquely identify a person. Today, in full digital era, huge numbers of people use individual recognition techniques based on the identification of human characteristics, not only in justice but in civil and military applications. In fact, the only way to conclusively identify an individual is to recognize the personal characteristics. These are defined biometric features and, the technology behind this identification is called Biometric. The term Biometric, from the greek bios (life) and meters (measure), in computer sense, means the automatic identification or verification of the identity of a person based on physical characteristics and/or behavioral (CNIPA, 2004). Biometric feature is described as a physiological or behavioral characteristic that can be measured and subsequently identified to confirm the identity of a person. We can then divide the biometrics in:  physical biometric: it is that based on data derived from measurements made on a person's physical characteristics such as iris, fingerprint, facial features, hand or other;  behavioral biometric: it is that based on aspects linked to behavioral characteristics such as, for example, the issue of voice, dynamic signing, or the type of gait. As each biometric process starts with a preliminary phase called "enrollment" in which, generally, the person must provide the biometric system, through a sensor, its characteristic physical and behavioral, which is then converted into a mathematical model (template), two operating modes of biometrics are: 9