Signal and Image Processing for Biometrics

Chapter 2Introduction to 2D Face Recognition ¹

2.1. Introduction

The ability of humans to recognize faces among others, and the ability to deal with gender, aging, and expressions, shows how extraordinarily the brain processes the information.

In computer visualization programs, the challenge of designing face recognition systems is to outperform, or at least equal, the abilities of the human brain. Within this context, face recognition methods and systems (Figure 2.1) are rapidly increasing, particularly through the development of information technology systems. Consequently, a large number of applications have been developed in various areas such as access control, video surveillance, analysis of facial and emotional expressions, and plastic surgery.

When considering security application, in particular for the identification of individuals, the performance of face recognition cannot compete with other methods such as those that use fingerprints or the iris, because facial features change over time, due to either voluntary or non-voluntary reasons (e.g. aging, gain/loss of weight, expressions, emotions, beards, makeup, and glasses). Moreover, face recognition performances are very sensitive to data acquisition conditions (e.g. illumination and pose).

In comparison to other methods, face recognition has some main advantages, such as good acceptability, low cost and contactless acquisition (i.e. using a camera).

Figure 2.1. Face recognition system (Zalix biometrics authorization) ...