2 User identification
2.1 2D face and 3D face recognition

Face recognition is a subset of large field of pattern recognition research and technology. Human face recognition has become one of the most important biometrics authentication methods in the past few decades, due to its potential for a wide variety of applications and areas (surveillance, home security, border control etc.). Biometric systems for personal identification, which are developed by several vendors, achieve very high face recognition accuracy. The most of these applications require [1]:

The main advantage in comparison with other approaches is that the face recognition does not require any voluntary action by the user since face images can be acquired from a distance by a camera. Next advantage is that the acquisition devices are cheap and are becoming a commodity.

The main drawback of the face recognition is its current relative ease with which can be defeated.

However, in comparison with speaker recognition, face recognition achieves much better results. In general there are three main approaches based on type of data which are used in the recognition process. We know methods based on 2D intensity image, 3D facial data and the technique which is using both types of data. The whole process of recognition consists of 3 main stages. The first one is acquisition and pre-processing, the second one is data registration and the third stage is recognition. Detail description of 2D and 3D face recognition can be found in module UserIdentification.

In modern systems multi-face recognition can be implemented. In that case all faces in the images are detected. Positions of detected faces are associated with the image. Next, the image is split in to several samples based on the positions of detected faces. These generated samples represent all faces in the scene and allow to track each face separately. The image split is a main step towards multi-face recognition within the systems.