On the use of the Kinect sensor for human identification in smart environments

This work is framed into the context of ambient intelligence applications, a general definition which usually refers to applications aimed at embedding technology in the environment and enriching it with the capability of recognizing people and adapting itself to user specific needs and preferences. One basic requirement for this kind of applications is unobtrusiveness: user recognition and environment personalization should not be perceived by the user, performed while he/she performs normal daily activities. Biometric recognition is particularly suited for this kind of applications, provided that suitable charac-teristics are exploited. The aim of this paper is to evaluate, in this application scenario, the possibility of performing automatic person identification using data collected by the Kinect sensor. In particular, starting from a sequence of frames, face and anthropometric measurements are considered for recognition of individuals. Extensive experiments show the feasibility of such technology for ambient intelligence applications.