Face-Based Distributed Visitor Identification System

In this paper the problem of the expansion of the simple face recognition system using the distributed system architecture is considered. The basic principle of the simple real-life face recognition system assumes the operating in specific localization using a single camera. People's faces are detected first and then recognized. Taking as a basis a primary face recognition system for the task of visitor identification the extension to additional localizations is analyzed. In order to maintain the simplicity of the components the final Distributed Visitor Identification System integrates the results from the individual subsystems and provides the data exchange. As a result face-based people re-identification is obtained. The resultant solution, by linking the information between separate localizations, enables new applications like for example the reported people flow analysis.