Real-Time Face Recognition from Surveillance Video - Intelligent Video Event Analysis and Understanding - page 193

Appendix: Further Implementation Details of the Experimental
System
The system used to conduct the experiments in this chapter was written in C++ us-
ing the O PEN CV libraries. The source code is available for free download from our
website ( http://www.electriceye.org.uk ), under the terms of the GNU
General Public License. The website also has some short videos illustrating our
experimental results.
An overview of the system architecture was described in section 2 and the rela-
tionship between front-end camera nodes and the back-end database and classifier
was shown in figure 2. Figure 31 shows a more detailed view of a single node and
its communication with the back-end. The node detects faces and extracts a feature
vector for each face. Object tracking could be carried out by the nodes, so that multi-
ple face samples can be tagged with an identifier, thus attributing the samples to the
Fig. 31 Distributed System Architecture: Communication between Front-End and Back-End