In the context of the rapid development of multimedia technologies, visual surveillance with traffic estimation and facial recognition is an important goal for many applications. The objective is to develop a human counting tool intended to provide statistical knowledge useful in objectively evaluating the efficiency of customer service in fast food.

The system will be able to accurately estimate the number of people passing through different areas and obtain an average, minimum and maximum value of time for serving customers at different times of the day or such information regarding different time intervals. Always, it will be possible to use in tracking mobile bots. The system will operate in a PC operating environment, being connected to a variable number of webcams in an Ethernet network.

Achieving the proposed objectives allows expanding the scope of use of several related fields, such as multimedia sequence image compression, video indexing for navigation, road traffic analysis, etc.