Neural network queue detector

Queue detector determines a queue of selected type objects in the controlled zone. The detector uses an object type identification algorithm based on high-accuracy neural networks. Neural network allows to categorize objects by type. This is a valuable feature that differs it from standard queue detector. A neural network detector can be configured to determine queue of the following types: person, car, bus, motorcycle, and bicycle.

This detector can be used to manage personnel or equipment operation for different service applications (retail shops, customs terminals, parking lots and etc.). Detector can be effectively used for transport and logistics, trading and any other application where needed information about the queue of any specific objects.

Using queue detector you can more effective manage retail space and employees schedule, optimize traffic flows etc. When several objects of specific type stay at a single point of the frame longer than a specified time, detector considers it as queue.

When system detects an object of specific type on the live video from camera, this object image on the screen will be marked with rectangular frame with its type and number in the queue.

The VideoNet system allows to pre-configure parameters individually for each of the detection zones of the queue detector, including entire image or selected image area.

You can configure detector for a specific camera and select type of object to be detected. You can select minimum size of the object to trigger detector. And select a number of objects which detector will consider as a queue and time since the queue have been occurred.

You can configure automatic response to the event, for example, system can activate recording of video and audio signal from the camera, send SMS-message or email, add record to video archive, launch application, produce audio alarm, generate report, send command to the actuating device or assign other event-based actions.