U-Filter is an AI-driven video analytics add-on designed specifically for VMS integration. It is responsible for analyzing and filtering large quantities of video feeds and making sure that only relevant events generate notification on the VMS side.
IP cameras generate alarms for any type of movement U-Filter is responsible to minimizes false alarm rates and maximizes event detection accuracy.
Alarm Receiving Centers provide surveillance and response services to a multitude of clients simultaneously. Monitored sites consist of critical infrastructure, transport facilities, warehouses, venues and more.
Their first line of defense is a sophisticated surveillance infrastructure consisting of hundreds of fixed and PTZ cameras equipped with some form of motion tracker or radar providing live feed to a central VMS.
While such an infrastructure excels at motion and thus, event detection as a downside it comes with an excessive number of false alarms. Various factors can produce false alarms, such as different weather and light conditions, small objects, foliage, animals, plus unusual movement and objects.
As more and more false alarms are generated, the surveillance system eventually becomes unreliable and security personnel may not react to real events because of that. To ensure that false alarm rates are kept at a minimum, surveillance infrastructures needs to utilize the benefits of AI-driven video analytics. This is where U-Filter comes into play.
U-Filter extracts short videos from VMS and sends them to the U-Filter server for analysis. It analyzes all alerts produced by other systems, identifies the ones that truly require further involvement, and forwards the information to the VMS as a notification alongside the relevant video frames.