Synapses by Cias classifies moving objects through Deep Learning video analysis, dynamically adapting the thresholds digital sensors. The field of view of the single connected camera can be divided into 8 freely configurable areas and can be combined with 32 field sensors. Synapses is able to distinguish within the video stream (stream via RTSP IP) the presence of different targets called "classes": man, animal, car, truck and other important targets that can generate alarms.
This information is sent to the connected sensors which will change the previously set detection thresholds to calibrate them accordingly. The result of this neural connectivity is an enhancement of detection rate along with a higher discrimination of false alarms. Synapses introduces a new concept in the intrusion detection sector, smartly connecting PIDS to Video-Analysis to provide enhanced perimeter protection as an outcome.