How AI NVR Can Identify Potentially Dangerous Objects in Train Stations

On August 16, 2019, two rice cookers were reported as suspicious objects in a subway station in Lower Manhattan and near a garbage can in Chelsea, New York City (NY Post). All were eventually deemed safe and non-explosive, but surely caused an unnerving sense of panic for the NY police, street pedestrians, and subway passengers. Witnesses, such as the ones who reported the suspicious activity, are encouraged to continue to relay such information to prevent catastrophes. With that said, for maximal public safety without delays and inconveniences, what can be done to make the investigations even more efficient?

First, AI NVR’s Face Recognition can identify a dangerous criminal/terrorist in publicly available wanted lists. By simply uploading images of suspicious individuals to the database, AI NVR will then alert the security authorities if that person appears on premise. In this scenario, if a suspicious individual indeed left a potentially dangerous object at the subway station, he or she could have potentially been identified in real time to prevent the anticipated disaster. Face Recognition can also be used after the fact. For instance, after the person who placed the rice cooker in that location left the scene, the police can upload their image, search for any sightings on other cameras to track and determine whether they proceed to act in a suspicious manner. One may even track a suspect back to his or her car, the vehicle can then be identified on cameras based on its color and possibly license plate number. The vehicle itself can then be tracked across all traffic intersections in the city using the AI License Plate Recognition function.

Video Search is another of several tools in AI NVR to help speed up the investigation in situations like these. This feature allows people to search for and be alerted of objects of interest, such as a backpack, handbag or suitcase in which the cooker (or weapon) was transported. Without manually viewing hours of videos, the security staff can easily search for the object dropped off at that camera, identify when and by whom it was dropped off, and track the person using such criteria as “person wearing blue with a black backpack” across all cameras within, say, a 10-mile radius. This whole process can be accomplished in less than 5 minutes if not seconds with AI NVR.

While there is no way to fully eradicate false alarms and/or false claims, AI NVR allows for a drastic reduction and a much easier way of finding a suspect. AI NVR pin-points faces, plates, objects, vehicles, etc. so that you don’t have to. The time and energy saved, as well as the overall efficiency of AI NVR may be lifesaving in comparison to the manual labor and time that it would take to find the suspect of a particular scene without the technology.