Computer Vision for Surveillance & Security Monitoring

Perimeter Surveillance & Security

By placing cameras around the perimeter of a structure/building, we will be able to monitor unauthorized entries. This could be people trying to scale the walls, enter through restricted gates, etc. We will also be able to monitor any unauthorized entry after closing hours and send instant alerts when it occurs.

Security Monitoring

We will be able to monitor the security of the building. This includes the manning of allocated security points. If at any point, these locations are not manned by any security personnel, we will be able to send instant alerts to the officials in charge. Additionally, the behaviour of the security personnel can be observed. For example, if a security personnel falls asleep on duty, the officials in charge will be alerted.

Surveillance & Alert system in Restricted Areas

We will be able to monitor unauthorized entries in restricted areas. If a specific uniform is required to enter, entering without the uniform will be considered a security breach, and officials will be alerted. We can set a predefined timeslot for entry; if anyone tries to enter outside the timeslot, officials will be alerted.

Threat Detection

CCTV cameras allow for the prediction of potential threats. For example, if a person is wearing a mask, officials will be immediately alerted. Any threatening behaviours will also be identified and reported. Arms & weapons such as guns, knives, etc., will also be detected, and alerts will be forwarded to the officials.

Traffic Surveillance

We will be able to detect all traffic violations through CCTV cameras installed across the city. This includes riding without a helmet, riding triples, one-way rule violations, riding on footpaths, etc. With ANPR (automatic number plate recognition), we will be able to tag the vehicle responsible for the violation.

Crowd Density Mapping

By placing cameras in threat-prone areas, we will be able to detect the crowd density and alert the police department if people group up during off-hours. The crowd density threshold, as well as the time, can be predefined.