Video surveillance is one of the main components in any current Law enforcement investigation. Be it an ad hoc surveillance or just one gathered from public cameras (traffic, security, private, etc.) the power of video is greatly appreciated among agents and officers, as it makes their live easier and helps to close more cases. But not everything related to video-surveillance is as rosy. You still have to analyze the recordings. And that’s where the problems begin.

 

Generally, in this day and age Law Enforcement uses video surveillance as one of the main helpers in any line of investigation. The amount of IP cameras that are already on the streets helps a lot, as well as the quality of said cameras, which is improving steadily. If you have someone on video committing a crime, it is that kind of irrefutable evidence that will likely end with someone behind bars. But, as we have hinted in the intro, there are also downsides to having access to a video surveillance network.

The amount of video to analyze

First of all, the sheer amount of video to be analyzed in any modern case can be daunting. Especially, if that analysis has to be done by hand. Traffic cameras, security camera, both public and private…the amount of footage to review can easily mount into the hundreds of hours.

For this reason alone, an automatic review of video material is extremely useful, if not unavoidable. Having to review the past 12 hours from 150 cameras is just not feasible by hand. The amount of agents that have to be involved in this imaginary task would be unfeasible. And when something is found (IF something is found) it may be too late to use as actionable information.

Video Analytics in Law Enforcement: The number of elements to analyze

And then there’s the complexity of the material to analyze: number plates, faces, objects, relationships between suspects, cars…no one can be perfectly concentrated during hours watching footage, and not miss something.

That is where Video Analytics in Law Enforcement, also called VCA (Video Content Analytics) comes to the rescue. Thanks to the advances in artificial intelligence over the past decade, now it is possible to feed video into the different analyzers and let them detect exactly what you want. The can detect faces, number plates, vehicle models, makes and colors, as well as objects (i.e. a handbag, a hat, even a wrist watch).

You get an alarm when something interesting is found, based on what you have told the system you are looking for. In the meantime, you can do high value law enforcement work, instead of watching a screen. After all the video has been analyzed, you get a summary of everything that has been found.

Detect, track, use as evidence

But AI systems like Intelion can do much more than just save you from watching video on a monitor: it can track suspects o cars through time and multiple cameras, creating an effective map with timestamps in all the locations where the suspect has been sighted, and when. This allows to reconstruct movements before and after the crime. Combined with the ability to create relationships between different people, Intelion is a powerful tool for a Safe city. 

All the evidence can be exported individually or as a report to add to the case file, complete with screenshots of important moments or detections. Solve cases more quickly without having to put in the long hours of boring screen time, and go right to the interesting parts of an otherwise uninteresting recording.

Conclusion

Video Analytics in Law Enforcement allows  to analyze dozens or hundreds of camera feeds or video recordings at the same time 24/7, helps to advance a case even if you are off the clock. And if you need to find something more specific, you can run all the videos through the system again, with a different set of analyzers, if the case it so requires. Intelion’s flexibility is almost endless.