Author Gary Nakanelua

Video Analytics in the Public Safety Industry

April 24 2019 | Technology, Analytics, Predictive Analytics

In February, Blueprint was invited to co-present with Microsoft and GridMeNow at Microsoft Ignite in Washington D.C. Our session was focused around the use of data services and analytics in Azure Government. In the session, we showcased a scenario that highlights the use of video analytics for object detection across multiple videos using one of our products, Reflect. Object detection identifies and tracks specific attributes with powerful facial and object analysis algorithms.

The scenario we presented begins with an email being sent from one of our Operations Analysts to our IT department about a missing laptop:

We have badge access systems and surveillance cameras set up at the entry and exits for all of our buildings in the Bellevue, WA area. In addition, we have surveillance cameras in sensitive areas of our buildings, such as the file room noted in the email above.

The email above would typically result in a group reviewing days worth of badge access log data and watching surveillance videos from the multiple cameras to determine where and when the missing laptop was last seen (can anyone else relate to this grueling effort in their roles?).  In addition, this particular event would get a high level of attention and urgency, because the missing laptop belongs to Ryan Neal... the President of our company.

However, instead of having to go through the hours of review, we used Reflect: Blueprint's video analytics tool. By using Reflect, an automated mobile notification was sent out about an hour after the laptop was first reported missing. Here's what the notification looked like:

You may notice that in the notification above, the missing laptop isn’t shown. The reason is that Reflect has correlated the laptop to a series of objects, and tracked those correlated objects across various videos.

Although Reflect has sent an automated notification, it is intended to augment the investigation that people would conduct in a similar situation. Rather than having to manually review hundreds of hours of video footage, Reflect recommends specific video clips for the user and highlights the areas of importance. In the video capture below, you can see Reflect run its analysis and the intelligence that it provides to people involved in performing the investigation.

The initial assumption was a missing laptop but what Reflect highlights is that the laptop was always accounted for. Without prior communication, the Operations Analyst couldn’t have known the laptop was given to the proper individual, and she followed protocol accordingly. However, Reflect saved the IT team potentially days worth of time in their investigation and saved a group of people from a few sleepless nights.

While this is one simple scenario, perception analytics is industry-agnostic. You can learn more about the use cases and features Reflect can offer here.

Do you currently rely on video footage in your job to ensure safety? Security? What about customer trends?  Path tracking or zone detection?  We love diving into the challenges our community has and identifying ways to streamline or make them more efficient. If you have live video or archived footage that you'd like analyzed, we'd love to show you how Reflect works! Contact us below.

Ready to start your next big thing?

Contact Us