HSToday Magazine Interview with Eric Olson At GovSec
What are you displaying at this year’s GovSec?At GovSec this year we’re displaying our PureActiv™ product. It’s a full command and control, everything from point-and-click camera control to high end video analytics to alarm management/security policy. All based on GIS technology.
Can you tell us in general terms how it works?There are several things, first is it has a real simple point and click camera control. We take things from the video side and we equate it into the map. So every pixel in the video space is mapped into the corresponding point in the map space. This makes for a very simple user interface. You can point anywhere on the map, click on it and it will automatically drive the camera to that location.
We also do full video analytics, including behaviors and object tracking. We then take things that happen in the video space; detect the object, classify whether it’s a human, animal or a truck and actually put that information in the map space. So the idea is the person is not looking at a big wall full of video cameras, what they are actually doing is looking at the cue from the map and our system will present them the correct camera for viewing. Over time you realize that humans aren’t the best to watch video cameras, so the software does monitoring for you. It detects and then brings you the information to react to.
How have video analytics technologies evolved?Video analytics have been around for a while, but in the past they weren't very robust. They had problems with utilizing different cameras, different lighting, background movement, etc. Today the algorithms have come a long way. Now you can literally lock onto an object with the camera shaking on a pole, vegetation moving in the background, lighting changes and clouds. With just a couple of pixels the software can pick out the object, classify it, track it and take action on that.
The other thing it’s done is really moved into sensor fusing. In addition to the video detection, you can also take things like fence sensors, radars, any kind of third-party sensor and take all that information together and make a more educated determination of the alarm. So video analytics have come a long ways.
Can you describe your user base?We do a lot of work with ports, water facilities, etc. We tend to focus on installs where there is a harsh environment: trees moving, lighting changes and vast expanses causing difficult video situations. We own and develop our own video analytics algorithms, so it's very easy for us to adapt to these types of situations. In the past, customers had a hard time using analytics due to poor lighting, difficult situations or low bandwidth locations where it was difficult for a remote camera to provide real time data back to the user. Our software can be smart, via an edge device and help in those situations.
Are there infrastructure requirements to deploy PureTech Systems devices and software?We work with analog cameras or IP cameras. So if a customer already has cameras in place, they can upgrade and get the intelligence software. They don’t have to put in brand new cameras to do that. We work both ways from business model. We can provide hardware, but there's nothing special about the hardware. So if the customer already has hardware in place or has some kind of an agreement with one of the big suppliers, they can provide their own.
The one exception to that is we do provide an edge device. An edge device is hardware that resides at the edge and mounts right next to the camera. That is the one piece of hardware we do provide. The edge device is for situations where there is limited infrastructure or there is little to no bandwidth out at the remote site. The edge device can process the information out at the edge and take actions based on what's occurring. So without any human interaction it can follow targets, set alarms, trip gates, etc. Once bandwidth becomes available, it can send that information back to the user