Genetec Visual Tracking

We have used it quite a lot, and found it to be very useful (especially when you have guards unfamiliar with the camera layout, or need to act quickly without thinking of what cameras go next). What kind of application are you looking at with it?

We show the basics of Visual Tracking at about 3:45 in this video:

It's fairly simple to set up and use, though it is definitely additional configuration work.

What do you want to know about it?

Thanks Ethan and others. This feature was being sold to my client as a facial recognition solution that would detect a known person and automatically follow them around the facility. The client is not actively monitoring their video and I don't see anything in there about facial recognition. Further, as it was described to me, it sounds like a fairy tale. Also, it would appear that if you are not actively monitoring the system, the feature isn't really all that useful. Agree?

There may be some biometric analytics (algorithms) involved in the back-end, but to say it is a facial-recognition solution is too far a stretch.  

I'm unfamiliar with Genetec's implementation but it seems somewhat similar to the camera navigator feature in Milestone. There, you can define your cameras on a map, and the relative positions and orientation of the cameras on the map are reflected as icons overlayed on the image to ease the navigation between nearby cameras. But there's no analytics involved and following someone is still a manual process, aided by visual cues/links.

One solution I'm still impressed by is from Snap Surveillance who have developed an analytic which intelligently maps out relationships between cameras, making it even easier to identify which cameras to jump to when following a subject. Not only do they automatically learn the relationships between camera views over time, but they enable quick and easy exporting of a "tracking session" which results as a seamless AVI cutting from camera to camera exactly as the operator did, like a storyboard.

They have a Milestone integration of course, but if they don't already also have a Genetec integration I imagine it may be nearly trivial to do so. It's certainly worth looking into anyway.

Yes, it is pretty much a way to navigate between cameras - but it has to be manually setup and defined when setting up the system, no analytics involved. You are able to draw zones on a camera view and link certain cameras to them, so when viewing video (live/playback) you can quickly jump to the next camera without going through a tree view or map display or something like that.

It sounds like that snap surveillance you described except you have to manually define each relationship. They did recently add the export feature where you can "record" the tile view of the operator and get a single spliced video file with all the camera views, one after another, as they were played back by the operator and following a subject.