If you want to really make it an automagic layout, you'd need to have a GPS chip, a compass/directional sensor, a tilt sensor, and some kind of electronic feedback from the lens to determine the FOV.
There is no doubt that mapping systems can be cool or helpful, but in reality they add little more than eye candy. Even if you have a camera covering 1000' of fenceline, simply naming the camera "Facility A, North Fence, Middle", or something similar gives the operators all the granularity they need position wise.
You also have to factor that any object in the FOV that you care about is going to be a moving target. So, unless you plan to also hook this to a missle launch system, having GPS-level granularity of object position yields data that is innaccurate before the operator even sees it. If you're using it to dispatch police/guards, it will be even more out of date by the time the guards are in a position to act on it.
This is something I put some thought and research into as well, but I'm not sure it's anything more than mrketing fodder at this point.
FLIR Security | 02/17/13 04:13pm
I would agree with Brian.... just because you can do GPS in surveillance cameras doesn't mean it adds any practical value.
...except maybe to help catch some of the Camover morons :)
Interesting idea. The principal(s?) of the company that built our current NVR system (Fast Video Security, AG, sold to Nice Systems in 2005) formed a new company, Fast Protect, AG, that does something similar.
Their Terra 3d product:
Terra 3d allocates to every pixel of a video image representing the ground real world coordinates!
| || || |
- Locate any object seen in video
- Measure distance between objects and heights in the video image
- Let a PTZ camera aim onto a location based on coordinates
- Enter the address of an object and let the camera aim onto it – available for fix, mobile or airborne observations
- Automated object classification based on object’s speed, direction towards north, and size is available
- Data fusion and correlation with other sensor types based on graphical inputs is availlable
I can see some use of the GPS enabled cameras in a City-wide video system that levereges privately installed video streams. It would permit the City system to determine in a flash if any there are any video resources that could be used for situational awareness and incident command as well as for forensic purposes; and it is a standardized format that anyone (or any system) can understand.
Jim - I agree that makes sense in theory.
However, I'm not aware of that many cities that have any kind of way to "register" cameras, and even at that I think that you could just as easily register your street address as having external CCTV cameras online. As I mentioned in my earlier post, without knowing full FOV data for a camera, a GPS position is no more valuable than a street address in that scenario.
Incident command brings into play a whole additional layer of complexity... Are the cameras externally accessible? What is port-forwarding infor for accessing them? Can you get directly to the cameras, or would you need the right VMS client? It's an interesting topic, but I struggle to think of many scenarios where the police wouldn't just show up on-site. Once at a specific location they might contact the building owner to determine if there is a way to see *interior* cameras, which likely wouldn't have been able to establish a GPS lock in the first place.
I can only see some worth for a GPS built-in, for when you have mobile camera's.
Like for dance festifals for example.