Typical percentage of motion/activity in various scenarios

[TheSauce 0]

I looked through forums here and Googled quite a bit and couldn't find anything on this. I am wondering if any of you have run across anything that gives "typical" motion/activity percentages for different installation environments. For example, something that says Airports on average have about X% of motion, schools have about Y% of motion, etc. Something like that would come in handy because I think a lot of times that motion/activity percentage is overestimated by a large amount when quoting a system which could result in you losing the job because you quoted more DVR's then the other guy or more storage.

 

What are your thoughts and if you have come across any white papers or anything like that dealing with this topic, please point me that way.

 

Thanks,

Sauce

[Soundy 1]

And how many angels can dance on the head of a pin?

 

The problem with trying to fill in something like "airports on average have x% of motion" is that there's nothing to base it on. First of all, what size airport? How many passengers do they actually move? What sort of time frame are you averaging on - per week, month, season, year? Are you accounting for high-traffic periods like holidays and long weekends? How about things like major flight cancellations caused by weather or other events, that can lead to more people simply milling about the terminal for extended periods?

 

And even if you can average that out, your individual installation will vary depending on camera placement, FOV, other non-passenger movement in the area (I've seen recordings run all night long because someone put up a hanging sign that triggered recording every time the HVAC kicked in - something that wasn't expected when the motion masking was initially set up). Things like headlights of cars going by at night, shining in through windows, can trigger recording as well - there's no way to average that across different sites.

 

If it's that big an issue to get it ACCURATE, it may be necessary to just monitor the traffic patterns yourself for a specific installation. Make up a mobile rig with a netbook and one or two IP cameras, set it up in the desired area with guestimate(s) of the desired coverage(s), and check the motion percentages for yourself.

[TheSauce 0]

I was hoping you would respond, Soundy. Your answer makes sense and I know that it can vary widely and maybe airports was a bad example. Schools was probably a better example, but I know that can very as well. I just see people sometimes calculating motion at a schools at 50% or more but in my experience, motion at a school doesn't usually go above 20% or 25% across a 30 day period (which seems to be what most end up recording for). I agree wholeheartedly that it's definitely better to overestimate and you can never be exact but if your customer is asking for 30 days and they end up getting 60 because you overestimated then they're likely to start questioning why they spent the money on twice as much storage (if doing a server/storage solution) or possibly twice as many DVR's if doing an analog solution.

 

I figured it was a long shot question to begin with. Maybe a better question should of been "how do you guys go about calculating how much motion a particular installation will see".

 

Sauce

[thewireguys 3]

Axis has a great calculator with example videos.

[jeromephone 6]

We find you almost have to do this after the fact until you get a history in that exact location it is hard to tell we have cameras in an ER area and they run 24/7 on motion. We have just kept enlarging the storage as cameras and the customer wanted higher resloution recordings. The good news is that storage is getting pretty cheap.