NVR LTN8608D-P8 Deep Learning Test

Eggin

Eggin

n3wb
Mar 15, 2019
5
0
Houston, TX 77063, US
I volunteered to come up with a security camera plan for our Property Owners Association in Houston TX. After dealing with my own cameras and researching this project I find myself addicted to surveillance systems and how A.I. can help prevent crime before it happens.

I am breaking this project into two phases. Phase one focuses on setting up the home owners with their own surveillance using some type of A.I. to prevent false notifications and provide instant push notification to their phones.

The second phase for overall neighborhood security will probably be based on dozens of LPR cameras.

For my first test I am looking at purchasing an LTS Deep Learning NVR model LTN8608D-P8. I want to be able to detect Humans entering the property, loitering on the sidewalk, and cars entering the driveway. Haven’t decided on a camera but it will probably be a LTS 4mp with a varifocal lens.

The point of this first phase is to give homeowners as much time as possible to prepare for a intruder. False notifications have to be kept to a minimum.

Any input on the LTN8608D-P8 would be a great help.


Thanks.
 
Eggin said:
I volunteered to come up with a security camera plan for our Property Owners Association in Houston TX. After dealing with my own cameras and researching this project I find myself addicted to surveillance systems and how A.I. can help prevent crime before it happens.

I am breaking this project into two phases. Phase one focuses on setting up the home owners with their own surveillance using some type of A.I. to prevent false notifications and provide instant push notification to their phones.

The second phase for overall neighborhood security will probably be based on dozens of LPR cameras.

For my first test I am looking at purchasing an LTS Deep Learning NVR model LTN8608D-P8. I want to be able to detect Humans entering the property, loitering on the sidewalk, and cars entering the driveway. Haven’t decided on a camera but it will probably be a LTS 4mp with a varifocal lens.

The point of this first phase is to give homeowners as much time as possible to prepare for a intruder. False notifications have to be kept to a minimum.

Any input on the LTN8608D-P8 would be a great help.


Thanks.
Click to expand...
It will not be reliable.
 
Eggin said:
Any thoughts on what would be reliable?
Click to expand...

At this time I would think something that is simple, robust and does not rely upon AI. Despite all the marvels of AI, we're just not there yet.

As for detection of humans entering the property: you need an alarm system. Magnetic contact switches on doors/windows and glass break and PIR movement detectors inside. If only one contact triggers an alarm it's probably nothing but if you have both a contact+PIR triggering the alarm there's very likely someone inside.

If you feel confident about not being hacked you could use indoor cameras for verification, although you have to realize you might not be able to do the verification 24/7 if you're self-monitoring.
 
How many people are in your association? You might want to look into contacting an actual security company and see what they can offer for how much. if you're doing this as a group it might be quite cost-effective. If you're with enough people you might even be able to get them to patrol your area at night. Or at least they could perhaps monitor your dozens of outdoor cams. That will be much more reliable than AI.

I'm wondering actually why you'd need dozens of LPR cams, do you really have that many entry/exit points? I'd think capturing license plates there might be enough, other cams wouldn't need to be LPR or am I missing something here?
 
Hi Eggin,
This is Jack Chan, I am the support manager here in LTS. I have taken over this forum.
How's the project going? Do you need any additional help?
Jack
 
You must log in or register to reply here.
Top Bottom