Critique my starter cameras/placement

[elwood]

My house is only accessible from one direction and one road which is around 120' from my house. For now i'd like to start small and eventually add a BI server with Deepstack and possibly integrate with my existing smart home devices. I'd like to get 2 or 3 quality cameras that would allow me to monitor and identify vehicles and/or persons coming onto my property with alerts. For now, i'd like to be able to use the built-in AI and SD slot until i can purchase and have a BI server running. My budget is around $500 for the initial cameras. I'd be willing to spend a little more if needed. I already have a sufficient PoE device and cat5e wiring.

I'm considering a 4231F-E2-M 3.6MM overview camera to trigger a 49225XA-HRN PTZ for tracking and identification, if possible. Both cams and a few accessories come in well under $600 from Andy. The PTZ would be installed on a corner bracket and the dual would be near it, possibly to the side or even directly above/below, haven't decided yet. I really like the idea of a PTZ with a large optical zoom, it's such a cool "toy" and i assume will work great if it performs as advertised. I know even the best implemented AI system will have a margin of error or fail completely so i will likely record 24/7, for sure when i get the BI server going.

Here's an overview of the area with the boob cams covering the entire entrance. There's lots of trees but they're trimmed quite a bit so i don't foresee too much trouble with the cameras seeing beyond them.

Let me know your thoughts and if you think i should do anything different or go with different cameras.

 

[looney2ns]

Study this: Cliff Notes

 

[Teken]

You need better lighting . . .

 

[elwood]

You need better lighting . . .

I have flood lights on dumb PIR's on each corner of the eaves on about a 15ft trigger. I would like to eventually use Frigate/Deepstack to trigger them when needed. I may also get some PoE IR's in the future. I need to get a couple cameras first to assess the need.

 

[sebastiantombs]

You're trying to cover too much with too few cameras. The two you show are fine for front door coverage, but expecting to get out to 120' effectively with them is not feasible. I have a similar situation, house about 120' from the road with one entrance and have a total of seven cameras covering the front. Even with that, solid identification everywhere is not possible.

 

[elwood]

You're trying to cover too much with too few cameras. The two you show are fine for front door coverage, but expecting to get out to 120' effectively with them is not feasable...

Thanks, i need to update the OP. After studying the wiki and various other research, i'll be using a minimum of 3 spotters + PTZ for front coverage. Just trying to come up with a plan to start small with better/best cameras and add on as the budget allows.

 

[Teken]

I have flood lights on dumb PIR's on each corner of the eaves on about a 15ft trigger. I would like to eventually use Frigate/Deepstack to trigger them when needed. I may also get some PoE IR's in the future. I need to get a couple cameras first to assess the need.

From a security stand point unless there’s a proven need to run stealth. The vast majority of installations benefit from lighting up the perimeter of the property. This serves several purposes as it relates to health & safety.

Health: Nobody will trip and fall on the property at night as its well lit. That’s great for the home owner and reduces possible insurance claims from you or others.

Safety: When an area is well lit that leaves very little room for someone to hide. Keep in mind there’s a difference between active vs passive lighting. Passive lighting is like curb appeal lighting that highlights the property and building structure.

This always on lighting enhances the homes beauty while offering critical lighting for the security cameras. This helps low light cameras to see better in colour while reducing motion blur and ghosting.

Of course the use of motion activated lights is important too! As this reacts to the immediate threat in the FOV. Once activated this tells the intruder they have been seen and the extra lighting will help the camera see better too.

The problem that normally arises is (IF) a camera is set to IR (Night Mode) and not properly calibrated to switch from night to day you can miss an event or the captured image will be distorted due to the IR lens switching over.

That’s why a properly deployed security video system has at least two cameras are locked in day vs night - at night. This of course affirms the importance of testing and calibrating the same to take into account smart IR and auxiliary lighting.

As auxiliary lighting can cause image blow out and defeat the whole purpose of video capture!

Solar powered lighting is so cheap now it’s one of the easiest things to add curb appeal to a home and property. While enhancing video security image quality it makes sense to use and deploy to reduce image blur and ghosting.

Solar powered lights come in all sizes, power output, and mounting options. From stick on / screw on / pole mount.

 

[elwood]

From a security stand point unless there’s a proven need to run stealth. The vast majority of installations benefit from lighting up the perimeter of the property. This serves several purposes as it relates to health & safety.

I'm leaning towards figuring out a way for Hubitat to automatically switch a camera to day mode when a corresponding light is activated. Not sure if there's anything in place to handle those commands but i will look into it.

Edit: looks like this will allow me to change the day/night profiles based on whatever condition i stipulate (I.E. on PIR#1 trigger, activate cam#1 day profile). Stability and reliability will have to be tested.

 

[Teken]

I'm leaning towards figuring out a way for Hubitat to automatically switch a camera to day mode when a corresponding light is activated. Not sure if there's anything in place to handle those commands but i will look into it.

Edit: looks like this will allow me to change the day/night profiles based on whatever condition i stipulate (I.E. on PIR#1 trigger, activate cam#1 day profile). Stability and reliability will have to be tested.

Just in time use takes lots of trial and error. Locking dedicated cameras to see in their respective day / night modes avoids missing an event because there’s no (IR) filter movement.

Obviously my reply is based on the latest 1 /1.8” ~ 1 .1.2” high performance low light cameras that can see in colour. This also applies to cameras that incorporate smart IR to avoid image saturation blow out.

If you have two cameras locked and well calibrated to operate in day & night. You’ll have a better chance in capturing details and objects moving quickly in and out of frame.

Keep in mind scheduling a camera to remain in a specific mode is the same as locking a camera in a day / night mode. The advantages of scheduling is the camera isn’t locked into that state 24.7.365.

Which also helps in avoiding IR switchover which 99.99999% of the time results in missing that money shot!