What's the Best Low Light IP Camera

[armstrong531]

I need two low light spotter camera for a rather large area. Not looking to recognize faces but looking to pick out details of a human from about 75 to 100ft in order to trigger the PTZ systems.

 

[wittaj]

The best low light camera is the 4K/X that is on the 1/1.2" sensor.

But even with that, a 2.8 or 3.6mm may struggle at being able to AI a human that far out at night unless you have some good light.

Your better bet would be to get the next best low light camera from the 5442 series like the Z4E and optically zoom it closer to get a better chance of it being able to recognize a human to swing the PTZ.

 

[The Automation Guy]

The best change of getting good quality low light footage is to use a large sensor with a small amount of pixels (low resolution).

We have a set of "minimums" that we like to suggest on the forum....
a 4k (8mp) camera should have a minimum sensor size of 1/1.2"
a 4mp camera should have a minimum sensor size of 1/1.8"
a 2mp camera should have a minimum sensor sizze of 1/2.8"

Any camera with those specs will have about the same low light performance. All things being equal, you will want to get the camera with the highest resolution however because that will provide the most detail which is why wittaj correctly suggested the 4K/X camera which is a 4k camera with a 1/1.2" sensor (and full time color to boot). Of course this is a huge generalization because the quality of the camera, sensor, lenses, etc and the proper choice of focal length for the situation will also play a factor in the overall performance.

 

[Teken]

I need two low light spotter camera for a rather large area. Not looking to recognize faces but looking to pick out details of a human from about 75 to 100ft in order to trigger the PTZ systems.

This is the wrong technology to use and follow.

The correct method is to install a sensing grid or sensor(s) that use ground loops, microwave, PIR.

The idea is simply tell the PTZ something has entered a zone / region and the above if integrated to the same can have the PTZ move to a preset - once triggered. When thermal / radar / microwave is deployed correctly it easily provides that trip wire to inform a PTZ to move to target.

 

[wittaj]

This is the wrong technology to use and follow.

The correct method is to install a sensing grid or sensor(s) that use ground loops, microwave, PIR.

The idea is simply tell the PTZ something has entered a zone / region and the above if integrated to the same can have the PTZ move to a preset - once triggered. When thermal / radar / microwave is deployed correctly it easily provides that trip wire to inform a PTZ to move to target.

If someone owns the entire property, then maybe yes this expensive approach is the more fail safe option.

But for most homeowners that are tying to get the PTZ to be looking the correct way for someone walking down the street or the public sidewalk, sensing grids and ground loops, etc. aren't feasible.

Many of us here, including myself, have had tremendous success using other cameras as spotter cameras.

Heck I have a $40 cheapo overview camera that works just fine as a spotter camera. The camera is garbage to identify or make out anything, but to get a clean enough outline of a person that DeepStack recognizes as a person and spins my PTZ, it works great. It never misses, so why spend more than needed?

 

[Teken]

If someone owns the entire property, then maybe yes this expensive approach is the more fail safe option.

But for most homeowners that are tying to get the PTZ to be looking the correct way for someone walking down the street or the public sidewalk, sensing grids and ground loops, etc. aren't feasible.

Many of us here, including myself, have had tremendous success using other cameras as spotter cameras.

Heck I have a $40 cheapo overview camera that works just fine as a spotter camera. The camera is garbage to identify or make out anything, but to get a clean enough outline of a person that DeepStack recognizes as a person and spins my PTZ, it works great. It never misses, so why spend more than needed?

Just to be clear for those who may read my reply and better understanding as to Why is based on real world experience, use, and deployment. In a residential area your method works quite well and I have used the same approach when it met the goals.

But if the OP has a large area say a farm land or acreage and the area exceeds 150 feet. The most effective and long term success comes from using ground sensors. Next is the use of thermal imaging vs PIR as the false free units costs a lot of money.

Obviously, on the highest end is the use of radar which lots of companies sell including both Dahua / Hikvision.

The benefit of radar sensing is it can track multiple targets at once and provide the speed, height, and distance. No other sensing technology (consumer grade) sensing can provide all four attributes. The conversation should be from the OP as:

I have an area that is 1000 feet wide and is 300 feet away from the PTZ. This area has X (trees, bushes, buildings, roads, polls, etc). I am seeing Y happen every week at this time / happens randomly

The use of sensor(s) are termed COT Close to Target and generally speaking are used to define a outer perimeter. This provides several things which is awareness but the most critical - time. More serious installations will have tiered rings or what some call Breach Points.

Breach points can vary based on the needs of the person / installation to allow reactive force / counter measures to be activated.

Generally speaking, if people want to keep this simple perimeter lighting is the most cost effective security tool to use. As it provides COT lighting, enhances video capture, increases health & safety to be able to see ones surroundings. Any fool can install a pole with a solar LED fixture with a PIR and just a single unit will offer ten times benefit over solutions.

Again, this comes down to understanding the environment, threat, and expected goals . . .

I've seen multi-million installations install some of the best in class video camera's only to fail to meet the goal.

Why???

No fencing to define the perimeter and keep the casual pedestrian . . . No lighting so those really expensive camera's where ghosting images all night long . . . No sensing of any kind from infrasonic, ultra sonic, microwave, radar, PIR, or ground loops . . .

We come along and install two miles of fencing and create a berm and guess what?!?

98% of all incidents stop - why??

Whelps, if the chicken can't cross the road - there's no chicken!

 

[Flintstone61]

I'll try and slip this thru the board meeting... 1/2 of them didn't know what Catalytic converter was. or where to find it on the car. Of course they all park inside. So as soon as I start the techno speak they gloss over.....so getting PIR, Radar and Grid stuff....won't pass the budget commitee. Fixed cam's seem to be the best at the Condo anyway. But I did dig myself a hole and buy a 49425 PTZ 2 weeks ago.