Front door ID camera IPC-HDBW4231F?

[just some dude]

I still have to play around with it a little bit with that type of lens you get a lot of barrel distortion on the edges and I don't want that in the area that I'm mostly concerned with. I'll do some adjusting and post another picture.

 

[just some dude]

Ok, I didn't adjust the position yet, but captured a delivery person. Here are three images one at 16 feet away, 8 feet away and 3 feet away. Bright background, WDR set at 25, 3D NR 10.

Camera IPC-HDBW4231F

Lens 2.8 mm

DORI Specs for this camera:

Detect 44m(144ft)

Observe 17m(56ft)

Recognize 9m(30ft)

Identify 4m(13ft)

I was thinking a tighter view would be better, in hindsight, maybe a 3.6mm would have been better in this application. (I did test with a 3.6 and 6mm camera). I may purchase a 3.6mm and repurpose this 2.8 in a different location All in all, I'm happy with the overall installation but I am not easy to please!

Hopefully some of the more seasoned members can let me know if this is what to expect from this camera, and/or suggest changes that may improve the setup.

 

[SLC]

Ok, I didn't adjust the position yet, but captured a delivery person. Here are three images one at 16 feet away, 8 feet away and 3 feet away. Bright background, WDR set at 25, 3D NR 10.

Camera IPC-HDBW4231F

Lens 2.8 mm

DORI Specs for this camera:

Detect 44m(144ft)

Observe 17m(56ft)

Recognize 9m(30ft)

Identify 4m(13ft)

I was thinking a tighter view would be better, in hindsight, maybe a 3.6mm would have been better in this application. (I did test with a 3.6 and 6mm camera). I may purchase a 3.6mm and repurpose this 2.8 in a different location All in all, I'm happy with the overall installation but I am not easy to please!

Hopefully some of the more seasoned members can let me know if this is what to expect from this camera, and/or suggest changes that may improve the setup.

Seems pretty good to me. Looks like you have a good ID shot at 8 feet away. How high did you mount it?

 

[just some dude]

Seems pretty good to me. Looks like you have a good ID shot at 8 feet away. How high did you mount it?

See page 1 of this thread for a visual, 58" to center of camera off porch.

 

[tangent]

I was thinking a tighter view would be better, in hindsight, maybe a 3.6mm would have been better in this application. (I did test with a 3.6 and 6mm camera). I may purchase a 3.6mm and repurpose this 2.8 in a different location All in all, I'm happy with the overall installation but I am not easy to please!

Hopefully some of the more seasoned members can let me know if this is what to expect from this camera, and/or suggest changes that may improve the setup.

Seems pretty good to me, face is nice and clear. In most cases I think 3.6 is the better choice than 2.8, but in this case it's certainly debatable. A little distortion is typical at 2.8.

If your porch wasn't so well lit at night, the 3.6 could certainly have advantages as the narrower fov would result in less IR reflection.

 

[mat200]

..
DORI Specs for this camera:

...

Looks good, btw I prefer not using DORI specs from Dahua - it seems that they us a calculation other than 100 ppf for the ID distance.

 

[Philip Gonzales]

Detect (8 ppf)
Observe (19 ppf)
Recognize (38 ppf)
Identify (76 ppf)

https://www.google.com/url?sa=t&source=web&rct=j&url=https://www.inaxsys.com/media/files/element/The_DORI_Distance_Inaxsys2017.pdf&ved=0ahUKEwi7koXO7uzXAhVW1IMKHUPfCkoQFgglMAA&usg=AOvVaw2TerG3-2TB5SC8RF0FjCCn

 

[mat200]

FYI -

Axis has a nice description, and clearly the numbers above seem to allow greater DORI distances, compared to their stated preference.

note - Using 100 ppf gives use a useful point between the 76 ppf and 150 ppf.

"One reason why we have chosen to use the face is its distinct identifying features. Furthermore, the variances in face widths are less than those of body lengths or widths, which results in a smaller margin of error. The average human face is 16 centimeters wide (= 6.3 inches wide). Following suggested operational requirements from SKL, the Swedish National Laboratory of Forensic Science, and supported by our own test results at Axis Communications, we have chosen to use 80 pixels [ 80 pixels across the face - or about 150 ppf ] as the requirement for facial identification for challenging conditions . (see Figure 3).

To some, this number might sound high, and in fact some vendors or independent sources recommend 40 pixels for a face or 100 pixels per foot for recognition. The argument behind the higher number is that for identification, there are limited other telltale signs. For recognition, previous knowledge adds factors such as how a person moves – a property easy to observe and recognize, but difficult to identify and describe accurately. To ensure sufficient video quality even if the object isn’t facing the camera straight on, or if the lighting is not optimal, the higher number provides an adequate safety margin."

Identification (challenging conditions) 80 px/face 5 px/cm 12.5 px/in [ this is 150 ppf, this is 80 pixels across the horizontal of an average face ]
Identification (good conditions) 40 px/face 2.5 px/cm 6.3 px/in [ this is 75.6 ppf ]

https://www.axis.com/files/feature_articles/ar_perfect_pixel_count_55971_en_1402_lo.pdf

 

[flynreelow]

Clean Install--

what I would really like to to, is convert the flat part to the left of the lens, and somehow install a wireless doorbell button. I know I could just stick one on, or mount to the outside, but i think it would be nice if I could hide the guts under the casing. looking at your plcture, not sure how much room there is with the casing off? do you think one would be able to house a small battery under the casing?

edit.. looks like this is located much higher than an average doorbell. at doorbell level, would you just not get a good facial recognition?

 

[Chase]

Great install

 

[mat200]

Clean Install--

what I would really like to to, is convert the flat part to the left of the lens, and somehow install a wireless doorbell button. I know I could just stick one on, or mount to the outside, but i think it would be nice if I could hide the guts under the casing. looking at your plcture, not sure how much room there is with the casing off? do you think one would be able to house a small battery under the casing?

edit.. looks like this is located much higher than an average doorbell. at doorbell level, would you just not get a good facial recognition?

Hi Flynreelow

I was also thinking about mods for this camera ( Dahua mini-dome wedge ipc-hdbw4231 )

There's really not much available space under the case, so if you want to do any modding you would have to use very small components.

 

[just some dude]

As far as mounting a switch for a doorbell inside, you could but I wouldn't.

Imagine every time someone goes to push the button all you see is a hand and arm! No thanks.

Mounting height- I would get similar ID results if the camera was + or - a foot from where it is.

 

[teamRAVEGREEN]

Good thread and great install @just some dude !

I've been prepping things around my house for some installs (mostly network and BI stuff) and tonight putting in a little more time thinking about camera placement. Originally I was thinking I would use a 5231 for my front door (ID) camera, but it is pretty big. I think I could still make it work, but I could also for sure use that 5231 somewhere else at my house.

So I came across this thread and I'm thinking I could MAYBE get away with 4231 at my front door, near the door bell because it is mostly shaded and pretty well covered up. I'll have to see about running cable to the location but I wanted to run the idea by you or anyone else on this thread. Front door faces SW, so maybe a little sun light but full Westward facing sun should be mostly blocked by the alcove I think.

I'm assuming that I could temporarily mount one of the 5231's on a mobile stand to test placement (as suggested on these forums) and because that camera is varifocal I could test at 2.8 and 3.6 to see which is better in my situation, is that correct? I mean, it should give me a pretty good idea which I would want?

I'm also just wondering if or how soon I'd regret a dome here...

 

[tangent]

Good thread and great install @just some dude !

I've been prepping things around my house for some installs (mostly network and BI stuff) and tonight putting in a little more time thinking about camera placement. Originally I was thinking I would use a 5231 for my front door (ID) camera, but it is pretty big. I think I could still make it work, but I could also for sure use that 5231 somewhere else at my house.

So I came across this thread and I'm thinking I could MAYBE get away with 4231 at my front door, near the door bell because it is mostly shaded and pretty well covered up. I'll have to see about running cable to the location but I wanted to run the idea by you or anyone else on this thread. Front door faces SW, so maybe a little sun light but full Westward facing sun should be mostly blocked by the alcove I think.

I'm assuming that I could temporarily mount one of the 5231's on a mobile stand to test placement (as suggested on these forums) and because that camera is varifocal I could test at 2.8 and 3.6 to see which is better in my situation, is that correct? I mean, it should give me a pretty good idea which I would want?

I'm also just wondering if or how soon I'd regret a dome here...

4231 just like 5231 can refer to multiple cameras.

The IPC-HDW4231EM-AS is a fixed focus turret that's a little smaller than the varifocal. The IPC-HDBW4231 is the 'wedge' aka mini-dome style of camera that this thread is primarily about.

 

[teamRAVEGREEN]

I see, thanks My post was a big long-winded too...

Basically I was trying to decide two things:
> First - if the wedge/mini-dome that @just some dude installed would work in my situation (pic above) or if I will regret trying to use it outdoors?
> Second - just looking for confirmation that I should be able to use my varifocal (HDW5231R-Z) to test positioning and for a fixed camera if 2.8 or 3.6 would be better for my situation - or am I missing something?

 

[tangent]

I see, thanks My post was a big long-winded too...

Basically I was trying to decide two things:
> First - if the wedge/mini-dome that @just some dude installed would work in my situation (pic above) or if I will regret trying to use it outdoors?
> Second - just looking for confirmation that I should be able to use my varifocal (HDW5231R-Z) to test positioning and for a fixed camera if 2.8 or 3.6 would be better for my situation - or am I missing something?

I'd put is closer to the door and not right above the doorbell button, less sun and you don't create a blind spot. It's IP67 rated, so it's fine outdoors. Direct sun will damage the dome some overtime.

You can use your varifocal camera to get a sense of which size lens you get, but it may not be exactly the same. I'd tend to go for the 3.6mm to get a little more detail and less fish eye distortion

 

[teamRAVEGREEN]

Depth in that pic is not great to get a sense for where I was trying to place the dot. I'm definitely going to test as close to the white trim around the door as possible - not directly above doorbell. Thanks for your reply!

 

[just some dude]

teamRAVEGREEN, I would repeat my install in a heartbeat! I'm still on the fence whether 2.8 or 3.6 is best for my location I'm leaning towards 3.6 as mentioned for better detail and less fisheye. No camera mounted in that position is going to give you a view of packages sitting at the door and good face shots at the same time in my opinion. I'm confident the mini wedge camera will have longevity in my location. If sheltered in yours also. All that said I would go 3.6 in Corridor mode for ultimate ID. You may have to run wide dynamic range on a schedule if bright background.

 

[looney2ns]

You could always mount the cam in the center of the doors, on the ceiling, that's only a tad over 6ft, run it in corridor mode and you would get face ID's as they come up the stairs, and packages on the porch. Use a 3.6mm.
Yes, you can use a Varifocal to give you a sense of how it would all work out.
It's always best to test a spot at least 24hrs or more before drilling holes.
Best laid plans on paper don't always pan out.

 

[teamRAVEGREEN]

Hey @looney2ns my front door is over 8' tall!

I'm going to go under the house this afternoon I think and I'll take a closer look at the doorbell wall. I don't recall how much the concrete slab interferes. Thanks for the advice on the optics and time for testing.

PS - I love your avatar!