Protect supports line crossing for Unifi cameras. I haven't tried it myself, but apparently latest version of Protect only supports ONVIF events for Motion Detection, and not Line Crossing, etc. It looks like someone from this forum should look into this, the "normal" Unifi users are not good in...
Guys, you are just a bunch of whiners. Dahua is, basically, a market leader, there are no real alternatives for their cameras.
So, logically, they want to use this position to make more money.
ONVIF Improvements:
Easier migration at scale
Motion/analytics support coming
Better than prior RTSP-only integrations
Also
AI Port X: 20+ streams per port
Check this thread out https://ipcamtalk.com/threads/a-healthy-competitor-g6-pro-turret.84296/ .
Even 1/1.2 G6 Pro looks worse then latest Dahua cameras.
Ubiquiti UX might be better for some people, but video quality, especially in low light conditions is just bad.
54PRO is much better than Reolink, trust me... :) It only looks bad when you compare it to 54IR, comparing it to all other cameras on the market is a differnt story.
I can definetly say, for example, that 54PRO is much better than G6 Pro Turret.
I think there are use-cases for both of them. I'm currently using 54IR and 54PRO, it really depends on your requirements.
Not everyone who is using the cameras report crimes to police. In many cases simple observation and early warning about some movements us more than enough.
For example, in...
BTW, G6 Pro Turret is huge, I think any varifocal 1/1.2 gonna be about the same size, it means it gonna be new mounting design :(
TBH, I would rather had 54PRO with 5442-S3 sensor.
It's... complicated
https://help.ui.com/hc/en-us/articles/28315005177239-Protect-AI-Port-FAQs
Notice "up to" everywhere :)
But, basically, it only supports one 5442 4MP stream.
I have one, tested, and can confirm.
Couple of comments:
1. It "just works" until it doesn't. When I'm testing G6 Pro vs S3 I can see how many events Ubiquiti is missing. If you don't have anything to compare with, it all looks really nice and cool, because you just don't know how it should work :)
2. BI is not for average user...
And even more non-helpful response, I'm experimenting with Google Gemini AI, asking it to analize my camera pictures, here is what it said about those two photos:
Prompt: Can you see any human or animal on the photo?
First photo: "Yes, I can see an animal in the photo.
There is a...
Everithing looks correct and the same for all cameras. It's defintely about moving shadows and branches. And as I said, it's usually happening on really windy days, when the patterns are prety chaotic.
As far as I understand this the result of using the "small" model, and that's why Dahua trying...
In general, I don't have too many false positives, I have ~20 S3s and only maybe 2 or 3 of them, located is most problematic areas, are having these issues.
But it's still annoying, that's why I looking for something better :)
A good example of the false positive scene is sunny and windy day in the forest. Tons of moving shadows, costantly chaging their shapes, sooner or later triggering camera IVS algorithms.
And it doesn't matter what I'm looking for, even if it's just a vehicle detection :(
My money is mainly on S3 at the moment :) Everything else is just a curiosity, tbh.
What I really want is S3 with better object detection. It's not bad as it is, but I'm getting some annoying false positives.