I first started with CPAI and it worked well for me. I then converted to using the onboard AI detection on my Dahua 54IRs to detect People, Vehicles and Animals. I've been pleased with the Dahua AI. I thought I'd miss the ability to detect different types of animals or vehicles, but I don't. How about any others, have you found one more accurate than the other etc? Thanks
For those that have used both... Is onboard AI detection more accurate than CPAI?
- Thread starter 105437
- Start date
I have found the AI of the cameras to be superior for my needs. I ran a lot of side-by-side concurrent testing and the camera always won.
Plus it is one less thing to worry about breaking. I have to restart CodeProject every few days as it just poops out for no reason that I see. And what if CodeProject quits being supported or BI moves to another program.
While some of that third party stuff is cool like tagging was it a dog or a bear, I don't need all that fancy stuff. If my camera triggers BI to tag an alert for human or vehicle and BI can accomplish what I need by way of a text or email or push or whatever, that is sufficient for my needs. I just want to be alerted if a person or vehicle is on my property and the camera AI does a fine job with that.
However, I do run BI AI on a few cameras so that it knocks out headlight shine so that the alert image includes the vehicle. The camera AI will trigger for a car, but the alert image was always just the headlights. I also run the ANPR AI module.
The true test....I have found the AI of the cameras to work even in a freakin blizzard....imagine how much the CPU/GPU would be maxing out sending all the snow pictures for analysis to CodeProject LOL. My non-AI cams in BI were triggering all night. This picture was ran through AI (without the IVS or red lines on it) and it failed to recognize a person in the picture, but the camera AI did. This pic says it all and the video had the red box over it even in complete white out on the screen:
See this thread on how using just Dahua AI may be sufficient for your needs (and other cameras with AI would perform similar):
Who uses Dahua AI capable cameras? Reliable AI for triggering events? Pro's/con's?
Plus it is one less thing to worry about breaking. I have to restart CodeProject every few days as it just poops out for no reason that I see. And what if CodeProject quits being supported or BI moves to another program.
While some of that third party stuff is cool like tagging was it a dog or a bear, I don't need all that fancy stuff. If my camera triggers BI to tag an alert for human or vehicle and BI can accomplish what I need by way of a text or email or push or whatever, that is sufficient for my needs. I just want to be alerted if a person or vehicle is on my property and the camera AI does a fine job with that.
However, I do run BI AI on a few cameras so that it knocks out headlight shine so that the alert image includes the vehicle. The camera AI will trigger for a car, but the alert image was always just the headlights. I also run the ANPR AI module.
The true test....I have found the AI of the cameras to work even in a freakin blizzard....imagine how much the CPU/GPU would be maxing out sending all the snow pictures for analysis to CodeProject LOL. My non-AI cams in BI were triggering all night. This picture was ran through AI (without the IVS or red lines on it) and it failed to recognize a person in the picture, but the camera AI did. This pic says it all and the video had the red box over it even in complete white out on the screen:
See this thread on how using just Dahua AI may be sufficient for your needs (and other cameras with AI would perform similar):
Who uses Dahua AI capable cameras? Reliable AI for triggering events? Pro's/con's?
biggen
Known around here
- May 6, 2018
- 2,767
- 3,170
Dahua IVS is magnitudes better than CPAI. It isn’t even comparable. Easier to deploy, setup, fine tune, and far more accurate unless you are training your own models with CPAI which is outside 99.999% of most people’s purview.
I’ll never buy another camera again that doesn’t have built in AI.
I’ll never buy another camera again that doesn’t have built in AI.
Appreciate the response! Valid point about CPAI working with BI v6 whenever it's released. Like I said, the camera AI is working great for me and it's faster to trigger notifications than CPAI. The only issue I'm having with one camera is that when it's raining at night, the camera detects a person. I've tried reducing the sensitivity but it remains.I have found the AI of the cameras to be superior for my needs. I ran a lot of side-by-side concurrent testing and the camera always won.
Plus it is one less thing to worry about breaking. I have to restart CodeProject every few days as it just poops out for no reason that I see. And what if CodeProject quits being supported or BI moves to another program.
While some of that third party stuff is cool like tagging was it a dog or a bear, I don't need all that fancy stuff. If my camera triggers BI to tag an alert for human or vehicle and BI can accomplish what I need by way of a text or email or push or whatever, that is sufficient for my needs. I just want to be alerted if a person or vehicle is on my property and the camera AI does a fine job with that.
However, I do run BI AI on a few cameras so that it knocks out headlight shine so that the alert image includes the vehicle. The camera AI will trigger for a car, but the alert image was always just the headlights. I also run the ANPR AI module.
The true test....I have found the AI of the cameras to work even in a freakin blizzard....imagine how much the CPU/GPU would be maxing out sending all the snow pictures for analysis to CodeProject LOL. My non-AI cams in BI were triggering all night. This picture was ran through AI (without the IVS or red lines on it) and it failed to recognize a person in the picture, but the camera AI did. This pic says it all and the video had the red box over it even in complete white out on the screen:
See this thread on how using just Dahua AI may be sufficient for your needs (and other cameras with AI would perform similar):
Who uses Dahua AI capable cameras? Reliable AI for triggering events? Pro's/con's?
TheWaterbug
Known around here
This is fascinating, and contrary to what my intuition would have suggested about whether the "smarts" belong in the camera or the computer. Apparently the cameras can do a lot with their tiny little brains!
I am guessing that, over time, the balance will shift in favor of the the smarts being in the PC and software, but I was wrong the first time, so
I am guessing that, over time, the balance will shift in favor of the the smarts being in the PC and software, but I was wrong the first time, so
No doubt the AI for a PC is getting better.
But it also comes down to cost.
The AI in the camera is included as part of the price of the camera (unless someone is buying a cloud-based camera, which we are not talking about).
The AI as we use in BI is being developed mostly for free by people that enjoy doing this sort of thing.
So in theory the AI of the camera should be ahead of someone doing it for free, but at the same time the free versions offer up lots of bells and whistles the AI camera doesn't.
I am sure AI programs that would knock the socks off camera AI and CodeProject exists, but at what price - probably out of most homeowner range.
People will argue for whichever use case they use.
Some argue it is better in one location and the update is just done once as opposed to each camera whenever their is an update.
But how often is the camera AI updated in a firmware? Not often.
For me, I have spent a lot of time working out field of views that are not trying to do too much and thus the camera AI is pretty accurate.
So in my case, I would rather have X number of cameras each running their own AI so if one camera goes down, I don't lose them all.
In the case of using CodeProject instead of camera AI, if CodeProject goes down on the computer and you are relying on it to do mission critical alerts, you now lost them on all your cameras.
But it also comes down to cost.
The AI in the camera is included as part of the price of the camera (unless someone is buying a cloud-based camera, which we are not talking about).
The AI as we use in BI is being developed mostly for free by people that enjoy doing this sort of thing.
So in theory the AI of the camera should be ahead of someone doing it for free, but at the same time the free versions offer up lots of bells and whistles the AI camera doesn't.
I am sure AI programs that would knock the socks off camera AI and CodeProject exists, but at what price - probably out of most homeowner range.
People will argue for whichever use case they use.
Some argue it is better in one location and the update is just done once as opposed to each camera whenever their is an update.
But how often is the camera AI updated in a firmware? Not often.
For me, I have spent a lot of time working out field of views that are not trying to do too much and thus the camera AI is pretty accurate.
So in my case, I would rather have X number of cameras each running their own AI so if one camera goes down, I don't lose them all.
In the case of using CodeProject instead of camera AI, if CodeProject goes down on the computer and you are relying on it to do mission critical alerts, you now lost them on all your cameras.
TheWaterbug
Known around here
I think the big win for computer-based LPR is the learning that comes from training on multiple data sets. Or I say "will come," since we're not there yet. But people are working on it, as we speak!
And who thought allowing oh, zero, one, eye, bee and eight in license plates was a good idea?
And who thought allowing oh, zero, one, eye, bee and eight in license plates was a good idea?