Can IVS and Video Metadata be used together on Camera AI web 5.0? Trying to set up trip wires and still have meta data shown on screen - how?
IVS and Video Meta data?
- Thread starter JPmedia
- Start date
People's hack for that is do IVS on the camera and video metadata on the NVR.
Mark_M
Known around here
I think it's the 7 and 8 series cameras which can do two AI rules at the same time.
As Wittaj said; the trick people (myself included) use is IVS by camera and Video Metadata by Recorder (for NVR5xxxx -I -EI - EI2).
But that being said:
Some newer 3 and 4 series cameras with AcuPick do effectively Video metadata (that's what Acupick is based off) and IVS at the same time.
As Wittaj said; the trick people (myself included) use is IVS by camera and Video Metadata by Recorder (for NVR5xxxx -I -EI - EI2).
But that being said:
Some newer 3 and 4 series cameras with AcuPick do effectively Video metadata (that's what Acupick is based off) and IVS at the same time.
I trying to implement the features on an IPC-B54IR-ZE. Not sure if it's an S2 or S3. I don't know if the Dahua 5216 NVR has video meta data. I haven't messed with the NVR settings, as I was trying to contain all rules to the camera to lighten the load of the NVR and maintain bandwidth
Mark_M
Known around here
The NVR5216 with either '-I', '-EI' or '-EI2' has video metadata.
Don't worry about using the NVRs AI and limiting bandwidth. 4k/8MP cameras use ~10mb/s at most. So that's all 16 channels with a 4K camera.
The times you go over is for higher res cameras. You can stick an 8K/32MP hubble on a 5 series NVR.
I do also. Solid reliable box.
But no they don’t have VMD by recorder.
This is the one feature and one AI function I bought the new EI2 for.
Unfortunately the new “Next” firmware is crap and some basic functionality doesn’t work.
This is from a 5442 Z4 S3
You really dont lose much without having VMD by Recorder IMHO
Targets are still recognized and labeled in AI playback as Human, Vehicle, non-vehicle...
I run VMD on my other Z4 but really not a lot of difference, and AI search takes longer as you have to search for the 3 categories independently
With VMD YOU have the nifty on screen counter but really, I've come to prefer IVS.
You really dont lose much without having VMD by Recorder IMHO
Targets are still recognized and labeled in AI playback as Human, Vehicle, non-vehicle...
I run VMD on my other Z4 but really not a lot of difference, and AI search takes longer as you have to search for the 3 categories independently
With VMD YOU have the nifty on screen counter but really, I've come to prefer IVS.
Last edited:
Mark_M
Known around here
Wont create a Face db
Shows no records on AI search even though they appear in the live "Events" tab
Various screens flash and stall with video player embedded in them
I havent made a full list becasue I shut it down and refuse to do anymore Beta testing for Dahua for free after the 54PRO fiasco
Shows no records on AI search even though they appear in the live "Events" tab
Various screens flash and stall with video player embedded in them
I havent made a full list becasue I shut it down and refuse to do anymore Beta testing for Dahua for free after the 54PRO fiasco
Understood. I was just curious if both can be used together and the limitations.
When I have some time, I would be curious on how to set up and display the screens with the detections with thumbnails. I presume that info (thumbnails) is written to the NVR drive(s)?