Oh okay perfect! I did all of that, I just wasn’t sure if I was missing something to connect AI tool to Blue Iris - will try and test out tonight!
As far as I know, that is not possible. You could set up two separate cameras pointing to the same cameras. So Driveway-Car and Driveway-Person. Then add both in AI tools. For the Car one, add the mask and check to only detect cars. Then for the Person one, do not add a mask but uncheck alerting for cars. You can then Hide one of the cameras in Blue Iris so you dont have to see or deal with it. In the AI Tools, have both cameras send the trigger alert to the one camera you want. So both Driveway-Car and Driveway-Person will send a trigger to Driveway-Person. I've never tried this but it seems like it would work. You also might be able to use one camera and just create two AI tools cameras that reference the same image created. Not sure what would happen when both try to process one file.
How do I add the summary/description & the confidence into the blue iris push notification? I see that AI Tool updates the stream with the identified items and confidence, but when I tried adding &summary into the 'text' section within BI alerts, it didn't work properly... Ideally, I am only looking to add 'Person + Confidence' to the push notification when a person is detected, and not all of the other things it finds (ex. couch, bed)It will only process in one camera- but Chris may change that if you ask and he can!
Entering the Environmental Variable as shown above wouldn't work because hyphens are not allowed. I'm going to read through the second link you provided. I really appreciate your help!
I changed the implementation to the windows deepstack executable and it's working now.
Damn it, that's my old post. I must try again in that case, as I never got it working...
We get viscio-scene from the deepstack tutorial of docker installation.@105437 - From the error, the problem still does not look like it is on the AITOOL end. I dont think you have VISION-DETECTION=True set correct as talked about above. I wouldnt know how to do that for QNAP.
Same error here, and I'm pretty sure it is because he is using VISION-SCENE=True not VISION-DETECTION=True:
GPU Docker Instance on WSL with Ubuntu-20.04 LTS - Returning null
I was maxing out my ancient CPU (i7 960 from back in the day) trying to run Deepstack, with image processing times being multiple seconds. Not workable. However, given the recent(ish) announcement of the CUDA toolkit being available in WSL 2, I now have a GPU based Deepstack Docker running. I...forum.deepstack.cc
Also I saw the Deepstack tab > STOP being used in the log before - to be clear, that is only for the Windows version running on the same machine. Nothing to do with the QNAP version. And you wouldn't see those PYTHON errors from qnap either.
Yep, that feature looks at the entire image and tries to decide what the scene is. (conference room), detection looks for an object in an image to determine if its there. Or at least that is how I read it...
Not the camera folder, but it does save it in the _settings. If you copy your bmp from camera to _settings does fix the issue?
That's it I don't have any mask bmp or otherwise so not sure what it is seeing? I will check and see what is in the _settings
BINGO! That was it. I must have been playing around with thar camera in a previous version, there was indeed a BMP in there. Thanks for help!
