MikeLud1
IPCT Contributor
You can test which one is faster for you using CodeProject.AI Explorer, I find .NET to be faster.
Blue Iris and CodeProject.AI ALPR
- Thread starter MikeLud1
- Start date
MikeLud1
IPCT Contributor
Another note if you use the Windows version of CodeProject.AI the ALPR module will also use your GPU and be faster.
prsmith777
Getting comfortable
I can't get the GPU to process plates. CPU works fine.
nvcc -- version says I have CUDA 11.8 installed. CP.AI info says I have 12.2 installed. Not sure if this is the problem
nvcc -- version says I have CUDA 11.8 installed. CP.AI info says I have 12.2 installed. Not sure if this is the problem
MikeLud1
IPCT Contributor
What is shown in CP.AI is what the GPU driver will support up to the CUDA version, it is not the CUDA version that is installed, nvcc -- version show what CUDA version is installed. So you have CUDA 11.8 installed.
Did you also install cuDNN, below is the link
Log in
developer.nvidia.com
prsmith777
Getting comfortable
I ran the CuDNN install script from the their website. I think it installed correctly not sure. This is the folder after running
MikeLud1
IPCT Contributor
prsmith777
Getting comfortable
Yes I have all those files in correct folder
I tried rebooting also, no change.
I tried rebooting also, no change.
prsmith777
Getting comfortable
Took a while. Had to sign up for NVidia developer. I did replace the files with the most current ones then rebooted... still no luck.
So I thought I'd uninstall CP.AI including all the folders and data and reinstall. Still no good.
I'm testing images using CP License Plate reader test page. Works fine with CPU, but not GPU.
So I thought I'd uninstall CP.AI including all the folders and data and reinstall. Still no good.
I'm testing images using CP License Plate reader test page. Works fine with CPU, but not GPU.
prsmith777
Getting comfortable
MikeLud1
IPCT Contributor
prsmith777
Getting comfortable
yes have tried enabling and disabling and enabling gpu.... no good
MikeLud1
IPCT Contributor
Does any of the modules work with the GPU?
prsmith777
Getting comfortable
Yes I have Object Detection(YOLOv5 6.2) with GPU enabled. I just tested a cat with ipcam-animal model and it works.
MikeLud1
IPCT Contributor
Open a command prompt and try running NVIDIA_SLA_cuDNN_Support.txt It should open the file like below.
prsmith777
Getting comfortable
Maybe we are getting somewhere now...getting this:.
Microsoft Windows [Version 10.0.19045.3324]
(c) Microsoft Corporation. All rights reserved.
C:\Users\BlueIris>NVIDIA_SLA_cuDNN_Support.txt
'NVIDIA_SLA_cuDNN_Support.txt' is not recognized as an internal or external command,
operable program or batch file.
C:\Users\BlueIris>
Microsoft Windows [Version 10.0.19045.3324]
(c) Microsoft Corporation. All rights reserved.
C:\Users\BlueIris>NVIDIA_SLA_cuDNN_Support.txt
'NVIDIA_SLA_cuDNN_Support.txt' is not recognized as an internal or external command,
operable program or batch file.
C:\Users\BlueIris>
MikeLud1
IPCT Contributor
Try running the install_CUDnn.bat again. After running reboot so that the paths get updated.
File Download - CodeProject
www.codeproject.com
Pentagano
Getting comfortable
I've just span up a new server with VMware esxi hypervisor with a windows 10 VM
Installed CPAI and all is online ready for testing
But when I click detect in the dashboard with an example jpeg no predictions are found. Tried all the models. Something is missing.
These are the trace logs: no errors just queued and dequeued.
Any ideas what I'm missing?
Thanks
Installed CPAI and all is online ready for testing
But when I click detect in the dashboard with an example jpeg no predictions are found. Tried all the models. Something is missing.
These are the trace logs: no errors just queued and dequeued.
Any ideas what I'm missing?
Thanks
prsmith777
Getting comfortable
cmd function now works but it brings up a blank notepad page
prsmith777
Getting comfortable
I am about ready to give up. I can leave LPR on CPU and use GPU for other detection. Computer should be fine. Thanks for your help.