New CodeProject.AI License Plate Recognition (YOLO11) Module

MikeLud1

IPCT Contributor
Apr 5, 2017
2,546
4,646
Brooklyn, NY
Since CodeProject.AI has not been receiving any updated modules I figured a way for me to provide updates and new modules. The fist module I add is a total new License Plate Recognition (YOLO11) module. To get my updated or new modules all that is needed is to replace one file see attached (appsettings.json). I recommend backing up the original appsettings.json file before replacing the file. After replacing the file restart the CodeProject.AI server then when on the Dashboard page do a CTRL + F5 to refresh the Dashboard page.

This will now use my available modules in the Install Modules tab. Existing modules will show as Private. If you uninstall a Private module and want to reinstall the module you will need to replace the appsettings.json file with the backed up appsettings.json file. After replacing the file restart the CodeProject.AI server then when on the Dashboard page do a CTRL + F5 to refresh the Dashboard page and the original list of modules will be restored.

Any new or updated module will be in this GitHub repo GitHub - MikeLud/MikeLud-CodeProject.AI-Modules: MikeLud-CodeProject.AI-Modules
The models will be stored on Hugging Face MikeLud (MikeLud)

Currently the new License Plate Recognition (YOLO11) module only works in Windows. The module will work with any GPU, CUDA is not needed. No Blue Iris setting changes are needed to use the new module. This new module I am planning on adding vehicle color, make, model and also State recognition.
If you see any issues with this module let me know.

1760293131524.png

1760292349662.png
1760293514821.png
 

Attachments

Presumably there is no reason to switch over if we are only using the other models (everything but license plate recognition)?
 
Hi Mike, thanks for sharing this! First impression was it almost works too well. I had to raise my min confidence % because it was picking up characters that weren't there. I changed it from 60 to 70 and it seems to be working well on a CPU only system.

Screen Shot 2025-10-12 at 2.08.43 PM.png
 
  • Haha
Reactions: jrbeddow
How about accuracy? Because it seems that in this newer version of license-plate detection, it requires multiple attempts to finally read the license plate correctly, at least when I do this manually in CodeProject AI Explorer. To clarify, when I say multiple attempts, I mean manually clicking the "detect license plate" button twice or more to get the readings correctly of the same sample image, and sometimes on the first try, it does not recognize the license plate at all, or maybe it does fully support license-plate in Canada.
 
Last edited:
How about accuracy? Because it seems that in this newer version of license-plate detection, it requires multiple attempts to finally read the license plate correctly, at least when I do this manually in CodeProject AI Explorer.
When I was running it in parallel with the old module I was see better accuracy. Can you post the image that you were using to test it manually. Also post a screenshot of you System Info