New CodeProject.AI License Plate Recognition (YOLO11) Module
- Thread starter MikeLud1
- Start date
Mike,
I've uploaded the snapshots here:
filebin.net
All of the ones that are named bad, I could not get them to give me a bad read, even when I matched the frame exactly as I see it in BI. The one that I was able to confirm a bad read is named confirmed.
Let me know if you need a few more of these.
Thanks
I've uploaded the snapshots here:
Filebin | rz06171vkmhvqyv5
Convenient file sharing. Registration is not required. Large files are supported.
filebin.net
All of the ones that are named bad, I could not get them to give me a bad read, even when I matched the frame exactly as I see it in BI. The one that I was able to confirm a bad read is named confirmed.
Let me know if you need a few more of these.
Thanks
Is the "Estimated Speed" OOB as part of Frigate or some additional plugin? Also, did you have to calibrate it somehow or is it just using some time of algorithm? Thanks.
Vettester
Known around here
- Feb 5, 2017
- 1,014
- 992
It’s part of Frigate. You do have to calibrate it by setting up a zone with measurements of all four sides.
To see If I can get the new ALPRYOLO11 module working on some of the users that are having issues running the module I created a totally new .NET module License Plate Recognition (YOLO11 .NET).
This new module defaults to Device ID 0 and has the option to change the Device ID to 1, 2, 3, or 4.
The accuracy for this new module is ok not as good as the License Plate Reader (YOLO11-ONNX) module. If all of the users are not having issues running the new module I will fix the accuracy.
This new module defaults to Device ID 0 and has the option to change the Device ID to 1, 2, 3, or 4.
The accuracy for this new module is ok not as good as the License Plate Reader (YOLO11-ONNX) module. If all of the users are not having issues running the new module I will fix the accuracy.
@MikeLud1, I still cannot run your 1.0.4 version of the license plate reader. The module starts sometimes but always stops immediately. Let me know what information I can provide you to help identify the cause.
Give License Plate Recognition (YOLO11 .NET) 1.1.0 a try and let me know if it works
OK, I'll give it a shot when I get home this evening. Thanks.
I was able to give License Plate Recognition (YOLO11 .NET) 1.1.0 a try and it behaves exactly like the 1.0.4 version of the license plate reader. Stops immediately.
This may be entirely incorrect however I've had good luck lately with using AI to troubleshoot. With your hardware and error log, Chatgpt seems to think the issue lies in the "half-precision" setting. Might be worth attempting the fix below at your own risk? EDIT: Once again @MikeLud1 is way ahead. I didnt realize he'd already addressed this issue with the new .NET module. NVM!
Last edited:
Thanks. I'll make a copy of the original moduleconfig.json, then I will give your suggestion a try. I can always revert back to the original moduleconfig.json.
Unfortunately, I don't have a file named "moduleconfig.json" anywhere on my machine. There was a modulesettings.json, but it didn't have the HalfPrecision parameter.
The "half-precision" setting in the License Plate Recognition (YOLO11 .NET) module is leftover from one of the other modules. It is not used and the module already does not use half-precision.
I just released License Plate Recognition (YOLO11 .NET) 1.1.1. This version uses the same onnxruntime-directml version as the Object Detection (YOLOv5 .NET) 1.14.0 module.
So hopefully if Object Detection (YOLOv5 .NET) 1.14.0 module works for you then License Plate Recognition (YOLO11 .NET) 1.1.1 will work for you.
So hopefully if Object Detection (YOLOv5 .NET) 1.14.0 module works for you then License Plate Recognition (YOLO11 .NET) 1.1.1 will work for you.
Ok, I installed License Plate Recognition (YOLO11 .NET) 1.1.1 and it is up and running. I'll let it go for a while and see how it does. Thanks!
Great! I’m hopeful that anyone who ran into problems with the older versions will find that License Plate Recognition (YOLO11 .NET) 1.1.1 works smoothly now.
Any plan to release object detection Yolo11? Anyway I appreciate your hardwork on license plate recognition (Yolov11) it work great on my cameras.
Once I have the ALPR module working I will finish the YOLO11 module.
It is using GPU (DirectML) and my GPU is Intel HD Graphics 630.