Skip to content

Latest commit

 

History

History

License_Plate_Detection_and_Recognition

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
.ipynb_checkpoints
.ipynb_checkpoints
 
 
test
test
 
 
Accuracy.png
Accuracy.png
 
 
Car_plate_detection_and_recognition.ipynb
Car_plate_detection_and_recognition.ipynb
 
 
License_plate_nvidia.h5
License_plate_nvidia.h5
 
 
License_plate_nvidia.json
License_plate_nvidia.json
 
 
License_plate_recognition_nvidianet.ipynb
License_plate_recognition_nvidianet.ipynb
 
 
Loss.png
Loss.png
 
 
README.md
README.md
 
 
digit_crops.png
digit_crops.png
 
 
final_result.jpg
final_result.jpg
 
 
grab_digit_contour.png
grab_digit_contour.png
 
 
lib_detection.py
lib_detection.py
 
 
plate_crop.png
plate_crop.png
 
 
vehicle_with_prediction.png
vehicle_with_prediction.png
 
 
wpod-net_update1.h5
wpod-net_update1.h5
 
 
wpod-net_update1.json
wpod-net_update1.json
 
 

README.md

The methodology to detect and recongize License Plate contains 3 main stages:

  1. Detect and crop the region of License Plate from rear care image using WPOD model

  1. Using common OpenCV techniques such as blur, threshold, findContour to crop all Plate's digits

  1. Using a Nvidia model which was trained to recognize plate digits. The accuracy of this training model as well as the result can be seen as below:


The End!