This project was developed as part of the following research, randomly generating synthetic images of Korean car license plates. We separated them by type according to their shape and background color. You can see the types in the following image
This project was submitted to the DCAI Conference in 2019
| Authors | Authorship | Occupation |
|---|---|---|
| Saidrasul Usmankhujaev | first author | MS student |
| Lee Seonwoo | co-author | PhD student |
| Kwon Jangwoo | corresponding author | Professor |
- This project was developed as part of the Korean License Plate Recognition with Combined Neural Networks
- Random generation is done with the help of the OpenCV library
- Images are placed by the specs of the image
Please read the Car plate notification on standards file for detailed legal information.
Please ensure you have OpenCV installed on your computer.
Before running a code, create a folder named train/test. Run the Python code to see the results, and you should indicate
-t -> train/test to save them in the proper folder,
-n -> the number of images to be generated,
-p -> type of plate to be generated. python licensePlateType.py -t train/test -n int(number) -p int(Type(1~6))@inproceedings{Usmankhujaev2019KoreanLP,
title={Korean License Plate Recognition System Using Combined Neural Networks},
author={Saidrasul Usmankhujaev and Sunwoo Lee and Jangwoo Kwon},
booktitle={International Symposium on Distributed Computing and Artificial Intelligence},
year={2019},
url={https://api.semanticscholar.org/CorpusID:195656548}
}