This project made for Satria Data 2023 elimination round. The goal is to make a program that can recognize plate number from an image.
This project reformatted using Black. The project was trained using Kaggle's with T4 x2 GPU. We wouldn't recommend to run this project on your local machine because it will take a lot of time and resources.
We got 97.33 score accuracy on leaderboard and finished on 33rd place.
.
├── data
│ ├── test_data
│ │ ├── *.png
│ │ └── DataTest.csv
│ └── train_data
│ ├── *.png
│ └── DataTrain.csv
├── Resources
│ └── Juknis Satria Data 2023.pdf
├── src
│ ├── playground
│ │ └── *.ipynb
│ ├── csvpreprocess.ipynb
│ ├── prediction.ipynb
│ └── preprocess.ipynb
├── .gitignore
├── LICENSE
└── README.md
data
folder contains the dataset for training and testing.Resources
folder contains the rules for this project.src
folder contains the source code for this project.src/playground
folder contains the source code for testing the model.src/csvpreprocess.ipynb
is the source code for preprocessing the dataset.src/prediction.ipynb
is the source code for predicting the dataset.src/preprocess.ipynb
is the source code for preprocessing the image.