This is the repository related to my Bachelor Thesis, regarding image-based chess board and piece recognition. The thesis will be linked here soon.
- install requirements from file
- create debug folder in the project
- update the config.yaml file with filepaths and data directories
- download models and data from here
Dependencies:
python3 main.py testing_images/21.jpg -b'CPS' -m'MobileNetV2'
Arguments are: - b for choosing the board recognition algorithm, either 'Mine' or 'CPS' - m for choosing the Model e.g. 'MobileNetV2', 'NASNetMobile', 'InceptionResNetV2', or 'Xception'
Before utilising checking the model paths or giving a specific model path in the config is recommended
- CPS from Maciej A. Czyzewski's Algorithm
- With improvements of Quintana et al.
