To train the model:
1. Follow the instructions file in _datasets to prepare the data first
2. Install all dependencies listed in requirements.txt; conda create -n lane-detection --file requirements.txt
3. Execute the data_creater.py file. This should create a train_labels folder
4. Execute model_trainer.py. This will save the model under a folder called saved_cnn_model
To evaluate the model:
1. Execute model_evaluator.py and an image window should appear along with a metric in the terminal
To host the Angular GUI:
1. Install npm (Node Package Manager)
2. In a shell type npm install -g @angular/cli
3. Navigate into the GUI folder and type ng serve. This will start the build and hosting process
To host the API:
1. Install the Python package, Flask, by using either Anaconda or pip.
2. In a shell or IDE, execute api.py