surgical-copilot

Create environment

1. Follow the official guide to install tensorflow and/or pytorch
2. To test GPU open Anaconda Prompt as admin and type

conda activate tf
python ./test/tf_GPU.py

to test with tensorflow or

conda activate tf
python ./test/pt_GPU.py

to test with pytorch

3. Install opencv using conda
4. Test cv2 with your webcam

python ./test/cv2_cam.py

Online images collection

Instructions to collect images while performing ductoscopy

1. Be sure the virtual environment is active and you are in "data collection" folder
2. Edit get_images.py and set labels and number of images
3. Run get_images.py
4. Collected images will be automatically stored in new created folders corresponding to the labels

Offline images labelling

1. If first use run setup_labelImg.py to clone the repository.
   
2. Run set_labels.py
3. Opend folder of the images you want ot label
4. Press "w" to start labelling and CTRL+S to save the labelled image