This thesis will survey existing computer vision methods, focusing on object segmentation of 2D data. An initial study of the healing process will be made to understand the processed data better. A tool will be developed for the manual annotation of imaging data. Collaborating with microbiologists will prepare the training data by specifically highlighting wounds. The study will also involve selecting suitable metrics for comparing the effectiveness of different segmentation methods. Based on these comparisons, the thesis will design a final system capable of monitoring the healing process over time, such as tracking the rate of wound ingrowth. This system aims to contribute valuable insights into healing dynamics, potentially influencing future therapeutic strategies.
Supervisor: Adam Novozámský
Collaborating workplace: Ústavu biochemie a mikrobiologie VŠCHT
Please refer to the user manual.
A .env file must be created in the project root with the following content:
DEBUG=False
DOMAIN=<DOMAIN>
ORIGIN=http://<DOMAIN>
SECRET_KEY=<YOUR SECRET KEY>
NUM_WORKERS=<(number of cores) * 2 + 1>
where is replaced with the domain on which the app will be running and replaced with a randomly generated secret key for Django. Set NUM_WORKERS to an approprivate number of workers corresponding to the number of available cores (there should be number_of_cores*2 + 1 workers).
The port on which the server will be running is by default 1337 (in docker-compose.yml) but it should be changed to 80 or other port depending on whether your server has another proxy on the way.
Then run the following commands:
docker-compose build
docker-compose up
First, run migrations, create a user and launch the backend server:
cd wound-healing/backend
pip3 install -r requirements.txt
python3 manage.py migrate
python3 manage.py createsuperuser
python3 manage.py runserverThen, launch the frontend:
cd wound-healing/frontend
npm install
npm run devFinally, head to http://localhost:8000/.