Image Processing Project

Authors: Giavinh Lam, Greg Hetherington and Petar Kenic

Created on March 23, 2019

About

Open-ended project designed for CIS4720. Web application which detects and computes the ripeness of a banana. React.js used for the frontend and Python 3 used for the backend (image processing algorithm, file operations...).

The React.js frontend is deployed on http://stark-gorge-37151.herokuapp.com/ (initial load may take long due to how free dynos work in Heroku) and the Python/Flask component (used as a server to handle requests and implements the banana ripeness detection algorithm) is deployed using pythonanywhere (on http://GV79.pythonanywhere.com).

Running the webserver

• install python dependencies

  pip install flask
  pip install flask-cors --upgrade
  

• navigate to /backend/src

  export FLASK_APP=server.py
  flask run
  

• common errors

  • use 'set' instead of 'export' for Windows
  • under PowerShell (ex. when using VS code) use the following command

    $env:FLASK_APP = "server.py"
    flask run
    

Updating npm files

• delete "package-lock.json" file

  npm install
  

Launching website

• navigate to root folder

  npm start
  

Notes:

• Images can be most file types (ie. PNG, JPG, TIFF, etc)
• Input image must have dimensions > 100 x 100 pixels
• Libraries that must be updated:
  • numpy
  • PIL
  • imageIO
  • cv2 (pip can be used to update these)