Skip to content

[IEEE LUMS CodinGuru 2020 Hackathon - Winner] Detect and track locations of TYLCV disease by sending a picture of a leaf to a web chat.

License

Notifications You must be signed in to change notification settings

ZoraizQ/agriscan

Repository files navigation

AgriScan

Deep Learning based Web App for Detection of Plant Diseases (SDG #2) DevPost

Send a picture to AgriBot of a tomato plant leaf (updates for more incoming). It will reply with disease classification predictions for your plant regarding Tomato Yellow Leaf Curl Virus (TYLCV). The location of the sender is registered and populated on a disease map for admins to pinpoint sources of early crop disease.

SpiderMites TargetSpot Healthy

View the Jupyter Notebook for the fast-ai CNN to see how the neural network processes data and its accuracy.

Prerequisites

Uses a JS to Python connector to run the CNN's python model scripts.

NPM, PIP3, MongoDB relevant dependencies.

Web Frameworks:

  • Express.js
  • Bootstrap
  • Multer
  • NodeJS
  • MongoDB
  • JS
  • HTML5
  • CSS3
  • VUEJS

Python Libraries:

  • fastai
  • numpy
  • torch

Installing

Test using the production link provided or using http://localhost:8080 after installing dependencies and running the following.

npm install
pip3 install -r requirements.txt
npm start

Facebook Technlogies Used

  • flow - vs code extension, type checking JS
  • pytorch (fastai) - for training the deep learning network

Built With

  • Hackathon Boilerplate provided
  • Trello for Project Management
  • Git for version control
  • Heroku for deployment

Acknowledgments

About

[IEEE LUMS CodinGuru 2020 Hackathon - Winner] Detect and track locations of TYLCV disease by sending a picture of a leaf to a web chat.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published