Penn Apps XXIII

This is our submission for Penn Apps XXIII.

We've created a proof-of-concept for detecting diseases using images of plant leaves. We use a neural net trained based on ResNet18 for classification. This ties to a Flask backend, which is called using a Raspberry Pi that can take pictures of leaves. The final result can be read on an OLED screen attached to the Raspberry Pi.

Credits:

API + ML: Naman
Arduino Stuff: Giancarlo
Project Website: Derrick
Presentation Slides: Steven

Links

1. New Plant Diseases Dataset
2. Original Kaggle Notebook