Inspiration
Developing an accessible tool for farmers to identify sick plants with cutting edge technology. Empowering growers and improving agriculture in a time of climate change.
What it does
We built a Machine Learning model with 44,000 training images of both sick and healthy leaves, from various plants. From a photo of a single leaf, it is able to predict whether the plant is healthy or not, and what disease it may carry. Our model achieved accuracy of 95.87%.
How we built it
We used the Machine Learning framework Pytorch to implement a feed forward neural network with 44,000 training images. It achieved accuracy of 0.9586. We built the web application with JavaScript and React, so all farmers with access to the Internet will be able to upload leaf images from their crops to determine whether they are healthy or not. The application also recognizes diseases, in the case that the plant is sick
Challenges we ran into
We had under 48 hours to create a functioning model using tools that were new to us. Finding the appropriate pre-trained model to apply to our case. Making our WebApp user-friendly for the general public.
Accomplishments we are proud of
Our model is doing pretty well and we pulled out a webapp prototype in a short amount of time
What we learned
Collaboration is an important strength for acting up against climate change.
Log in or sign up for Devpost to join the conversation.