Bio-Compute Explorer

The place that helps plants

Inspiration

The inspiration for this project was our love for plants and to take care of them. Additionally, we wanted to make a digital space that felt more like a real desk where you actually move stuff around, and we wanted to see if using AI could help people take better care of the plants they love.

What it does

Our website intensely checks a picture of a plant and gives a very accurate reading of the condition of it and how to fix the condition of the plant.

How we built it

We built a plant-health scanner using TensorFlow.js and MobileNet to run AI diagnostics locally within the browser. I developed a custom drag-and-drop interface using the FileReader API to allow for instant photo uploads and previews. The system then processes the image data to calculate a Vitality Rating and provide a clinical diagnosis.

Challenges we ran into

The main challenges we ran into were definitely making the project into a real website that looked appealing on different screens and getting that drag and drop mechanic to actually work without the browser just opening the image file in a new tab. We also had to figure out how to make a connection between raw AI data and actual plant advice.

Accomplishments that we're proud of

We are proud of the drag and drop feature and how the final results looked like.

What we learned

We learned a lot about neural networks work by using TensorFlow.js and MobileNet to let a browser see and diagnose plant health

What's next for Bio-Compute Explorer

I believe that we can enhance this program by making the AI smarter at recognizing the minute details in certain images and the website even easier to use. Currently, we are only giving a couple of advices for the user, and in the future, we would like to add even more with suggestions that could completely change the way the user's plants live.