Plantify

With the power of AI, make sure your plants are healthy, and disease free- for our future environment and to help your garden grow.

Comment

Inspiration

The control of plant leaf diseases is crucial as it affects the quality and production of plant species with an effect on the economy of any country. Automated identification and classification of plant leaf diseases is, therefore, essential for the reduction of economic losses and the conservation of species. I've strived to promote healthy, thriving plants for our environment, economy, and to support the value of nature even during the technological era we live in today.

What it does

Built on a CNN, we've identified and aggregated information about the disease through our deep learning model. Simply upload a photo of your plant leaf, and we'll give you information on whether it's healthy or not, what disease it has (if applicable), descriptions, preventions, and finally- supplements/fertilizers specific to helping that plant disease. There's also a direct link to buying the supplement if needed.

How we built it

Using 39 Different Categories built in a Pytorch framework, Plantify can address problems like limited scalability and realistic use inefficiencies. By using over 6,000 images of healthy and unhealthy fruit leaves and vegetable plant leaves to train our model, we measured roughly 95% accuracy with a potential real-time diagnosis of diseases of plant leaves. We used a flask server along with Jinja to help with the frontend, along with the use of Figma and plain CSS for the UI. The flask server integrated the results from the pre-trained model to classify the plant leaves.

Challenges we ran into

I ran into a great trouble when trying to figure out the particular architecture of the neural network and model. However, once I used batch gradient with softmax activation and the convolutional neural network, the extremely large dataset which was running locally on Jupyter Notebook took way too long to run (just on CPU, no gpu or tpu added). Luckily, I was able to train the model in time for it to work .

Accomplishments that we're proud of

I love how not only is this a basic classification but also we were able to use supplement and resource datasets to find potential solutions. Not only do users know what disease their plant has, but also how to REMEDY it.

What's next for Plantify

I want to add features that make it easy for users to buy, purchase and share products with others, in order to spread the love for gardening and farming and to make it a more efficient, less stressful and confusing process than it already is.

Share this project:

Updates