Our dad is an aspiring gardener, but for him, and many others, it is difficult to distinguish the mineral deficiency that each plant has, as well as keep track across different species. We wanted to find a way for amateur gardeners, in addition to farmers with hundreds of crops, to easily keep track of their plants nutritional needs and help them improve the quality and health of their plants.
What it does
The gardener can take a photo of their plant, and the app will instantly identify whether the plant has a deficiency or is healthy. These plant images will then be saved by date under 'My Plants,' where the user can go to track each plant's development, as well as understand the causes and effects of their nutrient deficiencies or celebrate the health of their plant. Our app also offers recommendations for remedies to treat nutrient deficiencies, under the user's 'Supply Shed.'
How I built it
We first wireframed the app in Figma to focus on user interface, and we then built our app in Swift, creating our own image classifier model with CreateML and the rest of the functionality for the app.
Challenges I ran into
Since this was our first time using CreateML, there was a bit of a learning curve when it came to creating our own training dataset of images and improving the models accuracy.
Accomplishments that I'm proud of
We are proud that we learned how to use CreateML in Swift and that we can now use this app in our own backyard.
What's next for Jardin
With more user images (collected with their permission, of course), Jardin can become more accurate in predicting mineral deficiencies. We also want to expand Jardin to detect water and sunlight deficiencies, which can greatly expand its functionality.