Plantr

Logo
Upload page
Search result page
Popup modal with sentiment analysis score and wikipedia info

Inspiration

We were inspired by the impact plants have on battling climate change. We wanted something that not only identifies and gives information about our plants, but also provides an indication about what others think about the plant.

What it does

You can provide an image of a plant, either by uploading a local image or URL to an image. It takes the plant image and matches it with a species, giving you a similarity score, a scientific name, as well as common names. You can click on it to open a modal that displays more information, including the sentiment analysis of its corresponding Wikipedia page as well as a more detailed description of the plant.

How we built it

We used an API from Pl@ntNet that utilizes image recognition to identify plants. To upload the image, we needed to provide a link to the image path as a parameter. In order to make this compatible with locally uploaded images, we first saved these into Firebase. Then, we passed the identified species through an npm web scraping library called WikiJS to pull the text content from the Wikipedia page. Finally, we used Google Cloud's Natural Language API to perform sentiment analysis on Wikipedia.

Challenges we ran into

Finding sources that we can perform sentiment analysis on our plant
Being able to upload a local image to be identified, which we resolved using Firebase
Finding the appropriate API/database for plants
Connecting frontend with backend

Accomplishments that we're proud of

Trying out Firebase and Google Cloud for the first time
Learning frontier image recognition and NLP softwares
Integrating API's to gather data
Our beautiful UI

What we learned

How to manage and use the Google Cloud's Natural Language API and the PlantNet API
There are a lot of libraries and API's that already exist to make our life easier