While we were doing our research about habitat preservation, we learned that a lot of people are disconnected from their own environment. We believe that the first step of environment preservation is to let people know what is around them. Additionally, native plants are on the decline, and we want to help preserve and protect these native species.
Users take photos of plants or animals they see around them and upload it to BloomWatch. Here, an AI algorithm searches the image for plants and animals and identifies the species with pinpoint accuracy. Users are awarded points for each new species they find. For example, an image of a plant would be 1 point, while an animal would be 3 points. A leaderboard of zip codes displays points values for specific areas. This encourages users to compete from all around the United States. This app serves as a crowdsourcing dataset, pulling what species are prominent in different areas.
We built our project is Visual Studio code, using python. For the AI we used Google's Gemini API, and we also used Google cloud to store usernames, passwords, and backyard statistics.
One challenge we ran into was prompting the AI to get the correct data in the correct format. IT required us to be super specific on our prompt and only ask what we needed. Additionally, setting up the Google cloud was also a challenge.
We are proud of how the AI turned out in the end. It was able to get the exact species of the image given. Also, the cloud software worked very well to store usernames, passwords, and statistics.
We learned different about conservation methods and how to stay active in the environment. Additionally, we also learned how to use databases and the Google Cloud.
We hope to send our data of species around the US to scientists and researchers. These people would analyze the data and see where native species are on the decline. Then, they would help put in conservation efforts in those areas.
- firebase
- geopy
- google-cloud
- nomatim
- numpy
- python
- streamlit
- Nomatim
