Streamline the identification process of invasive and non-native species in the Florida area.
An app and environmental tool developed to equip environmentalists and residents in the Sunshine State with the power to recognize possible threats towards ecosystem health and biodiversity. With the rapid spread of invasive species, Florida’s unique, expansive, and cherished ecosystems stand to lose their natural rich biodiversity to overpopulation and invasive competition for resources. Florida GreenGuard AI aims to empower the people of Florida to recognize potential threats to the state’s natural tropical beauty.
The app leverages the power of a machine learning image classification model trained on a plethora of species from thousands of collected images of non-Florida native flora thorugh tensorflow deep learning algorithms. Additionally, the app utilizes the OpenAI API via an open-source Kotlin support library to leverage the power of a large language model, generating helpful descriptions for each recognized plant. The app also features an image classification model to detect possible deficiencies in plant nutrients (nitrogen, phosphorus, magnesium, etc.)
The app features a page scrollable under the main species detection UI that serves as an information hub for invasive species and non-native species, their similarities and differences, as well as the specific hinderances of the current trained model.
With this tool, the hope is for environmentalists and interested residents alike to identify non-native plant species in Florida with high accuracy and simply begin to take the required steps to deal with wildlife that impact the state’s biodiversity.
*Note: A new OpenAI API Key will need to be generated and called for anyone who wishes to clone the project into a separate IDE.
Information garnered concerning invasive species within the model are based on the Florida Natural Areas Inventory Invasive Species Directory.
(Unaffiliated)