🐢 Turtle AI — Smart Turtle Monitoring System

Turtle AI is an AI-powered wildlife monitoring system designed for rehabilitation environments. It is locally-running application that captures snapshots from live RTSP camera feeds and uses an open-source LLM (Gemma 3n via Ollama) to detect signs of turtles in distress. When distress is detected, it sends alerts via Twilio and displays messages in the web interface.

---

🔧 Setup & Installation

Prerequisites

Twilio Account with:

• Account SID
• Auth Token

You must install the following before running this project:

• Python 3.9+
• Docker and Docker Compose v2
• FFmpeg
• Ollama (must be running locally)
• Gemma 3n model installed:

  ollama run gemma:3n:e4b

Installation

• Create a file named docker-compose.override.yml and add your specifiic environment variables (see docker-compose for format and examples)

git clone https://github.com/lynnaloo/turtle-ai.git
cd turtle-ai

---

🏃‍♀️ Run the Application

docker compose up --build

✔️ The app will now:

• Capture a frame from your RTSP camera every 10 minutes (or your custom interval)
• Analyze the image using Gemma via Ollama
• Alert you if a turtle appears distressed (via WhatsApp)

---

📷 System Architecture

📰 Presentation

Here is a YouTube teaser and demo of the project!