Vigision: Advanced AI-Powered Home Surveillance System

Vigision is an innovative Network Video Recorder (NVR) and AI Computer Vision software designed to transform traditional home camera systems into proactive guardians. Leveraging advanced Deep Learning technologies, Vigision offers features such as object detection, pose estimation, and intelligent motion detection to provide reliable, intelligent protection for your home and loved ones.

🚀 Product Features

• AI-Driven Alerts: Only significant movements trigger alerts, reducing unnecessary notifications.
• Fall Detection: Proactively monitors vulnerable family members.
• Self-Hosting: Host on local hardware using Docker for enhanced security and privacy.
• Global Accessibility: Access and manage your system from anywhere.
• Customizable: Configure camera systems and features based on personal needs.

---

🛠️ Technologies Used

• Frontend: React, TypeScript, Vite, Tailwind CSS, Axios
• Backend: Flask, SQLite, peewee, PyTorch, OpenCV, YOLO, AlphaPose
• DevOps: Docker, NGINX, ngrok, ZeroMQ
• Other Tools: FFmpeg, NumPy, pandas, Python

---

🖼️ Screenshots

---

📋 Prerequisites

Before getting started, ensure you have the following installed:

• Ubuntu or Ubuntu WSL on Windows
• Docker

---

🚀 Getting Started

Backend Setup

1. Open WSL Ubuntu.

2. Navigate to the project directory:

   cd vigision-nvr

3. Start the Docker containers:

   docker-compose up -d

4. Verify the backend version:

   make version

5. Run the backend:

   python3 -m vigision

Frontend and Development Setup

1. Open Visual Studio Code.

2. Attach to the running Docker container:

   • Select the vigision container.

   • Choose the workspace directory. The first time it prompts for the code path, enter:

     /workspace/vigision
     

   • (Optional) Check the path by executing:

     pwd

     inside the container.

3. Install dependencies and start the frontend:

   npm install && npm run dev

---

🎥 Demonstration Video:

Watch the Demo on YouTube