Active Alert System

Overview

The Active Alert System is a desktop application designed to detect various weapons using the YOLO (You Only Look Once) object detection model. Upon detection, it sends an emergency email with a screenshot of the detected object using the Gmail API. The application is designed to run on macOS, leveraging the power of Python, OpenCV, and machine learning models.

Features

• Weapon Detection: Detects weapons such as grenades, guns, knives, pistols, handguns, and rifles using a trained YOLO model.
• Real-Time Monitoring: Continuously monitors the live camera feed and highlights detected weapons with bounding boxes.
• Emergency Email Notifications: Sends an emergency email with an attached screenshot whenever a weapon is detected.
• Screenshot and Thumbnail Management: Captures full-sized screenshots and creates thumbnails for quick viewing in the "Records" window.
• User-Friendly Interface: Offers a simple and intuitive UI with fullscreen capabilities.

Installation

Prerequisites

• Python 3.7+
• pip (Python package manager)

Setup

1. Clone the repository:

   git clone https://github.com/gurliv21/Active-Alert2.git
   cd active-alert-system

2. Install required dependencies:

   pip install -r req.txt

3. Set up the environment variables:

   • Create a .env file in the root directory.
   • Add your Gmail API key and Gmail user details in the following format:

     API_KEY=your_gmail_api_key.json
     [email protected]
     

4. Download the YOLO model:

   • Ensure that best.pt (the trained YOLO model file) is placed in the root directory or specify the correct path in the code.

5. Run the application:

   python main.py

Usage

1. Home: The default window displaying the main UI.
2. Start Recording: Begins the real-time camera feed and weapon detection.
3. Stop Recording: Stops the real-time monitoring and returns to the Home screen.
4. Records: Displays thumbnails of detected weapons.
5. About Us: Displays information about the application.
6. Log out: Closes the application.

How It Works

1. Weapon Detection: The application uses the YOLO model to detect predefined classes of weapons. When a weapon is detected with a confidence level above 60%, it highlights the object in the video feed.
2. Emergency Alert: A screenshot of the detected weapon is captured and sent as an attachment via email using the Gmail API.
3. Thumbnails: For each detection, a smaller thumbnail is created and displayed in the "Records" window for quick review.

Technologies Used

• Python

• OpenCV: For real-time video processing.

• YOLO: For object detection.

• Gmail API: For sending email notifications.

• Tkinter: For the graphical user interface.

  Screenshots