Intruder Detection System

Description

This project is an intruder detection system that uses YOLO (You Only Look Once) for real-time object detection. The system is capable of monitoring multiple cameras and detecting harmful objects such as guns, knives, and other dangerous items. The application streams video feeds from connected cameras and uses YOLO to detect these objects in real-time.

Features

• Real-time video streaming from connected cameras.
• Object detection using YOLO for harmful objects like guns, knives, and more.
• Handles dynamic reconnection to cameras if they disconnect.
• Customizable harmful object classes for detection.
• Web-based interface for monitoring multiple camera feeds.

Requirements

• Python 3.x
• Flask
• OpenCV
• Ultralitycs YOLO
• NumPy

Screenshots

Setup

1. Clone the repository

First, clone the repository to your local machine:

git clone https://github.com/yourusername/intruder-detection.git

2. Create a Virtual Environment

Navigate to the project directory and create a virtual environment:

cd intruder-detection
python -m venv venv

3. Activate the Virtual Environment

On Windows:

venv\Scripts\activate

On macOS/Linux:

source venv/bin/activate

4. Install Dependencies

Install the required dependencies:

pip install -r requirements.txt

Sample requirements.txt

Flask==2.0.1
opencv-python-headless==4.5.3.56
numpy==1.21.2
ultralytics==8.0.0

5. Run the Application

Run the Flask application:

python app.py

By default, the application will run on http://127.0.0.1:5000/.

6. Access the Camera Feeds

Open your browser and go to http://127.0.0.1:5000/ to view the camera feeds. The system will display video streams from the available cameras, and harmful objects will be detected and highlighted.

Structure

• app.py: Main Python script to run the Flask application and handle video feeds and object detection.
• templates/index.html: The HTML template that provides the web interface to view the video feeds.
• static/css/styles.css: Styling for the user interface.
• yolo11n.pt: Pre-trained YOLO model file for object detection.

License

This project is licensed under the MIT License - see the LICENSE file for details.