Multi-Camera Surveillance with Time-Chunks

This project provides a Python-based application for managing multi-camera surveillance systems. The application allows users to record videos from multiple cameras, preview live camera feeds, and save recordings in time-chunked segments.

Features

• Supports multiple cameras simultaneously.
• Adjustable recording settings:
  • Frames per second (FPS)
  • Resolution (Width x Height)
  • Time-chunk duration (minutes)
• Live preview of camera feeds.
• Automatic file segmentation for easier storage and retrieval.
• Intuitive GUI built with PySide6.
• Saves recordings in the specified output directory.

Prerequisites

Ensure the following dependencies are installed:

• Python 3.8 or later
• PySide6
• OpenCV
• NumPy

You can install the required Python packages with:

pip install PySide6 opencv-python numpy

Usage

Running the Application

1. Clone the repository:

   git clone <repository_url>
   cd <repository_directory>

2. Run the application:

   python main.py

GUI Overview

1. Cameras Tab

   • Displays detected cameras.
   • Use checkboxes to select cameras for recording.

2. Preview Tab

   • Shows live previews of selected cameras.

3. Settings

   • Adjust FPS, resolution, and chunk duration.
   • Select an output directory for saving recordings.

4. Control Buttons

   • Start Recording: Begin recording from selected cameras.
   • Stop Recording: Stop recording and close active processes.
   • Close App: Stop recording (if active) and exit the application.

Key Settings

• FPS: Frames per second (0 uses the camera's default).
• Width and Height: Resolution for video recording (0 uses the camera's default).
• Chunk Duration: Length of each video segment in minutes.
• Output Directory: Directory where video files are saved.

Code Overview

Key Components

• CameraProcess: A multiprocessing class to manage individual camera streams. Each camera runs in its own process to ensure smooth performance.
• MainWindow: The main application window with tabs for camera selection, live preview, and settings.
• discover_cameras: Function to detect available cameras.
• start_new_segment: Creates new video files based on the specified chunk duration.
• update_previews: Updates live preview feeds using frames from the camera processes.

How It Works

1. The application detects available cameras.
2. Users select cameras and set desired recording parameters.
3. Each camera is managed by a separate process to record video chunks and update live previews.
4. The application saves recordings in .mp4 format in the specified directory.

Example Output Structure

output_directory/
├── camera_0_segment0_20250121_120000.mp4
├── camera_0_segment1_20250121_121000.mp4
├── camera_1_segment0_20250121_120000.mp4
└── camera_1_segment1_20250121_121000.mp4

Known Issues

• Preview updates may lag with high FPS settings or slow hardware.
• If no cameras are detected, ensure the cameras are properly connected and drivers are installed.
• It will probably crash when you click stop recording or close, but the videos are saved properly and thats all that matters. Just close the app the hard way and re-open it.

Contributing

Feel free to open issues or submit pull requests to enhance the functionality.

License

This project is licensed under the MIT License. See LICENSE for details.