Security Camera Application

Overview

This application turns your webcam into a smart security camera with motion detection capabilities. Built with Python and OpenCV, it offers both manual recording and automated sentry mode for surveillance.

Demo Video

Click the image to see Demo video!!

Features

1. Live Video Feed

• Real-time display from your webcam
• Professional camera frame overlay for better visual feedback
• FPS optimization for smooth video rendering

2. Manual Recording

• Start/stop recording with a simple spacebar press
• Videos saved with timestamp filenames
• AVI format with XVID codec for good quality and reasonable file size

3. Sentry Mode

• Automated motion detection and recording
• Intelligent background modeling to adapt to environmental changes
• Only records when motion is detected to save storage space
• Visual indicators when motion is detected

4. User Interface

• Clear status indicators for recording and sentry mode status
• Background model visualization for debugging
• Countdown timer when initializing sentry mode

How to Use

Prerequisites

• Python 3.x
• OpenCV (cv2)
• NumPy

Installation

pip install numpy opencv-python

Controls

• ESC: Exit the application
• Spacebar: Start/stop manual recording (when not in sentry mode)
• S: Toggle sentry mode on/off

Sentry Mode Setup

When activating sentry mode:

1. A 5-second countdown begins, allowing you to clear any moving objects from view
2. The system collects 1000 background frames to build a reliable motion detection model
3. Once initialized, the system automatically begins recording when motion is detected
4. Recording stops when motion ceases

Configuration

The following parameters can be adjusted at the top of the script:

video_name = "Laptop Webcam"        # Display name for the video feed
video_source_path = 0               # Camera index (0 for default webcam)
video_save_format = "avi"           # Output video format
video_save_format_fourcc = "XVID"   # Video codec

# Motion detection parameters
blur_ksize = (9, 9)                 # Blur kernel size
blur_sigma = 3                      # Blur sigma value
diff_threshold = 150                # Difference threshold for motion detection
bg_update_rate = 0.05               # Background update rate
fg_update_rate = 0.001              # Foreground update rate

How It Works

Motion Detection Algorithm

1. The application builds a background model by averaging multiple frames
2. For each new frame:
   • Apply Gaussian blur to reduce noise
   • Calculate the difference between the current frame and background model
   • Threshold the difference to identify potential motion
   • Apply erosion and dilation to reduce false positives
   • Update the background model adaptively
   • Determine if significant motion is present

Recording Management

• In manual mode, recording starts and stops with user input
• In sentry mode, recording automatically starts when motion is detected and stops when motion ceases
• All videos are saved with timestamped filenames

Limitations and Future Improvements

• Currently supports only webcam input; could be extended to support IP cameras or video files
• Motion detection parameters may need adjustment for different environments
• No cloud storage or remote notification features yet

License

The code is licensed under the GNU AGPL v3.0.

For more details, see the LICENSE file under each directory or visit GNU AGPL v3.0 License