Camera Tampering Detection System

An automated camera tampering detection system that uses image analysis and statistical techniques to detect when a camera has been tampered with.

Overview

This program monitors a webcam and detects signs of tampering such as:

• Camera being covered
• Camera being moved
• Sudden changes in lighting
• Other forms of interference

The system uses histogram comparison methods across different time periods to detect abnormal deviations from the camera's "normal state".

System Requirements

• Python 3.6 or higher
• Working webcam
• Python libraries:
  • OpenCV (cv2)
  • NumPy
  • Matplotlib
  • collections (standard library)

Installation

1. Clone or download this repository
2. Install the required libraries:

pip install opencv-python numpy matplotlib

Usage

Run the main file:

python camera_tampering_detection_cpu.py

When the program runs:

• A "Combined Feed" window will display with webcam feed (left) and analysis chart (right)
• Press 'q' key to exit the program

Important Parameters

The program uses the following parameters that you can adjust:

1. Buffer Sizes (In the __init__ method)

def __init__(self, short_term_size=3, long_term_size=36):
    self.short_term_pool = deque(maxlen=short_term_size)
    self.long_term_pool = deque(maxlen=long_term_size)

• short_term_size: Number of recent frames to analyze for short-term changes

  • Default: 3 (later adjusted to 3*5=15 in process_webcam())
  • Higher values = slower response but fewer false alarms

• long_term_size: Number of reference histograms to establish "normal" state

  • Default: 36 (later adjusted to 3*60=180 in process_webcam())
  • Higher values = longer learning time, more accuracy but slower startup

2. Detection Threshold (In the detect_tampering method)

threshold = 1
return d_norm > threshold, d_norm

• threshold: Threshold to determine when tampering occurs
  • Default: 1
  • Lower values = more sensitive (might have false alarms)
  • Higher values = less sensitive (might miss some tampering)

3. Processing Frequency (In the process_webcam method)

frame_interval = int(fps / 3)

• frame_interval: Number of frames to skip between analyses
  • Default: fps/3 (analyzes about 3 frames/second)
  • Lower values = analyze more frames, more sensitive but higher CPU usage
  • Higher values = analyze fewer frames, lower CPU usage but less sensitive

4. Histogram Sizes (Number of bins in histograms)

hist = cv2.calcHist([hsv], [0, 1], None, [50, 60], [0, 180, 0, 256])  # Chromaticity
hist = cv2.calcHist([stacked], [0, 1], None, [8, 8], [0, 256, 0, 256])  # L1R
hist = cv2.calcHist([grad_dir], [0], None, [16], [-np.pi, np.pi])  # Gradient

• Changing the number of bins affects the detail of analysis
• More bins = more detailed but resource-intensive and potentially noisy
• Fewer bins = more general, faster but might miss small details

How It Works

1. Initialization Phase:

   • System collects data to establish the "normal state"
   • During this phase, "Initializing..." will be displayed

2. Regular Detection:

   • Analyzes current frames and compares with reference data
   • Calculates normalized dissimilarity (d_norm)
   • If d_norm > threshold, displays "TAMPERING DETECTED"

3. Visual Display:

   • Real-time chart displays d_norm values
   • Threshold is set at 1.0 on the chart

Performance Tuning

• If CPU usage is too high: Increase frame_interval
• If detection is too sensitive (many false alarms): Increase threshold
• If not sensitive enough: Decrease threshold
• For faster initialization: Decrease long_term_size

Advanced Features

The program uses 3 types of histograms to analyze different aspects of the image:

1. Chromaticity Histogram: Analyzes color distribution (hue and saturation)
2. L1R Histogram: Combines total intensity (L1) and color range (R)
3. Gradient Histogram: Analyzes structure and edge directions

The combination of these methods helps the system detect various types of tampering.

Troubleshooting

• Can't open camera: Check webcam access permissions
• Poor performance: Adjust frame_interval higher
• False alarms: Increase threshold or short_term_size
• Not detecting tampering: Decrease threshold

Contributing

Contributions are welcome! Please create an issue or pull request if you have improvements.

License

MIT License

Copyright (c) 2025 LEDUCDAT

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.