Webcam Surveillance Suite

1
Webcam Surveillance Suite

• Saket Pusalkar
• Dipak Jogle

2
Motivation

• To make a general purpose Surveillance Suite
  which is
• Designed for end-users
• Easy to install use (minimum hardware
  dependencies)
• Extendable
• Portable
• Computationally efficient for real-time usage

3
Technologies Used

• Operating System Windows xp
• Development language Visual C (.NET framework)
• Component for hardware interaction Directshow,
  Directshowlib.dll

4
Directshow

• Concepts Filtergraph , filters , pins
• Modules in our Filtergraph

Image Store Filter
Image Capture Filter
Device Stream
Video Render Filter
Video Preview Filter
5
Software Capabilities

• Motion Alarm
• Frame Capture
• Area specific Sensitivity
• Meta-database

Image query interface
Image Browser
Sensitivity Function
Motion Capture Alarm
6
Algorithm

• Motion Detection using GMM Sensitivity function
• Create m models for every pixel of scaled down
  image
• For every new image, every pixel is compared with
  its models
• The comparison is thresholded according to
  sensitivity of the pixel
• Outliers are considered as motion changes

7
Conclusion

• The system is a useful tool for an end user with
  limited but effective functionalities for small
  scale surveillance
• A Simple Very useful extension could be head
  detection and capture
• This system can be used as a platform and the
  functionalities can be extended to perform
  various tasks like recognition, tracking which
  will lead to a very useful Surveillance System

8
References

• MSDN Library , Microsoft 2005
• N. Friedman and S. Russell, Image segmentation
  in video sequences A probabilistic approach, in
  Annual Conference on Uncertainty in Artificial
  Intelligence, University of California, Berkeley,
  CA, 1997
• Chris Stauffer W.E.L Grimson Adaptive background
  mixture models for real-time tracking 1998