privacycam: tomorrow’s smart video surveillance technology presented by - ankur chattopadhyay
TRANSCRIPT
PrivacyCam: Tomorrow’s Smart Video Surveillance
Technology
Presented By - Ankur Chattopadhyay
Overview Motivation behind PrivacyCam
The PrivacyCam (a description) Applications
Analysis of the technology Conclusion
Today’s Surveillance Systems
SurveillanceCameras
Existing Monitoring
Security Cameras: Trustworthy?
Dr. Roger Johnston’s Keynote Speech on Security Blunders Research shows security cameras:-Don’t prevent crime -Can sometimes help solving crimes-Can be useful for criminal prosecution“If you are going to be using security cameras to invade our privacy, then it
better be for the right reasons” (Dr. Johnston’s Quote)
“Privacy invasion” (violation of the civil rights of individuals)
CCTVs displace crime outside camera bounds
Privacy issues restrict the use of camera in
many such places Unmonitored zones become easy/soft
targets for criminal activity
Present Surveillance Problems
“Eyes” Of A Camera Always Watching You Everywhere
The Challenge At Hand The critical question faced:-
How to apply surveillance in the right way and at the right place to enhance security and at the same time protect our fundamental privacy rights?
The answer:- PrivacyCam
- A surveillance model that is sensitive to privacy and hides identities
- An intelligent technology that addresses the issue of privacy protection
PrivacyCam: The Technology
Uses applied cryptography Privacy through invertible
cryptographic obscuration (PICO) technique
Modeled as an embedded system- Built on a tiny Blackfin DSP (Digital
Signal Processor) board- Includes a small Omnivision CMOS
camera module- Runs on a real time operating system
(uCLinux)
PrivacyCam: How It Looks
Blackfin DSP Module
Omnivision CMOS Camera Module
Integrated System:
PrivacyCam
Algorithm ImplementationCapture image from surveillance
Detect the region(s), which need(s) to be protected, from within the captured
image
Generate encryption key(s) and store the information of the key and detected region(s)
Encrypt the detected region(s) using the encryption key(s), thereby hiding the identity of the individual(s) for maintaining privacy
Results: PrivacyCam
Output Footage
Applications
Incidents of violence & trespassing on campus
Locker Room Elder Care Center
Analysis Of Technology
Strikes a fine balance Compact and affordable
Adaptive and stable Real time performance
Can be used to address the issue of racial discrimination
in security systems
Conclusion: Looking Ahead
Novel patent-pending technology
Anti-Discriminatory Cam
Future Scope/Challenges: - Improved Security of Key(s): Combination of Keys - Enhanced ROI Detection
References A Chattopadhyay, T Boult, “Privacy Cam: a
Privacy Preserving Camera Using uCLinux on the Blackfin DSP”, IEEE CVPR (2007)
T Boult, “PICO: Privacy Through Invertible Cryptographic Obscuration” , Computer Vision, Interactive & Intelligent Environments (2005)
A Senior, S Pankanti, A Hampapur, L Brown, Y Tian, A Ekin, “Blinkering Surveillance: Enabling Video Privacy through Computer Vision”, IEEE Security & Privacy (2005)
Beatrice von Silva,Tarouca Larsen, “Setting the Watch: Privacy & Ethics of CCTV Surveillance”
Open for questions & comments -
Thanks!
Outline Of Applied Algorithm