Bicycle Theft Detection
Transcript of Bicycle Theft Detection
![Page 1: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/1.jpg)
17/07/2007 1
Bicycle Theft Detection
Dima Damen and David Hogg
Computer Vision Group, School of Computing
University of Leeds
International Crime Science Conference
British Library – London – 17th of July 2007
![Page 2: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/2.jpg)
17/07/2007 Dima Damen
Facts
� 500,000 Bicycles stolen annually in the UK
� 21,236 bicycles stolen in London (2006/7).
� 5% of the stolen bicycles returned to their owners. (2005)
� Highest rate of bicycle thefts in: the Netherlands, Sweden, Japan, Canada, New Zealand, England, Finland and South Africa
![Page 3: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/3.jpg)
17/07/2007 Dima Damen
Facts
Source: EUICS report, The Burden of Crime in the EU, A Comparative Analysis of the European Survey of Crime and Safety (EU ICS) 2005
![Page 4: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/4.jpg)
17/07/2007 Dima Damen
From the news…
� 22/5/2007: Cheltenham - £100,000 worth of
bicycles have been stolen over the past 9
months.
� 7/6/2007: York (290 bicycle thefts during May
2007) city sets up CCTV cameras over
bicycle racks.
� 22/6/2007: Oxford (1800 bicycle thefts during
the last year) city sets up CCTV cameras
over bicycle racks.
![Page 5: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/5.jpg)
17/07/2007 Dima Damen
From the news…
� 23/5/2007 – Catching Daniel Westrop…
“have been stealing commuters' cycles, often
two a day, for the past three years”!!
![Page 6: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/6.jpg)
17/07/2007 Dima Damen
The Task
![Page 7: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/7.jpg)
17/07/2007 Dima Damen
Relevant work
1. Metro Stations
Rota, M. and M. Thonnat. “Video sequence interpretation for visual
surveillance” 3rd IEEE Int. Workshop on Visual Surveillance. (2000).
![Page 8: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/8.jpg)
17/07/2007 Dima Damen
Relevant work
2. Airport Gates
Wang, Y., et. al. “A video analysis framework for soft biometry security
surveillance”. Int. Workshop on Video Surveillance and Sensor Networks. (2005).
![Page 9: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/9.jpg)
17/07/2007 Dima Damen
Relevant work
3. Abandoned Baggage
Ferryman, J., Ed. Ninth IEEE International Workshop on Performance
Evaluation of Tracking and Surveillance (PETS 2006), New York, IEEE, (2006)
![Page 10: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/10.jpg)
17/07/2007 Dima Damen
Bicycle Theft Detection
1. Tracking People
2. Detecting Bicycles
3. Deciding on drop-off and pick-up actions.
4. Comparing colour information
5. Raising warnings
![Page 11: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/11.jpg)
17/07/2007 Dima Damen
Bicycle Theft Detection
1. Tracking People
2. Detecting Bicycles
3. Deciding on drop-off and pick-up actions.
4. Comparing colour information
5. Raising warnings
![Page 12: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/12.jpg)
17/07/2007 Dima Damen
Bicycle Theft Detection
1. Tracking People
2. Detecting Bicycles
3. Deciding on drop-off and pick-up actions.
4. Comparing colour information
5. Raising warnings
Damen, D. and Hogg D. (Sept 2007) Associating People Dropping off and Picking up Objects.
British Machine Vision Conference (BMVC07), Warwick, UK.
![Page 13: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/13.jpg)
17/07/2007 Dima Damen
Bicycle Theft Detection
1. Tracking People
2. Detecting Bicycles
3. Deciding on drop-off and pick-up actions.
4. Comparing colour information
5. Raising warnings
![Page 14: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/14.jpg)
17/07/2007 Dima Damen
Bicycle Theft Detection
1. Tracking People
2. Detecting Bicycles
3. Deciding on drop-off and pick-up actions.
4. Comparing colour information
5. Raising warnings
![Page 15: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/15.jpg)
17/07/2007 Dima Damen
Experiments and Results
1. 1 hour recording
� 45 events –
22 pairs
234Non-Thief
10Thief
Non-ThiefThiefActual
Predicted
456Non-Thief
25Thief
Non-ThiefThiefActual
Predicted
2. 50 min recording
� 22 events –9 pairs
3. 9.5 hrs recording
� 40 events –
18 pairs
1166Non-Thief
24Thief
Non-ThiefThiefActual
Predicted
![Page 16: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/16.jpg)
17/07/2007 Dima Damen
Correct Example
![Page 17: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/17.jpg)
17/07/2007 Dima Damen
Theft Warning Example
![Page 18: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/18.jpg)
17/07/2007 Dima Damen
Strengths & Weaknesses
Strengths:
� Decrease in required
monitoring time.
Weaknesses
Recorded time: 11 hours and 30 minutes
Warning time: 13 minutes
![Page 19: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/19.jpg)
17/07/2007 Dima Damen
Strengths & Weaknesses
Strengths:
� Decrease in required
monitoring time.
� Raises warning, no
action taken.
Weaknesses
![Page 20: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/20.jpg)
17/07/2007 Dima Damen
Strengths & Weaknesses
Strengths:
� Decrease in required
monitoring time.
� Raises warning, no
action taken.
Weaknesses
� Person changing
clothing.
![Page 21: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/21.jpg)
17/07/2007 Dima Damen
Strengths & Weaknesses
Strengths:
� Decrease in required
monitoring time.
� Raises warning, no
action taken.
Weaknesses
� Person changing
clothing.
� Does not detect
suspicious behaviour
![Page 22: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/22.jpg)
17/07/2007 Dima Damen
Strengths & Weaknesses
Strengths:
� Decrease in required
monitoring time.
� Raises warning, no
action taken.
Weaknesses
� Person changing
clothing.
� Does not detect
suspicious behaviour
� System’s failure
cases…
![Page 23: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/23.jpg)
17/07/2007 Dima Damen
System’s Failure Cases
1. The thief wears the same clothing as the
owner.
![Page 24: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/24.jpg)
17/07/2007 Dima Damen
System’s Failure Cases
2. The thief drops another bicycle and picks a
better one at the same time.
3. The tracker loses track of people as they
pause
4. Theft cases of parts of the bicycle
![Page 25: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/25.jpg)
17/07/2007 Dima Damen
False Warning Example
![Page 26: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/26.jpg)
17/07/2007 Dima Damen
Conclusion
� 77% theft detection rate.
� 8.5% false negative rate.
� 1.9% of required monitoring time.
![Page 27: Bicycle Theft Detection](https://reader030.fdocuments.us/reader030/viewer/2022012615/619e3ac81a9e6b2d1a786005/html5/thumbnails/27.jpg)
17/07/2007 27
Thank you..
Contact Details:
Dima Damen
http://www.comp.leeds.ac.uk/dima