Salient event-detection in video surveillance scenarios

ABSTRACT

Surveillance video data is accumulating at a staggering rate, making its manual handling impossible. Therefore, automatic tools for analysis and processing of such data are highly needed. In most video surveillance scenarios the most interesting parts of the recorded data are those where an unusual event takes place. The rest of the data, which is actually representing the greatest part, relates to usual or normal activities that are of no real value to the security task and thus its viewing, storage and processing are pure waste of resources. Therefore, automatically finding the exact spot in a surveillance video sequence where an interesting event occurred is of great importance financially and to take timely actions.

In this project we investigate the detection of remarkable events in video surveillance scenarios. We look into how to distinguish events in surveillance scenarios, and further what is a remarkable event. We specifically focus our attention on the event of object dropping in public places such as airports and train stations. We try to answer some of the following questions: Is it possible to create a system for modeling salient events in surveillance scenarios? How does one determine what stands out as a remarkable event? How to distinguish between less remarkable events and more remarkable event taking place?

REFERENCES

• 1: M. P. A. Mecocci and A. Fumarola. Automatic detection of anomalous behavioural events for advanced real-time video surveillance. IEEE International Symposium on Computational Intelligence for Measurement Systems and Applications, 29-31 July 2003, pages 187--192.
• 2: D. Damen and D. C. Hogg. Associating people dropping off and picking up objects. The 18th British Machine Vision Conference, 10th-13th September 2007 University of Warwick, United Kingdom.
• 3: B. W. V. K. S. Fengjun Lv, Xuefeng Song and R. Nevatia. Left-luggage detection using bayesian inference.
• 4: L. Fuentes and S. Velastin. Advanced surveillance: From tracking to event detection. IEEE Latin America ransactions (Revista IEEE America Latina), September 2004, pages 1--1 Vol.2.
• 5: J. S. Hua Zhong and M. Visontai. Detecting unusual activity in video. Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2004), 27th June - 2nd July 2004, Washington D.C., USA, pages 819--826 Vol.2.
• 6: H. Hung and S. Gong. Detecting and quantifying unusual interactions by correlating salient motion. Advanced Video and Signal Based Surveillance, 2005, pages 46--51.
• 7: Neil Johnson , David Hogg, Learning the distribution of object trajectories for event recognition, Proceedings of the 6th British conference on Machine vision (Vol. 2), p.583-592, July 1995, Birmingham, United Kingdom
• 8: L. Latecki and D. de Wildt. Automatic recognition of unpredictable events in videos. Proceedings of the 16th International Conference on Pattern Recognition (ICPR 2002), 11th-15th August 2002, Quebec, Canada, pages 889--892 vol.2.
• 9: M. R. M. Bhargava, Chia-Chih Chen and J. Aggarwal. Detection of abandoned objects in crowded environments. IEEE Conference on Advanced Video and Signal Based Surveillance (AVSS 2007), 5th-7th Sept. 2007, London, England, pages 271--276.
• 10: M. S. Nixon and A. S. Aguado. Feature extraction and image processing. Academic Press, Linacre House, Oxford, United Kingdom, 2002.
• 11: P. Remagnino and G. A. Jones. Classifying surveillance events from attributes and behaviour. Proceedings of the British Machine Vision Conference BMVC, 10-13 September 2001, Manchester, United Kingdom, page Section 8: Modelling Behaviour.
• 12: Silvia Ferrando , Gianluca Gera , Carlo Regazzoni, Classification of Unattended and Stolen Objects in Video-Surveillance System, Proceedings of the IEEE International Conference on Video and Signal Based Surveillance, p.21, November 22-24, 2006 [doi>10.1109/AVSS.2006.32]
• 13: Sijun Lu , Jian Zhang , David Dagan Feng, Detecting unattended packages through human activity recognition and object association, Pattern Recognition, v.40 n.8, p.2173-2184, August, 2007 [doi>10.1016/j.patcog.2006.12.013]
• 14: Sergio Velastin , Paolo Remagnino, Intelligent Distributed Video Surveillance Systems (Professional Applications of Computing) (Professional Applications of Computing), Institution of Engineering and Technology, 2006

INDEX TERMS

Primary Classification:
  I. Computing Methodologies
  I.4 IMAGE PROCESSING AND COMPUTER VISION
      I.4.8 Scene Analysis

Additional Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
  I.5 PATTERN RECOGNITION

General Terms:
Algorithms, Experimentation, Theory

Keywords:
analysis, detection, drop, event, extraction, feature, object, salient, surveillance, tracking, video