Visual Detection of People Movement Rules Violation in Crowded Indoor Scenes
The paper presents a camera-independent framework for detecting violations of two typical people movement rules that are in force in many public transit terminals: moving in the wrong direction or across designated lanes. Low-level image processing is based on object detection with Gaussian Mixture Models and employs Kalman filters with conflict resolving extensions for the object tracking. In order to allow an effective event recognition in a crowded environment, the algorithm for event detection is supplemented with the optical-flow based analysis in order to obtain pixel-level velocity characteristics. The proposed solution is evaluated with multi-camera, real-life recordings from an airport terminal. Results are discussed and compared with a traditional approach that does not include optical flow based direction of movement analysis.
Keywordsmovement event detection Gaussian Mixture Model Kalman filters optical flow
Unable to display preview. Download preview PDF.
- 1.Dong, N., Jia, Z., Shao, J., Xiong, Z., et al.: Traffic abnormality detection through directional motion behavior map. In: 7th IEEE International Conference on Advanced Video and Signal Based Surveillance, pp. 80–84 (2010)Google Scholar
- 2.Veeraraghavan, H., Schrater, P., Papanikolopoulos, N.: Switching Kalman filter-based approach for tracking and event detection at traffic intersections. In: Proc. IEEE Mediterrean Conference on Control and Automation Intelligent Control, pp. 1167–1172 (2005)Google Scholar
- 3.Tusch, R., Pletzer, F., Mudunuri, M., Kraetschmer, A., et al.: LOOK2 - a video-based system for real-time notification of relevant traffic events. In: IEEE International Conference on Multimedia and Expo Workshops, p. 670 (2012)Google Scholar
- 4.Spirito, M., Regazzoni, C.S., Marcenaro, L.: Automatic detection of dangerous events for underground surveillance. In: Proc. IEEE Conf. Adv. Video Signal Based Surveillance, pp. 195–200 (2005)Google Scholar
- 5.Ellwart, D., Czyzewski, A.: Camera angle invariant shape recognition in surveillance systems. In: Proc of the 3rd International Symposium on Intelligent and Interactive Multimedia: Systems and Services, Baltimore, USA, vol. 6, pp. 33–40 (2010)Google Scholar
- 6.Takahashi, M., Naemura, M., Fujii, M., Satoh, S.: Human action recognition in crowded surveillance video sequences by using features taken from key-point trajectories. In: Computer Vision and Pattern Recognition (CVPR) Workshops, pp. 9–16 (2011)Google Scholar
- 7.Laptev, I., Marszalek, M., Schmid, C., Rozenfeld, B.: Learning realistic human actions from movies. In: IEEE Conference on Computer Vision and Pattern Recognition (2008)Google Scholar
- 8.Eshel, R., Moses, Y.: Homography based multiple camera detection and tracking of people in a dense crowd. In: IEEE Conference on Computer Vision and Pattern Recognition (2008)Google Scholar
- 10.Czyzewski, A., Szwoch, G., Dalka, P., et al.: Multi-stage video analysis framework. In: Lin, W. (ed.) Video Surveillance, pp. 147–172. InTech, Rijeka (2011)Google Scholar
- 11.Stauffer, C., Grimson, W.E.: Adaptive background mixture models for real-time tracking. In: Proc. of IEEE Conf. on Computer Vision and Pattern Recognition, pp. 246–252 (1999)Google Scholar
- 12.Czyzewski, A., Dalka, P.: Moving object detection and tracking for the purpose of multimodal surveillance system in urban areas. In: Tsihrintzis, G.A., Virvou, M., Howlett, R.J., Jain, L.C. (eds.) New Direct. in Intel. Interac. Multimedia. SCI, vol. 142, pp. 75–84. Springer, Berlin (2008)CrossRefGoogle Scholar