Abstract—There are many applications for the detection of anomalies; in this paper we propose a new method for the detection of abnormalities in crowded scenes. In our method, we present a hand technique for temporal tracking of different people during movement in a video sequence, using the technique of Gaussian mixture model (GMM). This method is based on the blob detector analyzing foreground functioning of cells based on which is created a statistical modelling of the individual item. The Gaussian mixture model (GMM) is used as a position vector to extract different motion characteristics. In addition to detecting the abnormal behavior of the crowd we propose a new simple method. We use the differential method of Lucas and Kanade to estimate abnormal events observed in a surveillance video. It presents an algorithm to accelerate the process of abnormal motion detection based on a local adjustment of the velocity field by calculating the light intensity between two images to detect the abnormal movement.
Index Terms—Anomaly detection, crowd analysis, Gaussian mixture model, pyramids of Lucas and Kanade, video surveillance.
G. Mariem is with Higher Institute of Computer Science and Multimedia of Gabes (ISIMG), Gabes, Tunisia (e-mail: mariem21gnouma@gmail.com.
E. Ridha and Z. Mourad are with Research Groups on Intelligent Machines (REGIM), University of Sfax National Engineering School of Sfax (ENIS), Tunisia (e-mail: ridha_ejbali@ieee.org, mourad.zaied@ieee.org).
[PDF]
Cite:G. Mariem, E. Ridha, and Z. Mourad, "Detection of Abnormal Movements of a Crowd in a Video Scene," International Journal of Computer Theory and Engineering vol. 8, no. 5, pp. 398-402, 2016.
