Sie befinden Sich nicht im Netzwerk der Universität Paderborn. Der Zugriff auf elektronische Ressourcen ist gegebenenfalls nur via VPN oder Shibboleth (DFN-AAI) möglich. mehr Informationen...
In this paper we introduce a novel method to detect and alarm pedestrian fall action in video surveillance by using frame difference method and pedestrians marker frame. Firstly, we used the fixed camera to capture video image information, and applied the morphological erosion and dilation method to reduce the noises. Then combined the frame difference method and the background difference method to mark the target contour, and analyzed the contour information to fix the pedestrian location. Finally, by analyzing the motion position and trajectory of the centroid of pedestrian and analyzing length-to-width ratio of pedestrians marker frame and residence time to detect pedestrian fall action and other abnormal motions, and give the alarm in time. The experiments show that the proposed method captures the pedestrian fall action successfully.