Automated video surveillance for snatching detection using majority rule network and gabor features

In this paper, we propose an intelligent security camera system for automated detection of snatching incidents in which a bicycle is used. In addition, the effectiveness of the Basic Snatching Action Model (BSAM) and Gabor features for automated detection of snatching incidents is presented. The localization of moving objects in a video stream and human behavior estimation are the key techniques applied in the proposed system. Gabor features are determined from video streams and, using a majority rule network (MRN) composed of various artificial intelligence (AI) systems, the video streams are automatically classified into criminal or non-criminal scenes. In our experiments, we considered some scenarios of snatching incidents in which the perpetrator uses a bicycle. The experimental results show that the proposed system can effectively detect criminal scenes with high accuracy. © 2016 The Institute of Electrical Engineers of Japan.