Abnormal crowd behaviors in high density situations can pose great danger to public safety. Despite the extensive installation of closed-circuit television (CCTV) cameras, it is still difficult to achieve real-time alerts and automated responses from current systems. Two major breakthroughs have been reported in this research. Firstly, a spatial-temporal texture extraction algorithm is developed. This algorithm is able to effectively extract video textures with abundant crowd motion details. It is through adopting Gaborfiltered textures with the highest information entropy values. Secondly, a novel scheme for defining crowd motion patterns (signatures) is devised to identify abnormal behaviors in the crowd by employing an enhanced gray level co-occurrence matrix model. In the experiments, various classic classifiers are utilized to benchmark the performance of the proposed method. The results obtained exhibit detection and accuracy rates which are, overall, superior to other techniques.

Abstract-In order to achieve automatic prediction and warning of hazardous crowd behaviors, a Spatio-Temporal Volume (STV) analysis method is proposed in this research to detect crowd abnormality recorded in CCTV streams. The method starts from building STV models using video data. STV slices -called Spatio-Temporal Textures (STT) -can then be analyzed to detect crowded regions. After calculating the Gray Level Co-occurrence Matrix (GLCM) among those regions, abnormal crowd behavior can be identified, including panic behaviors and other behavioral patterns. In this research, the proposed STT signatures have been defined and experimented on benchmarking video databases. The proposed algorithm has shown a promising accuracy and efficiency for detecting crowd-based abnormal behaviors. It has been proved that the STT signatures are suitable descriptors for detecting certain crowd events, which provide an encouraging direction for real-time surveillance and video retrieval applications. Keywords-Crowd abnormality, Spatio-Temporal Volume, STT Signature I. BACKGROUNDThe data obtained from CCTV cameras are 2D image sequences called frames. Each frame of these video data contains spatial information of various visual patterns. However, a single frame doesn't contain useful temporal information such as motions and trajectories of those patterns. In order to explore the temporal features, two or more consecutive frames need to be analyzed together. Horn and Schunck [1] first introduced Optical Flow Field to describe instant motion between two consecutive frames, fellow researchers [2] have conducted experiments using optical flow features to realize crowd abnormality detection. However, the computational complexity of optical flow is high, which limits its real-time performance. Other approaches such as motion field have been used to extract spatio-temporal information, yet suffering from object tracking problems caused by occlusion [3]. In this research, low-level and 2D texture features are utilized to avoid excessive computation because they only require pixel-wise calculation. This approach has ensured the computational time is independent from the actual crowds' density and event types. Following sections will explain how to reconstruct and utilize texture features for abnormal crowd behavior detection in details.

Abstract-With the purpose of achieving automated detection of crowd abnormal behavior in public, this paper discusses the category of typical crowd and individual behaviors and their patterns. Popular image features for abnormal behavior detection are also introduced, including global flow based features such as optical flow, and local spatio-temporal based features such as Spatio-temporal Volume (STV). After reviewing some relative abnormal behavior detection algorithms, a brandnew approach to detect crowd panic behavior has been proposed based on optical flow features in this paper. During the experiments, all panic behaviors are successfully detected. In the end, the future work to improve current approach has been discussed.

The research proposes an innovated structure of the density map-based crowd counting network augmented by information entropy. The network comprises of a front-end network to extract features and a back-end network to generate density maps. In order to validate the assumption that the entropy can boost the accuracy of density map generation, a multi-scale entropy map extraction process is imported into the front-end network along with a fine-tuned convolutional feature extraction process, In the back-end network, extracted features are decoded into the density map with a multi-column dilated convolution network. Finally, the decoded density map can be mapped as the estimated counting number. Experimental results indicate that the devised network is capable of accurately estimating the count in extremely high crowd density. Compared to similar structured networks which don’t adapt entropy feature, the proposed network exhibits higher performance. This result proves the feature of information entropy is capable of enhancing the efficiency of density map-based crowd counting approaches.