Traffic Anomalies in Surveillance Videos: Recent Trends
V. Prasath, K. Deepak, S. Chandrakala · Journal of Computational and Theoretical Nanoscience · 2020
There is an increasing demand for automated traffic surveillance with a wide range of threats in road safety and less man power to monitor them. Especially, detecting anomalous behavior in traffic surveillance is challenging because of the presence of occlusion, weather conditions, and pose variations. This paper reviews the recent trends on vision based traffic anomaly detection. Various features, modeling techniques and data sets used in traffic anomaly detection are reviewed.