A Survey and Analysis of Crowd Anomaly Detection Techniques

In recent years, Crowd Anomaly detection has become a field of critical importance, especially so after the recent surge in technological development in the modern era. An increasing number of situations or places require surveillance which is also required to detect anomalies so as to prevent unforeseen accidents. Crowd Anomaly detection usually works by analyzing the surveillance scene in real time using various concepts and techniques such as AI, ML, optical flow analysis, streak flow analysis etc. to identify anomalies in a crowd of people. If an anomaly is found, the relevant authorities are immediately alerted. In this survey paper, few of these methodologies are analyzed by comparing their merits and demerits.