Survey of Anomaly Detection Techniques in Intelligent Surveillance Systems

Abstract

Smart surveillance systems are used to monitor specific regions such as houses, businesses, and borders. For a long time, abnormality detection in time-series data has been a significant study area. In today's society, unusual conduct signifies a threat or risk to others. An anomaly is something that deviates from what is anticipated, common, or usual. The most unexpected aspect of human behavior is determining if it is dubious or normal. It is extremely difficult to continually monitor public locations. Effective monitoring methods are critical to consistently increase the demand for successful inspection of open areas, handling difficulties such as traffic congestion, and ambiguities in crowded scenes. Handling millions of people during Hajj, intruder detection, fall detection, and so on. Raising knowledge of intelligent video surveillance technology and improving the degree of intelligent video surveillance has resulted in a reduction in data observation time. And analysis time has been reduced by half. Technology had advanced far too quickly. When introducing machine learning, artificial intelligence, and deep learning into the system, anomaly detection methods were developed using statistical models.