Published Paper Details:

One of the main objectives of the study and application of video surveillance is the detection of anomalous events. Surveillance cameras are being used more often in public areas such as roadways, crossroads, banks, and shopping centers in an effort to increase public safety. One of the most important tasks in video surveillance is recognizing anomalous occurrences, such as crimes, traffic accidents, or illegal behavior. Abnormal events usually occur significantly less frequently than normal activities. The goal of a workable anomaly detection system is to identify the window of time when the anomaly is occurring and to rapidly notify users of behavior that deviates from expected patterns. Consequently, anomaly identification can be considered a fundamental method of comprehending videos by distinguishing abnormalities from typical patterns. After an anomaly has been discovered, it can be categorized into one of the specialist activities by applying classification techniques. An overview of anomaly detection is provided in this article, with a focus on applications related to banking operations. A wide range of stakeholders, including employees, clients, debtors, and outside parties, participate in or are impacted by the routine, recurrent, and ad hoc activities and transactions involved in banking operations. Although things could take time to work out, early detection can significantly reduce any potential bad effects and, in certain cases, even completely avoid them. Time series based anomaly detection is used to locate persons in undesired times. This paper identifies both common and uncommon events using an anomaly detection technique based on machine learning. The biometric identity of the face is captured and cross-referenced with faces of well-known criminals. If a match is found, it is easy to identify and capture the offenders