Realization of event detection control on underground metro lines by video analysis method

Abstract

In recent years, both academic and technological studies related to security practices have been one of the most widely mentioned topics. There are many different studies for security applications where images taken from the camera are analyzed. Even the recent epidemic, etc. It can be seen that by using camera images to help diagnose diseases, people are leading new applications in the field of use regarding the possibility of having any disease. In this study, a potential application study was carried out to allow users to travel safely by making video analysis studies with the images taken from the cameras used for security purposes in metro lines. In the video analysis method, specific scenarios were defined to the system with various algorithms written in the software and event detection control was performed. Motion detection, steady state detection, identification of suspicious packages, panic-rush situations, suspicious facial expressions and unusual situations crowded subway in video analysis methods used in line. The general working principles of the given analysis methods are specified and especially motion detection and suspicious facial recognition technology defined as a scenario to the system are emphasized. By sending e-mails to the scenarios realized with the video analysis method, it is provided to automatically report the incidents on the system. With this study, it is thought that it is a potential study to be used in event detection in transportation vehicles in crowded cities and in many similar areas. It has been determined that the scenarios and approaches specified in the study are redesigned for different areas and topics, and it is likely that similar successful studies will continue.