Real-time event detection system for intelligent video surveillance

• Main Authors: Chua, Timothy John A., Co, Andrew Jonathan W., Ilustre, Paolo Javier S.
• Format: text
• Language: English
• Published: Animo Repository 2006
• Subjects: Video recordings > Production and direction; Video recordings industry; Surveillance radar; Electronic surveillance
• Online Access: https://animorepository.dlsu.edu.ph/etd_honors/246
• Institution: De La Salle University

The use of CCTV cameras in various locations brings about an increase in security. Unfortunately, it is quite tedious for security personnel to monitor 50-100 live feeds simultaneously and for extended periods of time. To solve this problem, the use of automated surveillance must be considered. In this undertaking, a real-time event detection system is created. The system takes live feeds from an analog camera (in frame form), via a digitizer connected to the USB port of a computer. Various image processing techniques are then applied to the frame in order to separate the foreground from the background. Useful information is then extracted from the foreground and is used in either identifying objects of interest, creating correspondence between the consecutive frames or analyzing the behavior of the object itself. Upon the detection of commissions of suspicious actions, also referred to as alarm events, the system notes the occurrence by means of a log, which is visible through the graphical user interface (GUI) of the system.