Automated Surveillance of Unattended Bags for Complex Situations

Unattended bags continues to be a threat for important places like government buildings, airports, subways, and secured areas as they have become key targets for terrorist attacks. Video surveillance has always been a vital component of security systems. Normally, these video recordings are checked only after a criminal activity is performed for obtaining evidences. In this research, an automated video surveillance system is proposed that identifies an unattended bag, generates an alarm signal and displays the image of the person who carried this bag into the scene. This solution improves the security effectiveness, while reducing the number of people needed to monitor video screens. This system also eliminates incorrect detections of other strangers walking near the bag or sitting near the bag for a short amount of time.