Automated Video Surveillance at Night

ObjectVideo and Prof. Jianbo Shi from the University of Pennsylvania propose an automated activity recognition system for video surveillance at night. The deliverable is software that performs real-time threat analysis on incoming video streams and alertssecurity personnel of impending danger. The software will operate on legacy camera systems, including thermal, near-IR, and visible wavelength cameras. There are three key technical challenges. (1) Development of learning algorithms, so that the softwarecan automatically classify unusual behavior without user specification. (2) Development of suitable computer vision algorithms so that the system can hand off targets between multiple thermal cameras, i.e. without color information. (3) Development ofsuitable computer vision algorithms for robust video object detection, tracking, and classification that operate as well at night as during the day. ObjectVideo already has significant experience with computer vision-based automated video surveillancetechnologies and their application to real-world physical security and force protection challenges. Video is an excellent sensor modality for surveillance, physical security, and force protection. It is highly intuitive for a user, cheap, and widelyapplicable. Recent advances, such as low-light and thermal cameras, make video a viable option even at night. Video''s only drawback is that it is manually intensive for a human operator to monitor. Using computer vision technology, a computer canmonitor video signals and automatically detect threatening events. The benefits of the proposed system are that a nighttime video system will become a proactive security alarm system, alerting guards in real-time to unusual behavior and deterring crime,rather than a mere forensic tool, used for after-the-fact analysis.There are many commercial applications of this technology. Sensitive installations, such as port facilities, power utilities, and large estates, have expressed interest in: A guard that doesn't sleep and can monitor thousands of video feeds simultaneously. A system to detect unusual activity that humans may not be able to detect. For instance, while a human cannot easily detect a terrorist circling and surveying the facility, walking through multiple camera views, the proposed system can. A system that can detect unspecified, `unusual' behaviors and specified behaviors, such as loitering, dropping a suitcase, crossing a virtual tripwire, and entering through an exit.