Integrating video surveillance and building automation systems can improve security, ensure timelier maintenance, and use the potentially vast amount of video information for other whole-building automation schemes. In the latter applications, the video data needs to be further processed and filtered before being presented to human operators or other computer systems. This processing may allow the monitored video scene, regardless of venue or time of day, to be analyzed to detect suspicious objects or movement and then generate an appropriate alarm. The analysis even can be extended to detect the presence of small amounts of fire or smoke in large spaces before the danger can be detected by traditional flame or smoke sensors.