With the video surveillance system developing to digital and intelligent direction gradually in recent years, intelligent video surveillance system more and more often appeared in a number of applications. It not only enhanced the automation of traditional video, but also made important contributions to the economic development, the national defense and the social order.This paper researched on a simple intelligent video surveillance and alarm system, mainly analyzed and researched on the moving object detecting and tracking algorithm. As the background that video surveillance is vulnerable to interference, the paper proposed to establish a mixed-Gaussian model with parameters of color, hue, value, to detect the foreground of video images. It effectively eliminated the vibration disturbance in the video surveillance background, and light influence produced by optical change. The paper also discussed the theory and realization process of Kalman filter track algorithm and Mean Shift emphatically algorithm. The system combined Intel Corporation's OpenCV Library to the MFC applications, and had realized the moving object detecting and tracking by using the re-development functions of OpenCV. OpenCV had greatly simplified the research and development process, and reduced the system's work, because of its high degree packaging in video data.When detected a moving object, the system retained the sound and light. Simultaneously, it applied wireless communication technology into the system, used Siemens modem module to realize a kind of wireless communication warning way, provided a new alarm method to the system. |