The research of moving object detection based on complex background

The complex background environment of the target detection has a great challenge. In this paper, a target detection method that based on the uncertainty of feature extraction and the distinction between the goals for the neighborhood is proposed. The method is based on Gaussian mixture model for background subtraction combined with edge detection. It uses Gaussian model for the background image model, and adopts background updating method, improves the stability and reliability of the algorithm. Experiments show that the method of illumination, background motion, foreground and background of each other has a strong adaptability.