A practical technique for moving object detection and tracking

摘要

A method of achieving background is presented which detect the moving object with statistics from complex scene. Compared with other methods of achieving image background, this approach can achieve and update simultaneously background. The method of moving object detection has the advantages of simplicity in computation, strong resistance to noise. The state transition matrix and observation matrix of Kalman filter are then defined to track the detected object. The experiment indicates that the method proposed has stability of forecast.