Абстрактный

The subtraction in the application of taekwondo image information based moving objects detection algorithms

Li Chen1, Yan Dai

In the case that the camera is fixed, the most common method which is used in carrying out the real-time detection for the moving object of the image sequence is the background subs traction. This algorithm can be estimated the background model without a moving target. The position of the moving object is determined according to calculating the difference between the image frames and the background model, and the detection result is used to update the background model. In a variety of background subtraction algorithms, themain difference lies in adopting the background model and updating algorithm. At present, as for the background model, the statistical model is mostly used to describe the probability distribution of the brightness, and the most practical application is normal probability distribution. As for the background updating algorithm, the different test detection results is basically given the different coefficients to distinguish tending to retain or change the original distribution.