Reseaech On Background Subtraction For Sports Video

Background subtraction with human motion monocular images is the first step of human motion analysis. The result of this step will affect the progress of other steps behind it and the final result of human motion analysis greatly. An important purpose of our research is to endow the computer with the ability being similar to the human vision. So, the computer can feel the motion object in the view and apprehend the behavior of the human more easily.This paper is mostly studying on the background subtraction with human motion monocular images. Based on discussing popular methods, the following several works have mainly been done in this paper, include:1,Studies on the representative theories and algorithms of the background subtraction with human motion monocular images. And this paper analyses the predominance and deficiency of these theory and algorithm. These algorithms include differential images, Running Gaussian average, the Mixture of Gaussians. The basic principle and steps of realization are expounded. Also the data of the evaluation are given.2,Analyses the peculiarity of the background subtraction with motion monocular images in the special scene of the ice stadium. The cause of the difference is that the lights are more easily reflected by the face of ice. The factors affected the result of subtraction include the reflected lights, shadows and inverted images.3,On the basic of studying on the representative theories and algorithms, Proposes an effective preprocessing algorithm to fit for the special scene of the ice stadium. The algorithm creates the background model with the improved Gaussian mixture model, and then brightness distortion and chromaticity distortion are introduced into the model to subtract the background.4,Resolve the problem of minute shift of background image caused by the movement of the camera. To prove if the background changed by computing the Harris corner points and matching image.5,Contrast the result of evaluation and analyze the data of evaluation. Evaluation shows that the proposed algorithm of background subtraction is highly effective and it can cast the reflected light, shadow and inverted image well. The algorithm adapts to the special scene of the ice stadium.