Improving Robustness and Accuracy in Moving Object Detection Using Section-Distribution Background Model
Yi Tang, Weiming Liu, Liang Xiong · 2008
In this paper, a new integrated approach for moving object detection is proposed. In initialization, a statistical algorithm is used to obtain the section-distribution background model which provides a scheme of choosing every parameter value in algorithm. The model is also updated real time in order to adapt to changes of illumination and objects in the scene. After applying a threshold to separate candidates in foreground and background, a shadow detection scheme is also introduced in this paper. It is based on HSV color space information and makes use of our background model. Finally, a comparison has been made among our algorithm and other algorithms. The results show improving robustness and accuracy of the model using our update algorithm.