The extraction of moving object in real-time Web-based video sequence
Xianping Fu, Zhao Wang, Dequn Liang, Jianping Jiang · 2005
This work presents an efficient moving object segmentation algorithm for real-time Web-based multimedia communication systems. To extract the moving objects regions from the sequence, the foreground of moving objects isn't used directly, but the background image is obtained firstly, based at motion information. This algorithm is motivated by the assumption that moving objects usually have intensity changes between successive frames. First, obtaining background mask technique is used to construct a reliable background image from the accumulated frame difference information between successive frames in a sequence. The moving object region is then separated from the background region by comparing the current frame with the constructed background image. Then, a postprocessing using morphological filter operation is used to remove noise regions and produce a more smooth shape boundary.