A hierarchical semantic video object tracking algorithm using mathematical morphology

In this paper, we propose a hierarchical segmentation method for tracking a semantic video object using mathematical morphology and watershed algorithm. At each hierarchy, we perform 3 basic steps: First, we extract markers on the simplified current frame. Second, region growing by a modified watershed algorithm is performed to get the segmented regions for the current frame. Third, we classify the regions into 3 categories, i.e., inside, outside and uncertain regions, by comparing the region pdf value with the predetermined threshold value. The region pdf value is acquired by the proposed pdf map which is calculated from a motion field estimated in forward direction. For the remaining uncertain regions at the end of a hierarchy, the above three steps are repeated at the lower hierarchy with less simplified frames until every region is decided to a certain region. The proposed algorithm provides prospective results in video sequences such as Miss America, Clair and Akiyo