A novel video object segmentation algorithm based on object structure and chromatic analysis in the rainy situations is proposed in this paper. Firstly, we separate the pixels of both foreground and background from the image sequence by using the change- detection approach. Thus, we can obtain the initial mask of object, which consists of shadow regions and reflective regions on the wet ground. The shadow region is caused of the object projection and mirroring on the wet ground generates the reflective region. Both of two regions can be removed by analyzing the structure and color of the moving object. Finally, the boundary of moving object is refined by using connecting component labeling and morphological operations. Experimental results show that our proposed algorithm can efficiently extract the moving objects, especially under the rainy environments.
展开▼