Motion Objects Segmentation and Shadow Suppressing without Background Learning

Figure 5

Some results for the campus sequence. The first row is the current frame; the second row is the segmented masks.