Motion Objects Segmentation and Shadow Suppressing without Background Learning

Figure 4

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