Figure 10From: Pattern-based video coding with dynamic background modeling Generated pattern templates using the first three decoded frames of each video sequence. QP = 30 where P i (i = 1, 2,…, 8) indicates the pattern number.Back to article page