Figure 10From: Robust Recognition of Specific Human Behaviors in Crowded Surveillance Video Sequences Target regions and 2D color histograms. Back to article page