Figure 10
From: Human Motion Analysis via Statistical Motion Processing and Sequential Change Detection
Handclapping-boxing video: (a) frame 30, (b) frame 100, (c) Activity Area of clapping and boxing, (d) likelihood ratio estimated for each pixel, (e) likelihood ratio values for all pixels, handwaving-handclapping video: (f) frame 10, (g) frame 254, (h) Activity Area, (i) likelihood ratio estimated for each pixel, (j) likelihood ratio values for all pixels, used for change detection.