Moving Object Detection for Dynamic Background Scenes Based on Spatiotemporal Model

Figure 7

Comparison of the detection results. (a) Input frames of the six videos of “dynamic background” in changedetection.net, and they are the 2000th frame in “Boats,” the 955th frame in Canoe, the 1892th frame in Fall, the 1147th frame in “Fountain01,” the 745th frame in “Fountain02,” and the 2462th frame in “Overpass” from top to bottom. (b) Ground truth of (a). (c) Results of ViBe. (d) Results of CodeBook. (e) Results of the proposed method.

(a)

(b)

(c)

(d)

(e)