Moving Object Tracking in Satellite Videos by Kernelized Correlation Filter Based on Color-Name Features and Kalman Prediction

Figure 4

Overview of the MOT in our experiments. There are six scenes of the satellite videos. In every scene, there is only one object. (a) The moving airplane-1 is selected as the object; (b) the moving airplane-2 is selected as the object; (c) the moving airplane-3 is selected as the object; (d) the moving airplane-4 is selected as the object; (e) the moving rocket is selected as the object; (f) the moving ship is selected as the object.

(a)

(b)

(c)

(d)

(e)

(f)