Visual Attention and Motion Estimation-Based Video Retargeting for Medical Data Security

Figure 4

SIFT detection between adjacent frames. In (a)–(c), the green lines indicate the connection of feature points in the background, and the red lines indicate the connection of feature points in the foreground. All feature points in (a) and (b) belong to the background area. In (c), only one feature point is in the foreground area and the other feature points are in the background area.

(a)

(b)

(c)