(a) Image A
(b) Image B
(c) Image C
Figure 4: Hashing distance for different regions with perceptual saliency corresponding to three images in Figure 3.