Autofocus on Depth of Interest for 3D Image Coding

Figure 14

Visual quality comparison over depth map of UndoDancer view 1 frame 250 at 0.014 bpp: (a) original depth map, (b) input binary depth mask, (c) depth coded by classic LAR ( dB), and (d) depth coded with DoI refinement of depth zone within and with ( dB).

(a)

(b)

(c)

(d)