Graphics Reference
In-Depth Information
Figure 7.25. Four cameras observe the striped object. The visual hull is the shaded region
formed by intersecting the backprojected silhouettes (gray lines on the image planes). The visual
hull is always larger than the actual object.
(a)
(b)
Figure 7.26. (a) Images of a mannequin acquired by seven calibrated cameras, and the corre-
sponding automatically extracted silhouettes. (b) The resulting visual hull from a frontal and side
view. The coarse 3D approximation is reasonable, but more cameras would be required to carve
away the extraneous voxels.
Search WWH ::




Custom Search