**Aloimonos and
Basu 1986 :** The problem of
using image contours to infer the shape, orientation and motion of surfaces is
treated for the case of a binocular observer. In particular, the following
results are reported. a) A binocular observer understands the shape and depth
of a planar contour, without having to solve the point to point correspondence
problem between the left and right images of the contour. b) A binocular
observer understands the shape and depth of a planar contour, based on the
change of the perceived areas and perimeter between the left and the right
images of the contour under consideration. c) A binocular observer can recover
the 3-D motion of a moving planar contour, without having to solve he dynamic
or static correspondence problem, instead, the correspondence between the
different (dynamic) contour images ( as a whole) is required. d) Finally, some
of the above results are generalized for the case of a monocular observer. In
particular, a translating monocular observer can recover the shape of a contour
in view without using any point-to-point correspondences.