(a) Original image (b) Edge image (c) Valley image (d) Peak image
Figure 6.2 Energy fields over whole face image (Benn, 1999)
processing speed achieved via optical interconnect, but computers like that are not ready yet
(on our budgets at least). We can reduce the number of combinations by introducing constraints
on the relative size and position of the shapes, e.g. the circle should lie wholly within the
parabolas, but this will not reduce the number of combinations much. We can seek two alter-
natives: one is to use optimization techniques. The original approach (Yuille, 1991) favoured
the use of gradient descent techniques; currently, the genetic algorithm approach (Goldberg,
1988) seems to be most favoured and this has been shown to good effect for deformable
template eye extraction on a database of 1000 faces (Benn, 1999) (this is the source of the
images shown here). The alternative is to seek a different technique that uses fewer param-
eters. This is where we move to snakes, which are a much more popular approach. These
snakes evolve a s