The concepts outlined in Chapter have
been applied to select preferable descriptors of shape [].
Selection of points that describe a given shape, as explained in Chapter
, was perpetually altered and evaluated
to find shape models and examples that require a smaller set of data
to be passed as an encoded message5.2.
To express the process at a moderate pace, each time points on the
curve that traces the shape are selected, a different model is ultimately
constructed. A good and compact statistical model is one whose legal
variations are relatively small and possibly so are the number of
its control points. Such a model is found using a general optimisation
regime under which points are reparameterised. MDL can be used as
a replacement for similarity in an objective function that is iteratively
evaluated for each such points reparameterisation. The minimisation
process will described in reasonable detail in Section
on optimisation. The more genuine part of this seminal work is the
use of an existing information theoretic measure, namely MDL, to guide
an autonomous search for good models. This work will be explained
with respect to current research in later chapters and especially
in Subsection sub:Returning-to-Shape which takes a more focused
scope. One alternative way of realising what this method is based
on is to look at its objective function.