Object location in digital images is a wellknown
computer vision research topic. In the
so-called bin-picking problem, the position of a
specified object is to be determined in an image
with several objects. The common approach
is to describe both the parsed object model and
the segmented image in terms of features and their relationships (so-called relational descriptions
(Shapiro and Haralick, 1981)) and to find
the corresponding features by a comparison of the
attribute values of the features and their relationships.