A computational framework for the flexible packing of arbitrary planar
shapes under visual control is described. Our aim in this work has be
en to produce an efficient packing strategy that is flexible enough fo
r a wide variety of industrial uses and which can be implemented in fa
st, moderately priced hardware. We have deliberately adopted a systems
approach, versus a purely algorithmic one, since we are concerned wit
h industrial vision problems in which significant problem constraints
exist. The packing procedure that we devised consists of two major com
ponents. The first is a geometric packing technique that is based on m
orphological image processing operations. This is used in conjunction
with a heuristic packing procedure. Some of the factors considered at
the heuristic level include shape ordering and shape orientation, both
of which must be carried out prior to applying the geometric packer.
The heuristic procedure also deals with problem constraints that are s
pecific to a given application. Various issues arising from this appro
ach, such as the properties and performance of the procedure, are disc
ussed within the background of some sample applications. The ideas out
lined are currently being used in the development of a visually contro
lled intelligent packing work cell.