Silhouettes contain much of the shape information necessary to recognise simple objects. For this reason they have been used widely in machine vision systems that check the shape of industrial parts or provide grasping information to robots. The digital version of a silhouette is a binary image i.e. an array of 1's and 0's. Figure 1 shows a digital image. In a binary image version of this, 0 represents a dark area and 1 represents a light area. The positions of pixels are defined with respect to the coordinate system shown in the figure.
A significant problem of image processing is the large amount of storage needed for image data
and the consequent computational load of performing calculations on large numbers of pixels.
There has therefore been much effort in developing data representations or codings which are
more compact but still capture the relevant features of the data. One promising data structure is
the quadtree. The quadtree encoding of the binary image shown in Figure 1 is given in Figure 2.