So far we have shown methods to segment an image and obtain various regions that share similar characteristics. An important next step is to understand the shape, size and geometrical characteristics of these regions.
The regions in an image may be circular such as an image of coins or edges in a building. In some cases, the regions may not have simple geometrical shapes like circles, lines etc. Hence radius, slope etc. alone do not suffice to characterize the regions. An array of properties such as area, bounding box, central moments, centroid, eccentricity, euler number etc. are needed to describe shapes of regions.
In this chapter we begin the discussion with a label function that allows numbering ...