ABSTRACT
Measurements that can be performed on each of the individual features in an image may be grouped for convenience into four general classes: brightness (including color values and parameters such as density that can be related to brightness), location (both absolute position and relative to other features present), size, and shape. Within each class, quite a variety of specific measurements are possible, and there are also several ways to perform the operations, with results that are not always the same. Most image analysis systems offer at least a few measures in each class. Users find themselves at some time or another having to select from several different measurement parameters. The problem is frequently to decide which of the measured parameters is most useful or appropriate for solving a particular problem.
