ABSTRACT
Template matching and cross-correlation Recognition and classification are essentially complementary functions that lie at the “high end” (i.e., that require the most complicated algorithms) in the field of image analysis (Duda and Hart 1973). Classification is concerned with establishing criteria that can be used to identify or distinguish different populations of objects that appear in images. These criteria can vary widely in form and sophistication, ranging from example images of prototypical representatives of each class, to numeric parameters for measurement, to syntactical descriptions of key features. Recognition is the process by which these tools are subsequently used to find a particular feature within an image. It functions at many different levels, including such different processes as finding a face in an image, or matching that face to a specific individual.
