Issues of Spatial Representation in Computational Vision

Zusammenfassung

Representation is an ubiquitous word which expresses multifaceted aspects of quite different, primarily non-technical, discourse domains such as philosophy, art, psychology and politics. Consequently one can hardly expect a general and exhaustive definition but instead (more or less hollow) phrases like „… a term with a broad general connotation of making present something or somebody that is not present …“ (Encyclopaedia Brittanica, 14th edition, Vol. 19, 1968, p. 152).² In the context of Artificial Intelligence, and particularly of computational vision, a crisp and widely accepted definition of what representation in general is meant to be has not been given up until now except for rather non-commital statements such as „… representation may be viewed as’ a set of conventions about how to describe things’ …“ (WESZKA 79, Chapter 3.a) or „A representation is a formal system for making explicit certain entities or types of information, together with a specification of how the system does this.“ (MARR 82, Chapter 1.2, p. 20).³

This contribution is a slightly extended and revised version of Chapter 4.3 in STIEHL 87.