ABSTRACT
This chapter explains the reasoning behind the context-driven analysis and its most important basic principles. The Cognition Network Language (CNL) has been developed to enable context-driven analysis, in particular for images. In most of the image analysis tasks, the goal is to detect and quantify objects. Some simple context-driven procedures were already implemented early in the history of image analysis. In particular, in the image analysis of radiological images, such as CT images of humans, it was obvious that the spatial relations between the organs are quite well-defined and that one can make use of it for image analysis. Analyzing tissue slides also requires context-driven processing. It is, however, more difficult to define the network of context steps than for radiological scans. The chapter investigates the relation between the state-of-the-art image analysis techniques and CNL technology and describes the most important components of context-driven analysis by CNL.
