Graphical Knowledge Management in Graphics Recognition Systems

This paper deals with the problem of graphical knowledge management (formalization, modelling, representation and operationalization) in graphics recognition systems. We present here a “generic” formalism for graphical knowledge, allowing various modellings for a given graphical shape. We use a modelling library based on this formalism for the management of our graphical knowledge. The use of this library allows to request graphical knowledge databases, according to the processings’ requirements on graphical primitives. Like this, this approach allows interoperability between processings, especially for their combination. We present a “short” system use-case of our approach to illustrate the interoperability between processings.