An interactive Case-Based Reasoning System for the development of Image Processing Applications

Valérie Ficet-Cauchard (1) , Christine Porquet (1) , Marinette Revenu (1)

1 Equipe Image - Laboratoire GREYC - UMR6072

In this paper, an interactive system for the development of Image Processing applications is described. This system is intended to provide some assistance to Image Processing experts during the construction as well as the execution of their applications. Through the system's graphical interface, an application can gradually be built and represented as a tree of tasks, and strategies can be made explicit so as to become reusable in future applications. In order to increase the assistance provided by the system, developers must be in a position to reuse parts of pre-existing treatments. To implement reusability, a Case-Based Reasoning approach is used. After defining two sets of criteria related to the task's description and the context of images, that enable us to classify a task in order to know when and how to reuse it, we describe our similarity functions and propose a search / adaptation algorithm for the retrieval of appropriate cases.

Domaines: Traitement des images [eess.IV]; Intelligence artificielle [cs.AI]

Valérie Ficet-Cauchard, Christine Porquet, Marinette Revenu. An interactive Case-Based Reasoning System for the development of Image Processing Applications. EWCBR'98, 1998, Dublin, Ireland. 11 p. ⟨hal-00869568⟩