Design of a log sorting X-ray system based on domain ontologies
Résumé
This study aims to highlight gains due to establishment and use of trade ontologies in order to design and improve the functional model of an inspection system using X-ray tomography. The application case study is an automated vision system allowing qualitative enhancement of sawmill stem supply. The goal is reached with company trade knowledge modeling according to final product quality. NIAM-ORM modeling method is used to obtain ontological models of Lumber quality and Log quality. The confrontation of both models by knowledge engineers allows creation of genetic model and then functional model. The study suggests also an aggregation of a wood matter model in order to enhance the functional model. So, it lets easily improve the global system by clarifying knowledge hold by the company, favoring communication between the vision system integrator (supplier) and the company (customer) with a better comprehension of needs and issues associated. It gives a structured expert knowledge of the company that the integrator can easily reuse in his system.