New Process Ontology-Based Character Recognition

We present in this paper, the general description of a classical process of character recognition, and subsequently we describe our contribution for a new process that is characterized by a step of creating of domain ontology, representing the Latin alphabet in the form of concepts, where each concept represents a grapheme (features). In this work we consider that there is a robust extraction module of primitives that will serve us to instantiate our ontology, by the different graphemes and their spatial relationships that exist in the processed document. Our ontology is generic and can support other languages by enriching it by new specific spatial relationships. Our process represents a new approach in the domain of character recognition, we have not yet developed a final product but our idea is promising in this domain.