A symbol classifier able to reject wrong shapes for document recognition systems

Abstract

We propose in this paper a new framework to develop a transparent classifier able to deal with reject notions. The generated classifier can be characterized by a strong reliability without loosing good properties in generalization. We show on a musical scores recognition system that this classifier is very well suited to develop a complete document recognition system. Indeed this classifier allows them firstly to extract known symbols in a document (text for example) and secondly to validate segmentation hypotheses. Tests had been successfully performed on musical and digit symbols databases. © Springer-Verlag Berlin Heidelberg 2000.