Physical layout analysis of complex structured arabic documents using artificial neural nets

Abstract

This paper describes PLANET, a recognition method to be applied on Arabic documents with complex structures allowing incremental learning in an interactive environment. The classification is driven by artificial neural nets each one being specialized in a document model. The first prototype of PLANET has been tested on five different phases of newspaper image analysis: thread recognition, frame recognition, image text separation, text line recognition and line merging into blocks. The learning capability has been tested on line merging into blocks. Some promising experimental results are reported. © Springer-Verlag 2004.