In this paper we experiment and validate an approach for the automatic generation of browsable technical documents. The application demonstrated in this work is restricted to cutaway diagrams with a particular visual representation. The global approach in which it is integrated is of a far more general interest, however. Our principal aim is to automatically generate an XML description of a graphical document that will allow for further use of its content by establishing links within the document or with other documents, thus constructing a corpus of interconnected documents sharing zones of similar content or semantics.
Index Terms:
Keywords: XML, document analysis, indexing, segmentation
Citation:
Ernest Valveny, Bart Lamiroy, "Scan-to-XML: Automatic Generation of Browsable Technical Documents," icpr, vol. 3, pp.30188, 16th International Conference on Pattern Recognition (ICPR'02) - Volume 3, 2002