HAL will be down for maintenance from Friday, June 10 at 4pm through Monday, June 13 at 9am. More information
Skip to Main content Skip to Navigation
Conference papers

Introduction of statistical information in a syntactic analyser for document image recognition

André O. Maroneze 1 Bertrand Coüasnon 1 Aurélie Lemaitre 1
1 IMADOC - Interprétation et Reconnaissance d’Images et de Documents
UR1 - Université de Rennes 1, INSA Rennes - Institut National des Sciences Appliquées - Rennes, CNRS - Centre National de la Recherche Scientifique : UMR6074
Abstract : This paper presents an improvement to document layout analysis systems, oering a possible solution to Sayre's paradox (which states that an element must be recognized before it can be segmented; and it must be segmented before it can be recognized). This improvement, based on stochastic parsing, allows integration of statistical information, obtained from recognizers, during syntactic layout analysis. We present how this fusion of numeric and symbolic information in a feedback loop can be applied to syntactic methods to improve document description expressiveness. To limit combinatorial explosion during exploration of solutions, we devised an operator that allows optional activation of the stochastic parsing mechanism. Our evaluation on 1250 handwritten business letters shows this method allows the improvement of global recognition scores.
Document type :
Conference papers
Complete list of metadata

Contributor : Aurélie Lemaitre Connect in order to contact the contributor
Submitted on : Friday, February 18, 2011 - 10:03:05 AM
Last modification on : Thursday, January 7, 2021 - 4:35:32 PM
Long-term archiving on: : Tuesday, November 6, 2012 - 2:15:52 PM


Files produced by the author(s)


  • HAL Id : hal-00567077, version 1


André O. Maroneze, Bertrand Coüasnon, Aurélie Lemaitre. Introduction of statistical information in a syntactic analyser for document image recognition. Document recognition and Retrieval XVIII - Electronic Imaging, Jan 2011, San Francisco, United States. pp.7874 04. ⟨hal-00567077⟩



Record views


Files downloads