Handwriting knowledge based on parameterization for writer identification

Título:	Handwriting knowledge based on parameterization for writer identification
Autores/as:	Romero, Carlos F. Travieso, Carlos M. Ferrer, Miguel A. Alonso, Jesús B.
Clasificación UNESCO:	3307 Tecnología electrónica
Fecha de publicación:	2009
Editor/a:	1876-1100
Publicación seriada:	Lecture Notes in Electrical Engineering
Conferencia:	European Computing Conference
Resumen:	This present paper has worked out and implemented a set of geometrical characteristics from observation of handwriting. In particular, this work has developed a proportionality index together with other parameters applied to handwritten words, and they have been used for writer identification. That set of characteristics has been tested with our offline handwriting database, which consists of 40 writers with 10 samples per writer. We have got a success rate of 97%, applying a neural network as classifier.
URI:	http://hdl.handle.net/10553/44073
ISBN:	9780387848136
ISSN:	1876-1100
DOI:	10.1007/978-0-387-84814-3_1
Fuente:	Lecture Notes in Electrical Engineering[ISSN 1876-1100],v. 27 LNEE, p. 3-13
Colección:	Actas de congresos

Vista completa