Reading handwritten German words in historical documents

Author

Steinke, K. ; Yuanchen Zhang

Author_Institution

Univ. of Appl. Sci. & Arts, Hanover, Germany

fYear

2012

fDate

16-18 Oct. 2012

Firstpage

1294

Lastpage

1298

Abstract

The research project “Herbar Digital” was started in 2007 with the aim to digitize 3.5 million dried plants on paper sheets belonging to the Botanic Museum Berlin in Germany. Frequently there are printed labels on the sheets with handwritten annotations. The annotations are written by the determiner or the finder of the plant. They often describe the plant and give information about its name and where it was found. So procedures have to be developed in order to read the most important handwritten words on the sheets. A HMM-approach, a Fourier-approach and a DTW-approach are compared. With a limited number of words a recognition rate of about 95% is obtained by all three methods.

Keywords

Fourier series; handwriting recognition; handwritten character recognition; hidden Markov models; history; museums; text analysis; word processing; Botanic Museum Berlin; DTW approach; Fourier approach; Germany; HMM approach; Herbar Digital project; dried plant digitization; dynamic time warping; handwritten German word reading; handwritten annotations; historical documents; paper sheets; Estimation; Feature extraction; Handwriting recognition; Hidden Markov models; Time series analysis; Training; DTW; Fourier; HMM; handwriting recognition; writer recognition;

fLanguage

English

Publisher

ieee

Conference_Titel

Image and Signal Processing (CISP), 2012 5th International Congress on

Conference_Location

Chongqing

Print_ISBN

978-1-4673-0965-3

Type

conf

DOI

10.1109/CISP.2012.6469910

Filename

6469910