A segmentation method of merged numerals by wavelet analysis and GVF model

Author

Pan, Jian-jia ; Tang, Yuan-yan

Author_Institution

Dept. of Comput. Sci., Hong Kong Baptist Univ., Kowloon

Volume

2

fYear

2008

fDate

30-31 Aug. 2008

Firstpage

502

Lastpage

506

Abstract

Aiming for some limit of the recognition-free approach to the segment merged numerals, a new segmentation method is proposed based on GVF mode in this paper. Firstly, this paper presents a new segment model based on wavelet multi-scale analysis and GVF model, and then we apply this energy model to segment the merged numerals. This method got the candidate segment line by contour method, and then confirmed the segment path based on energy model. The segmented digits have a more natural shape than others.

Keywords

handwriting recognition; image recognition; image segmentation; wavelet transforms; GVF model; contour method; handwriting; merged numerals; recognition-free approach; segmentation method; wavelet analysis; wavelet multiscale analysis; Active contours; Character recognition; Computer science; Handwriting recognition; Image segmentation; Pattern analysis; Pattern recognition; Shape; Sorting; Wavelet analysis; GVF Model; Merged Numeral; Segmentation; Wavelet Analysis;

fLanguage

English

Publisher

ieee

Conference_Titel

Wavelet Analysis and Pattern Recognition, 2008. ICWAPR '08. International Conference on

Conference_Location

Hong Kong

Print_ISBN

978-1-4244-2238-8

Electronic_ISBN

978-1-4244-2239-5

Type

conf

DOI

10.1109/ICWAPR.2008.4635832

Filename

4635832