Share Email Print

Proceedings Paper

Handwritten text line segmentation by spectral clustering
Author(s): Xuecheng Han; Hui Yao; Guoqiang Zhong
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Since handwritten text lines are generally skewed and not obviously separated, text line segmentation of handwritten document images is still a challenging problem. In this paper, we propose a novel text line segmentation algorithm based on the spectral clustering. Given a handwritten document image, we convert it to a binary image first, and then compute the adjacent matrix of the pixel points. We apply spectral clustering on this similarity metric and use the orthogonal kmeans clustering algorithm to group the text lines. Experiments on Chinese handwritten documents database (HIT-MW) demonstrate the effectiveness of the proposed method.

Paper Details

Date Published: 8 February 2017
PDF: 6 pages
Proc. SPIE 10225, Eighth International Conference on Graphic and Image Processing (ICGIP 2016), 102251A (8 February 2017); doi: 10.1117/12.2266982
Show Author Affiliations
Xuecheng Han, Ocean Univ. of China (China)
Hui Yao, Ocean Univ. of China (China)
Guoqiang Zhong, Ocean Univ. of China (China)

Published in SPIE Proceedings Vol. 10225:
Eighth International Conference on Graphic and Image Processing (ICGIP 2016)
Yulin Wang; Tuan D. Pham; Vit Vozenilek; David Zhang; Yi Xie, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?