DocumentCode :
423785
Title :
Mixed Chinese/English document auto-processing based on the periodicity
Author :
Wang, Kai ; Jin, Jian-Ming ; Pan, Wu-Mo ; Shi, Guang-Shun ; Wang, Qing-Ren
Author_Institution :
Inst. of Machine Intelligence, Nankai Univ., Tianjin, China
Volume :
6
fYear :
2004
fDate :
26-29 Aug. 2004
Firstpage :
3616
Abstract :
A novel approach based on the periodicity is presented. Opposite to previous approaches, global characteristic is employed. To verify the effectiveness of new approach, two systems are implemented. Experiment shows that error rate drops from 0.163% to 0.104% when new algorithm is employed, more than 1/3 of errors are excluded.
Keywords :
document image processing; image segmentation; natural languages; optical character recognition; character segmentation; document autoprocessing; document image processing; global characteristic; language discrimination; mixed Chinese/English document; periodicity; Character recognition; Document image processing; Feature extraction; Globalization; Image segmentation; Machine intelligence; Natural languages; Optical character recognition software;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Machine Learning and Cybernetics, 2004. Proceedings of 2004 International Conference on
Print_ISBN :
0-7803-8403-2
Type :
conf
DOI :
10.1109/ICMLC.2004.1380423
Filename :
1380423
Link To Document :
بازگشت