Lexicon-free handwritten word spotting using character HMMs

被引:193
作者
Fischer, Andreas [1 ]
Keller, Andreas [1 ]
Frinken, Volkmar [1 ]
Bunke, Horst [1 ]
机构
[1] Univ Bern, Inst Comp Sci & Appl Math, CH-3012 Bern, Switzerland
基金
瑞士国家科学基金会;
关键词
Handwriting recognition; Keyword spotting; Hidden Markov Models; HIDDEN MARKOV-MODELS; RECOGNITION; ALGORITHM;
D O I
10.1016/j.patrec.2011.09.009
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
For retrieving keywords from scanned handwritten documents, we present a word spotting system that is based on character Hidden Markov Models. In an efficient lexicon-free approach, arbitrary keywords can be spotted without pre-segmenting text lines into words. For a multi-writer scenario on the IAM off-line database as well as for two single writer scenarios on historical data sets, it is shown that the proposed learning-based system outperforms a standard template matching method. (C) 2011 Elsevier B.V. All rights reserved.
引用
收藏
页码:934 / 942
页数:9
相关论文
共 47 条
[1]   Word matching using single closed contours for indexing handwritten historical documents [J].
Adamek, Tornasz ;
O'Connor, Noel E. ;
Smeaton, Alan F. .
INTERNATIONAL JOURNAL ON DOCUMENT ANALYSIS AND RECOGNITION, 2007, 9 (2-4) :153-165
[2]  
[Anonymous], 1971, The SMART Retrieval System-Experiments in Automatic Document Processing
[3]  
[Anonymous], 2008, Proc. ICFHR
[4]   Special issue on the analysis of historical documents [J].
Antonacopoulos, Apostolos ;
Downton, Andy C. .
INTERNATIONAL JOURNAL ON DOCUMENT ANALYSIS AND RECOGNITION, 2007, 9 (2-4) :75-77
[5]  
Barrett C, 1997, COMPUT APPL BIOSCI, V13, P191
[6]  
Bhardwaj A., 2008, 2nd Intl Workshop on Cross Lingual Information Access, P48
[7]  
Bunke H, 2007, ADV PATTERN RECOGNIT, P165, DOI 10.1007/978-1-84628-726-8_8
[8]  
Cao HG, 2007, PROCEEDINGS OF THE SIXTH INTERNATIONAL CONFERENCE ON ADVANCES IN PATTERN RECOGNITION, P135
[9]  
Chan J., 2006, P IEEE C COMPUTER VI, P1455, DOI DOI 10.1109/CVPR.2006.269
[10]  
Chen F. R., 1993, ICASSP-93. 1993 IEEE International Conference on Acoustics, Speech, and Signal Processing (Cat. No.92CH3252-4), P1, DOI 10.1109/ICASSP.1993.319732