Abstract.
This paper describes a method for estimating a confidence value (CV) by which we can express the potential correctness of handwritten Kanji character recognition candidates. An accumulated confidence value (ACV), calculated as the sum of CVs, is also applied to reduce the number of candidates. Such reduction is vital to increasing the speed of such applications as Kanji address recognition, and it also reduces the probability of misreadings in linguistic postprocessing. Sorted sets of character candidates, ranked in increasing order of each candidate’s distance value, are used as feature vectors. A CV is defined as the a posteriori probability with respect to each rank. To obtain good quality approximations of probability density functions (PDFs), we introduce a subspace within which correct data can easily be separated from erroneous data and then estimate PDF parameters over this subspace. Next, we use an ACV as a measure for expressing a threshold for candidate acceptance in Kanji character recognition. The efficiency of the proposed method is evaluated in an experiment using IPTP CD-ROM2 Japanese address images, and a comparison with the results for a conventional method shows that a roughly 35% reduction in the number of candidates is obtained without reducing the number of correct candidates.
Similar content being viewed by others
References
Akiyama K (1996) A new reject decision method for statistical pattern recognition. In: Proceedings of the 5th international workshop on frontiers in handwritting recognition, Colchester, UK, 2 September 1996, pp 239-242
Bouchaffra D, Govindaraju V, Srihari SN (1999) A methodology for mapping scores to probabilities. IEEE Trans Patt Anal Mach Intell 21(9):923-927
Dempster A, Laird N, Rubin D (1977) Maximum likelihood from incomplete data via the EM algorithm. J R Stat Soc Ser B Methodol 39(1):1-38
Fukushima T, Shimomura H, Mori Y (1995) An address elements search algorithm for a handwritten address reader. In: Proceedings of the IPSJ annual conference, Tokyo, 15 March 1995, 4D-6:65-66 (in Japanese)
Hamanaka M, Yamada K, Tsukumo J (1993) Normalization-cooperated feature extraction method for handprinted Kanji character recognition. In: Proceedings of the 3rd international workshop on frontiers in handwritting recognition, Buffalo, NY, 25 May 1993, pp 343-348
Huang YS, Suen CY (1995) A method of combining multiple experts for the recognition of unconstrained handwritten numerals. IEEE Trans Patt Anal Mach Intell 17(1):90-94
Ishidera E, Nishiwaki D, Yamada K (1997) Unconstrained Japanese Address Recognition Using a Combination of Spatial Information and Word Knowledge. In: Proceedings of the 4th international conference on document analysis and recognition, Ulm, Germany, 18 August 1997, 2:1016-1022
Kittler J, Hatef M, Duin RPW, Matas J (1998) On combining classifiers. IEEE Trans Patt Anal Mach Intell 20(3):226-239
Lin X, Ding X, Chen M, Zhang R, Wu Y (1998) Adaptive confidence transform based classifier combination for chinese character recognition. Patt Recog Lett 19(10):975-988
Liu C, Nakagawa M (2000) Precise candidate selection for large character set recognition by confidence evaluation. IEEE Trans Patt Anal Mach Intell 22(6):636-642
Liu C, Koga M, Fujisawa H (2001) Lexicon-driven handwritten character string recognition for Japanese address reading. In: Proceedings of the 6th international conference on document analysis and recognition, Seattle, 10 September 2001, pp 877-881
Nakayama Y, Yokozuka S (1995) A study on a certainty of character recognition. In: Proceedings of the 1995 IEICE general conference, Fukuoka, Japan, 27 March 1995, D-541, p 267 (in Japanese)
Rumelhart D, Hinton G, Williams R (1986) Learning internal representations by backpropagating errors. In: Nature 323(99):533-536
Sato A, Yamada K (1998) A formulation of learning vector quantization using a new misclassification measure. In: Proceedings of the 14th international conference on pattern recognition, Brisbane, Australia, 1:322-325
Shürmann J (1996) Pattern classification: a unified view of statistical and neural approaches. Wiley, New York
Tsutsumida T, Kawamata F, Yamaguchi S, Nagata K, Wakahara T (1996) The third IPTP character recognition competition and study on multi-expert systems for handwritten Kanji recognition. In: Proceedings of the 5th international workshop on frontiers in handwriting recognition, Colchester, UK, 2 September 1996, pp 479-482
Author information
Authors and Affiliations
Corresponding author
Additional information
Received: 29 October 2001, Accepted: 30 September 2003, Published online: 1 April 2004
Correspondence to: Eiki Ishidera
Rights and permissions
About this article
Cite this article
Ishidera, E., Nishiwaki, D. & Sato, A. A confidence value estimation method for handwritten Kanji character recognition and its application to candidate reduction. IJDAR 6, 263–270 (2003). https://doi.org/10.1007/s10032-003-0118-8
Issue Date:
DOI: https://doi.org/10.1007/s10032-003-0118-8