Mathematical symbol recognition with support vector machines

Christopher Malon, Seiichi Uchida, Masakazu Suzuki

Research output: Contribution to journalArticle

32 Citations (Scopus)

Abstract

Single-character recognition of mathematical symbols poses challenges from its two-dimensional pattern, the variety of similar symbols that must be recognized distinctly, the imbalance and paucity of training data available, and the impossibility of final verification through spell check. We investigate the use of support vector machines to improve the classification of InftyReader, a free system for the OCR of mathematical documents. First, we compare the performance of SVM kernels and feature definitions on pairs of letters that InftyReader usually confuses. Second, we describe a successful approach to multi-class classification with SVM, utilizing the ranking of alternatives within InftyReader's confusion clusters. The inclusion of our technique in InftyReader reduces its misrecognition rate by 41%.

Original languageEnglish
Pages (from-to)1326-1332
Number of pages7
JournalPattern Recognition Letters
Volume29
Issue number9
DOIs
Publication statusPublished - Jul 1 2008

Fingerprint

Support vector machines
Optical character recognition
Character recognition

All Science Journal Classification (ASJC) codes

  • Software
  • Signal Processing
  • Computer Vision and Pattern Recognition
  • Artificial Intelligence

Cite this

Mathematical symbol recognition with support vector machines. / Malon, Christopher; Uchida, Seiichi; Suzuki, Masakazu.

In: Pattern Recognition Letters, Vol. 29, No. 9, 01.07.2008, p. 1326-1332.

Research output: Contribution to journalArticle

Malon, Christopher ; Uchida, Seiichi ; Suzuki, Masakazu. / Mathematical symbol recognition with support vector machines. In: Pattern Recognition Letters. 2008 ; Vol. 29, No. 9. pp. 1326-1332.
@article{e30319e4a2c949a1915b8f90aaf7c38e,
title = "Mathematical symbol recognition with support vector machines",
abstract = "Single-character recognition of mathematical symbols poses challenges from its two-dimensional pattern, the variety of similar symbols that must be recognized distinctly, the imbalance and paucity of training data available, and the impossibility of final verification through spell check. We investigate the use of support vector machines to improve the classification of InftyReader, a free system for the OCR of mathematical documents. First, we compare the performance of SVM kernels and feature definitions on pairs of letters that InftyReader usually confuses. Second, we describe a successful approach to multi-class classification with SVM, utilizing the ranking of alternatives within InftyReader's confusion clusters. The inclusion of our technique in InftyReader reduces its misrecognition rate by 41{\%}.",
author = "Christopher Malon and Seiichi Uchida and Masakazu Suzuki",
year = "2008",
month = "7",
day = "1",
doi = "10.1016/j.patrec.2008.02.005",
language = "English",
volume = "29",
pages = "1326--1332",
journal = "Pattern Recognition Letters",
issn = "0167-8655",
publisher = "Elsevier",
number = "9",

}

TY - JOUR

T1 - Mathematical symbol recognition with support vector machines

AU - Malon, Christopher

AU - Uchida, Seiichi

AU - Suzuki, Masakazu

PY - 2008/7/1

Y1 - 2008/7/1

N2 - Single-character recognition of mathematical symbols poses challenges from its two-dimensional pattern, the variety of similar symbols that must be recognized distinctly, the imbalance and paucity of training data available, and the impossibility of final verification through spell check. We investigate the use of support vector machines to improve the classification of InftyReader, a free system for the OCR of mathematical documents. First, we compare the performance of SVM kernels and feature definitions on pairs of letters that InftyReader usually confuses. Second, we describe a successful approach to multi-class classification with SVM, utilizing the ranking of alternatives within InftyReader's confusion clusters. The inclusion of our technique in InftyReader reduces its misrecognition rate by 41%.

AB - Single-character recognition of mathematical symbols poses challenges from its two-dimensional pattern, the variety of similar symbols that must be recognized distinctly, the imbalance and paucity of training data available, and the impossibility of final verification through spell check. We investigate the use of support vector machines to improve the classification of InftyReader, a free system for the OCR of mathematical documents. First, we compare the performance of SVM kernels and feature definitions on pairs of letters that InftyReader usually confuses. Second, we describe a successful approach to multi-class classification with SVM, utilizing the ranking of alternatives within InftyReader's confusion clusters. The inclusion of our technique in InftyReader reduces its misrecognition rate by 41%.

UR - http://www.scopus.com/inward/record.url?scp=43249087434&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=43249087434&partnerID=8YFLogxK

U2 - 10.1016/j.patrec.2008.02.005

DO - 10.1016/j.patrec.2008.02.005

M3 - Article

VL - 29

SP - 1326

EP - 1332

JO - Pattern Recognition Letters

JF - Pattern Recognition Letters

SN - 0167-8655

IS - 9

ER -