Presentation 2002/3/7
Document recognition by real-time classifications of character images and reduction of correction labor of recognition results
Eriko ANDO, Masakazu SUZUKI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In the process of the recognition of documents, we perform a real-time classification of character images. After the classification, the result of character recognition is determined by majority in each class, so that the character recognition rate is improved. We propose two methods of reduce the labor to correct the recognition results and report the experimets to estimate the performance. One of the method is the automatic correction method for the caracters belonging to a same class. When a user correct a character, then all the other characters belonging to the same class are corrected simultaneously. In the second method, after the classification of character images for all the pages of the volume to be recognized, the system lists the classification results of the characters and present a method of cluster-wise correction of the recognition results, before performing the structure analysis of mathematical expressions, to improve the efficiency of digitization work of mathematical documents.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) OCR / clustering / reduction of correction labor of recognition results
Paper # TL2001-62
Date of Issue

Conference Information
Committee TL
Conference Date 2002/3/7(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Thought and Language (TL)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Document recognition by real-time classifications of character images and reduction of correction labor of recognition results
Sub Title (in English)
Keyword(1) OCR
Keyword(2) clustering
Keyword(3) reduction of correction labor of recognition results
1st Author's Name Eriko ANDO
1st Author's Affiliation Faculty of Mathematics, Kyushu University()
2nd Author's Name Masakazu SUZUKI
2nd Author's Affiliation Graduate School of Mathematics, Kyushu University
Date 2002/3/7
Paper # TL2001-62
Volume (vol) vol.101
Number (no) 710
Page pp.pp.-
#Pages 8
Date of Issue