Presentation 2006-02-23
Form Layout Analysis Based on Size Invariant Features
Tomohisa SUZUKI, Akihiro UDA, Hiroyuki MIZUTANI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) One of the important requirements of the character recognition for printed forms is to find locations of fields to be recognized. In the most conventional method designed for OCR sheets reading, these fields are located by predefined coordinates. Nevertheless, this conventional method is not appropriate for non-OCR sheets, due to inexact format designs. In this paper, we introduce a new algorithm suitable for non-OCR forms. We exploit structural features of printed form for this purpose. A dynamic programming technique is adopted for form and template matching with cost function defined as difference of structural feature vectors. The experiments show 90% correct results in locating process for field extraction with 200 test sheets.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Form / Layout / Size
Paper # TL2005-49,PRMU2005-184
Date of Issue

Conference Information
Committee TL
Conference Date 2006/2/16(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) Form Layout Analysis Based on Size Invariant Features
Sub Title (in English)
Keyword(1) Form
Keyword(2) Layout
Keyword(3) Size
1st Author's Name Tomohisa SUZUKI
1st Author's Affiliation Toshiba Solutions Corporation()
2nd Author's Name Akihiro UDA
2nd Author's Affiliation Toshiba Solutions Corporation
3rd Author's Name Hiroyuki MIZUTANI
3rd Author's Affiliation Toshiba Solutions Corporation
Date 2006-02-23
Paper # TL2005-49,PRMU2005-184
Volume (vol) vol.105
Number (no) 612
Page pp.pp.-
#Pages 6
Date of Issue