| Presentation | 2006-10-20 A Method of Logical Structure Analysis for Form Images with Various Layouts by Belief Propagation Akihiro Minagawa, Yusaku Fujii, Hiroaki Takebe, Katsuhito Fujimoto, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In this paper, we propose a method for analyzing logical structure of an input form image with various layouts on a form category, such as an invoice, application form, etc. A generic logical structure, representing a set of logical elements and relationships between them, is also input to our method, and a word corresponding to each logical element is extracted in the input form image. Advantages of our method are as follows: First, simple and intuitive graphical model corresponding to the generic logical structure is constructed, by representing character information and its position information as relationships between a word and a logical elements and relationships between words conditioned by a set of logical elements, respectively. Second, the specific logical structure for the input form with various layouts can be analyzed independent of its layout structure by using belief propagation on the above graphical model. Finally, the logical structure can be analyzed robustly to characters/words recognition errors and missing elements. Effectiveness of this method was confirmed by the experiment for three categories of forms. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | form recognition / optical character recognition (OCR) / logical structure / logical structure analysis / graphical model / belief propagation (BP) |
| Paper # | PRMU2006-107 |
| Date of Issue |
| Conference Information | |
| Committee | PRMU |
|---|---|
| Conference Date | 2006/10/13(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 | Pattern Recognition and Media Understanding (PRMU) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | A Method of Logical Structure Analysis for Form Images with Various Layouts by Belief Propagation |
| Sub Title (in English) | |
| Keyword(1) | form recognition |
| Keyword(2) | optical character recognition (OCR) |
| Keyword(3) | logical structure |
| Keyword(4) | logical structure analysis |
| Keyword(5) | graphical model |
| Keyword(6) | belief propagation (BP) |
| 1st Author's Name | Akihiro Minagawa |
| 1st Author's Affiliation | Fujitsu Laboratories Ltd.() |
| 2nd Author's Name | Yusaku Fujii |
| 2nd Author's Affiliation | Fujitsu Laboratories Ltd. |
| 3rd Author's Name | Hiroaki Takebe |
| 3rd Author's Affiliation | Fujitsu Laboratories Ltd. |
| 4th Author's Name | Katsuhito Fujimoto |
| 4th Author's Affiliation | Fujitsu Laboratories Ltd. |
| Date | 2006-10-20 |
| Paper # | PRMU2006-107 |
| Volume (vol) | vol.106 |
| Number (no) | 301 |
| Page | pp.pp.- |
| #Pages | 6 |
| Date of Issue | |