Automatic content extraction of filled form images based on clustering component block projection vectors

Hanchuan Peng; Xiaofeng He; Fuhui Long

doi:10.1117/12.527345

Automatic content extraction of filled form images based on clustering component block projection vectors

Hanchuan Peng, Xiaofeng He, Fuhui Long

Research output: Contribution to journal › Conference article › peer-review

3 Scopus citations

Abstract

Automatic understanding of document images is a hard problem. Here we consider a sub-problem, automatically extracting content from filled form images. Without pre-selected templates or sophisticated structural/semantic analysis, we propose a novel approach based on clustering the component-block-projection-vectors. By combining spectral clustering and minimal spanning tree clustering, we generate highly accurate clusters, from which the adaptive templates are constructed to extract the filled-in content. Our experiments show this approach is effective for a set of 1040 US IRS tax form images belonging to 208 types.

Original language	English
Pages (from-to)	204-212
Number of pages	9
Journal	Proceedings of SPIE - The International Society for Optical Engineering
Volume	5296
DOIs	https://doi.org/10.1117/12.527345
State	Published - 2004
Externally published	Yes
Event	Document Recognition and Retrieval XI - San Jose, CA, United States Duration: 21 Jan 2004 → 22 Jan 2004

Keywords

Clustering
Document analysis
Form processing
Image classification
Image understanding

Access to Document

10.1117/12.527345

Cite this

@article{e6933f20bafb4dedb8af9ff9a1d6ade2,

title = "Automatic content extraction of filled form images based on clustering component block projection vectors",

abstract = "Automatic understanding of document images is a hard problem. Here we consider a sub-problem, automatically extracting content from filled form images. Without pre-selected templates or sophisticated structural/semantic analysis, we propose a novel approach based on clustering the component-block-projection-vectors. By combining spectral clustering and minimal spanning tree clustering, we generate highly accurate clusters, from which the adaptive templates are constructed to extract the filled-in content. Our experiments show this approach is effective for a set of 1040 US IRS tax form images belonging to 208 types.",

keywords = "Clustering, Document analysis, Form processing, Image classification, Image understanding",

author = "Hanchuan Peng and Xiaofeng He and Fuhui Long",

year = "2004",

doi = "10.1117/12.527345",

language = "英语",

volume = "5296",

pages = "204--212",

journal = "Proceedings of SPIE - The International Society for Optical Engineering",

issn = "0277-786X",

publisher = "SPIE",

note = "Document Recognition and Retrieval XI ; Conference date: 21-01-2004 Through 22-01-2004",

}

TY - JOUR

T1 - Automatic content extraction of filled form images based on clustering component block projection vectors

AU - Peng, Hanchuan

AU - He, Xiaofeng

AU - Long, Fuhui

PY - 2004

Y1 - 2004

N2 - Automatic understanding of document images is a hard problem. Here we consider a sub-problem, automatically extracting content from filled form images. Without pre-selected templates or sophisticated structural/semantic analysis, we propose a novel approach based on clustering the component-block-projection-vectors. By combining spectral clustering and minimal spanning tree clustering, we generate highly accurate clusters, from which the adaptive templates are constructed to extract the filled-in content. Our experiments show this approach is effective for a set of 1040 US IRS tax form images belonging to 208 types.

AB - Automatic understanding of document images is a hard problem. Here we consider a sub-problem, automatically extracting content from filled form images. Without pre-selected templates or sophisticated structural/semantic analysis, we propose a novel approach based on clustering the component-block-projection-vectors. By combining spectral clustering and minimal spanning tree clustering, we generate highly accurate clusters, from which the adaptive templates are constructed to extract the filled-in content. Our experiments show this approach is effective for a set of 1040 US IRS tax form images belonging to 208 types.

KW - Clustering

KW - Document analysis

KW - Form processing

KW - Image classification

KW - Image understanding

UR - https://www.scopus.com/pages/publications/8844259471

U2 - 10.1117/12.527345

DO - 10.1117/12.527345

M3 - 会议文章

AN - SCOPUS:8844259471

SN - 0277-786X

VL - 5296

SP - 204

EP - 212

JO - Proceedings of SPIE - The International Society for Optical Engineering

JF - Proceedings of SPIE - The International Society for Optical Engineering

T2 - Document Recognition and Retrieval XI

Y2 - 21 January 2004 through 22 January 2004

ER -

Automatic content extraction of filled form images based on clustering component block projection vectors

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this