DOCUMENT ANALYSIS - FROM PIXELS TO CONTENTS

被引:44
作者
SCHURMANN, J
BARTNECK, N
BAYER, T
FRANKE, J
MANDLER, E
OBERLANDER, M
机构
[1] DAIMLER BENZ INST INFORMAT TECHNOL, RES, AEG RES CTR, W-7900 ULM, GERMANY
[2] DAIMLER BENZ INST INFORMAT TECHNOL, AEG TELEFUNKEN RES CTR, DOCUMENT ANAL GRP, W-7900 ULM, GERMANY
[3] DAIMLER BENZ INST INFORMAT TECHNOL, CTR INFORMAT TECHNOL, AEG RES INST, DEPT PATTERN RECOGNIT, W-7900 ULM, GERMANY
关键词
PATTERN RECOGNITION; DOCUMENT ANALYSIS; CHARACTER RECOGNITION; LAYOUT ANALYSIS; CONTEXTUAL PROCESSING; DOCUMENT MODELING; DOCUMENT UNDERSTANDING;
D O I
10.1109/5.156473
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
The paper presents the conceptual framework for solving the task of document analysis, which, in essence, consists in the conversion of the document's pixel representation into an equivalent knowledge network representation holding the document's content and layout. The overall system is structured into several levels of abstraction. Starting on the pixel level, the formation of elementary geometric objects is described on which layout analysis as well as the definition of character objects is based. Character recognition accomplishes the mapping from geometric object to character meaning in ASCII representation. On the subsequent level of abstraction words are formed and verified by contextual processing. Modeled knowledge about complete documents and about how their constituents are related to the application form the highest level of abstraction. The various problems arising at each stage are discussed. The dependencies between the different levels are exemplified and technical solutions put forward.
引用
收藏
页码:1101 / 1119
页数:19
相关论文
共 21 条
[1]  
BAIRD HS, 1987, 40TH P SPSE C S HYBR, P21
[2]  
BARTNECK N, 1990, P ADV TECHNOLOGY C, P297
[3]  
BARTNECK N, 1987, THESIS TU BRAUNSCHWE
[4]  
BAYER T, 1990, PREPROC SSPR WORKSHO, P47
[5]  
BAYER T, 1986, MUSTERERKENNUNG, P56
[6]  
BAYER T, 1987, 5TH P SCAND C IM AN
[7]   VITERBI ALGORITHM [J].
FORNEY, GD .
PROCEEDINGS OF THE IEEE, 1973, 61 (03) :268-278
[8]  
HORAK W, 1985, IEEE COMPUT, P51
[9]  
HULL JJ, 1985, 2ND P IEEE C ART INT
[10]   A SYSTEM FOR INTERPRETATION OF LINE DRAWINGS [J].
KASTURI, R ;
BOW, ST ;
ELMASRI, W ;
SHAH, J ;
GATTIKER, JR ;
MOKATE, UB .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1990, 12 (10) :978-992