US20070201768A1 - Method And System For Acquiring Data From Machine-Readable Documents - Google Patents

Method And System For Acquiring Data From Machine-Readable Documents Download PDF

Info

Publication number
US20070201768A1
US20070201768A1 US10/573,429 US57342904A US2007201768A1 US 20070201768 A1 US20070201768 A1 US 20070201768A1 US 57342904 A US57342904 A US 57342904A US 2007201768 A1 US2007201768 A1 US 2007201768A1
Authority
US
United States
Prior art keywords
data
document
database
string section
extracted
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Abandoned
Application number
US10/573,429
Other languages
English (en)
Inventor
Matthias Schiehlen
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Open Text Inc
Original Assignee
Individual
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Individual filed Critical Individual
Assigned to OCE DOCUMENT TECHNOLOGIES GMBH reassignment OCE DOCUMENT TECHNOLOGIES GMBH ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: SCHIEHLEN, MATTHIAS
Publication of US20070201768A1 publication Critical patent/US20070201768A1/en
Assigned to CAPTARIS, INC. reassignment CAPTARIS, INC. ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: OCE DOCUMENT TECHNOLOGIES GMBH
Abandoned legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/40—Document-oriented image-based pattern recognition
    • G06V30/41—Analysis of document content
    • G06V30/416—Extracting the logical structure, e.g. chapters, sections or page numbers; Identifying elements of the document, e.g. authors
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/98—Detection or correction of errors, e.g. by rescanning the pattern or by human intervention; Evaluation of the quality of the acquired patterns
    • G06V10/987—Detection or correction of errors, e.g. by rescanning the pattern or by human intervention; Evaluation of the quality of the acquired patterns with the intervention of an operator

Definitions

  • the preferred embodiment invention relates to a method and a system for acquiring data from machine-readable documents, the data being assigned to a database, in which individual data are extracted from the document as automatically as possible and are entered into corresponding database fields, the method and system according to the present invention relating to the acquisition of data in the case in which data cannot be extracted with the necessary degree of reliability for one or more particular database fields of a document.
  • Methods and systems for acquiring data from machine-readable documents are known.
  • the systems have a scanner with which documents are optically scanned.
  • the data files produced in this way are machine-readable documents, and as a rule contain text elements.
  • the text elements are converted into coded text with the aid of an OCR device.
  • predetermined forms or templates are assigned to the data files, so that on the basis of the forms data files containing particular items of information from the text can be determined in a targeted manner. These items of information are stored for example in a database.
  • Methods and systems of this sort are used for example in large firms in order to read invoices.
  • the data extracted in this way can be communicated automatically to an accounting software program.
  • WO 98/47098 another system is known for the automatic acquisition of data from machine-readable documents.
  • a scanner is used to optically scan forms.
  • a line map of the form is created automatically.
  • all lines are acquired, and all graphic elements are converted into a line structure. Other elements, such as for example text sections, are filtered out.
  • All vertical lines form the basis for creating a vertical key
  • all horizontal lines form the basis for creating a horizontal key.
  • a string matching method is described in which two strings are compared and a cost measure is calculated that is indirectly proportional to the similarity of the strings. If the two strings are identical, the magnitude of the cost measure is zero. The more the strings differ, the greater is the magnitude of the cost measure. The cost measure is thus an expression of the similarity of the two strings.
  • This and similar methods are also known under the names approximate string matching, Levenshtein method, elastic matching, and Viterbi algorithm. These methods are part of the field of dynamic programming.
  • the read document is displayed on a display screen and the data can be read out only by marking corresponding fields in the read document.
  • additional master documents are automatically produced on the basis of the marked read documents, or existing master documents are correspondingly corrected. This system is easy enough to use that no special computer or software knowledge is necessary.
  • a method that supports an operator in the generation of electronic templates for a form recognition system arises from U.S. Pat. No. 5,317,646.
  • a form not provided with data (what is known as a master form) is shown on a screen, and the user can identify the data fields with a pointer device.
  • the coordinates that bound the corresponding region are automatically detected after which a single point within this region has been selected by the operator. Templates for the automatic form recognition can be created simply and quickly with this method.
  • a two-stage method in which form templates can be initially input and documents can be automatically read out using the input form templates arises from US 2002/141660 A1.
  • Form templates to be input are scanned, and the operator indicates input fields with a cursor. The position and size of the input fields is stored. The operator can also determine the data type associated with each data field. Given automatic reading of forms, these are scanned in and automatically read out using the data fields contained in the stored form documents. In the event that an error occurs in the readout, the operator can correct the errors via the keyboard.
  • U.S. Pat. No. 6,028,970 concerns a method and a system for automatic text recognition (OCR).
  • OCR automatic text recognition
  • the system comprises an error correction module (“error correction logic module”).
  • error correction module is applied to clearly detectable data errors in order to correct these. These corrections are executed automatically. Not only errors of individual letters are hereby detected, but rather errors in context are analyzed and correspondingly corrected. An error that cannot be automatically corrected can be communicated to the operator by means of an error message. The operator can then assess and, if applicable, correct the text generated by means of the text recognition.
  • a method for acquiring data from a machine-readable document for assignment to fields of a database individual data are extracted substantially automatically from the document and entered into the corresponding database fields. If data cannot be extracted from the document with a desired degree of reliability for one or more particular database fields, then the steps are executed of displaying the document onto the display screen, displaying on the display screen the at least one or more database fields for which the data cannot be extracted with the desired degree of reliability, and executing a proposal routine with which string sections in the vicinity of a pointer movable by a user on the display screen are selected, marked, and proposed for extraction.
  • FIG. 1 shows a method for acquiring from a document data that cannot be extracted automatically
  • FIGS. 2-6 each show copies of display screen representations corresponding to individual method steps of the method indicated in FIG. 1 ;
  • FIG. 7 shows a method for extracting data arranged in tables
  • FIGS. 8, 9 each show a table with marked data
  • FIG. 10 shows a system for executing the method according to the preferred embodiment.
  • data can be acquired from a plurality of machine-readable documents, the data being assigned to a database in that individual data are extracted from the document as automatically as possible and are entered into corresponding database fields. If data cannot be extracted with the necessary degree of reliability for one or more particular database fields of a document, for example because an error has been determined, caused for example by the fact that no data or false data are present in the document at the point at which the data are to be read, or that during the reading in of this document using an OCR method one or more characters are falsely converted, then according to the preferred embodiment the following steps are executed:
  • the document is displayed on the display screen so that the user can read it.
  • the database field is indicated for which the data cannot be extracted with the necessary degree of reliability. In this way, the user is informed of the database field for which the data must still be extracted from the document shown on the display screen.
  • string sections in the vicinity of a pointer, movable on the display screen by the user can be selected, marked, and proposed for extraction.
  • the user need merely move the pointer on the document shown on the display screen into the vicinity of a string section that contains the data for the indicated database field.
  • the data are then automatically selected, marked, and proposed for extraction.
  • the user can then transfer or incorporate the proposed string section into the database field merely by actuating a particular key.
  • the method according to the preferred embodiment for acquiring data from machine-readable documents is a development of the methods described above with which data can be extracted from documents and stored in a database by machine.
  • the method according to the preferred embodiment thus begins when data cannot be reliably extracted.
  • the expression “not reliably extractable” includes both fundamental errors in the reading of data that make a reading of the data impossible, and also read data that are mapped to the database field while taking into account context information, the quality of the mapping being determined during this process.
  • Such mapping methods include for example the string matching method named above. If the mapping quality achieved here is too low, the automatically read-in data are evaluated as insufficiently reliable and are rejected.
  • the method begins with step S 1 .
  • FIG. 2 shows a display screen representation immediately after the determination that data could not be extracted with the necessary degree of reliability; here the document 1 is shown in a window 4 / 1 on the right side of the display screen representation. Two windows 4 / 2 and 4 / 3 are situated on the left side. Window 4 / 2 contains an overview of the documents that are to be processed, and in window 4 / 3 the individual database fields are indicated in which data are stored that are to be read from document 1 .
  • none of the database fields could be filled with data, for which reason the individual database fields 3 are provided with the designation “empty”. However, it is also possible for data to be missing only in a few database fields, or only in a single database field.
  • the database field “InvoiceNumber” is marked darker in comparison to the other database fields 3 , which indicates to the user that data are to be extracted from document 1 for this database field 3 .
  • the term “InvoiceNumber” is indicated in a larger font, additionally indicating to the user the database field for which data are to be extracted.
  • the user can now position a pointer 5 that he preferably situates in such a way that it is located as close as possible to the string section for which the user assumes that the content is to be stored in the corresponding database field.
  • a pointer 5 that he preferably situates in such a way that it is located as close as possible to the string section for which the user assumes that the content is to be stored in the corresponding database field.
  • data are to be extracted for the database field “InvoiceNumber,” so pointer 5 is positioned in the vicinity of invoice number “ 4361 ” (step S 3 ).
  • pointer 5 can be moved in window 4 / 1 using a mouse 6 or via inputs on a keyboard 7 .
  • a proposal routine begins that comprises a plurality of method steps.
  • This proposal routine can on the one hand be initiated in that pointer 5 is not moved for a predetermined time interval, whereupon the proposal routine is then automatically executed, or it can be initiated by actuating a particular mouse button or keyboard key.
  • step S 4 it is first checked whether there is located in the immediate vicinity of the pointer a string section having a concept suitable for database field 3 , insofar as concept information has been previously assigned to the corresponding type of the database field.
  • This concept information includes the syntax and/or the semantics of the database field.
  • Information concerning syntax includes for example the number of numerals and/or letters and/or specified formats of the string section that is to be read.
  • date fields, amount fields, and address fields have as a rule particular formats.
  • Semantic information includes specified terms that can be entered into the corresponding database field. This is useful for example for currency indications, or if the article designation of a particular supplier that can supply a limited number of articles is to be read in. The corresponding article designations are then stored in a lexicon and can then be unambiguously recognized.
  • the two string sections “ 4361 ” and “ 02 . 08 . 2002 ” are situated in the vicinity of pointer 5 .
  • the latter string section has the syntax of a date, and for this reason it is rejected for the extraction of the invoice number.
  • the string section “ 4361 ” corresponds to the syntax of an invoice number. Therefore, in step S 4 it is decided that a string section having a suitable concept is present, and for this reason the method sequence next goes to step S 5 .
  • the string section “ 4361 ” is marked ( FIG. 3 ). In the present exemplary embodiment, the marking takes place through a colored highlighting or background of the string section and through the drawing in of a frame 8 .
  • step S 6 the individual character situated closest to pointer 5 is determined, which, in the present exemplary embodiment according to FIGS. 2-4 , is the “1.”
  • the boundaries of the string section containing this character are determined according to general rules. These boundaries can for example be determined by empty characters or empty spaces in the document 1 , or by particular punctuation marks or other markings in document 1 . If corresponding boundary markings are recognized, the string section situated between them is selected and marked. In the exemplary embodiment shown in FIGS. 2 and 3 , on each side of string section “ 4361 ” there are situated empty spaces, via which an unambiguous selection of the marking of the string section is possible, according to the general rules as well.
  • step S 7 Independent of whether the string section has been selected or marked according to step S 5 or according to step S 6 , the method sequence goes to method step S 7 , with which the string section is displayed in an additional frame 9 as a coded text, and is displayed in an enlarged fashion in another frame 10 ( FIGS. 3, 4 ).
  • document 1 is present as a graphic data file, e.g. in the .pdf, tif, .gif, or .jpg format.
  • the coded text is here also examined for concepts, and the corresponding information is stored.
  • the section corresponding to the string section is removed from this coded text and is shown in frame 9 . In this way, the user recognizes whether the string section has been correctly converted into coded text.
  • the string section is shown in a graphic format in an enlarged representation, so that the user can also recognize details in the string section.
  • step S 7 the proposal routine is terminated.
  • step S 8 the user judges whether the selected and marked string section is fundamentally suitable for transferring into the database field. If this is not the case, pointer 5 is repositioned (S 3 ) and the proposal routine (S 4 -S 7 ) is executed again. If, in contrast, the selection of the string section is fundamentally suitable, the user judges whether the marked area is also correct (step S 9 ). If this is not the case, the user can manually process the marking of the string section and/or can edit the coded text in frame 9 (step S 10 ). With the editing of the coded text, errors resulting from an incorrect OCR conversion can be removed. When these corrections (adapt area, edit) are made, the marked area and the contents of frames 9 and 10 are automatically adapted.
  • step S 11 the data contained in the selected string section are transferred into the corresponding database field ( FIG. 4 ). This transferring of the data is initiated by user actuation of a predetermined mouse button or key on the keyboard. Subsequently, the method for extracting data for a database field is terminated (S 12 ). If data are to be read for additional database fields, the method begins again with step S 1 . In FIG. 5 , the next database field to be read (“Invoice Date”) is indicated.
  • the activity of a user in the manual transferring of data from a document into a database field is limited to the positioning of the pointer, the checking of the automatically proposed selection and the possible correction of the area, and the actuation of a key in order to transfer the data.
  • the selection and the marking of the area of the string section to be selected are carried out automatically by the method according to the preferred embodiment.
  • FIGS. 2 to 5 show the transfer of data into an individual database field. However, by taking into account concept information, it is also possible to extract data for a plurality of database fields with a single string section.
  • FIG. 6 shows a corresponding exemplary embodiment, in which the complete address is marked and read as a string section, the address being automatically segmented into the individual database fields name, company, street, postal code, and city.
  • This method begins with step S 15 .
  • step S 16 the values of the table in the first table row are extracted according to the above method through the positioning of the pointer, the automatic selection and marking of the string section, and the transferring of the data into corresponding database fields.
  • FIG. 8 shows a table in which the string sections of the first table row are marked that have been transferred into the corresponding database fields.
  • These database fields have the structure of a table; for example, they are applied as a two-dimensional data field, so that during the extraction of the data into these database fields the method recognizes automatically, on the basis of the database field, that data are being read out from a table.
  • a row of a table can also extend over a plurality of pages if the table correspondingly extends over a plurality of pages. If the data of the first table row has been completely extracted, the user can initiate the automatic extraction of the further table entries using a predetermined input. If this input is actuated by the user, then, in step S 17 , first a list is created of all string sections that are situated under the first table row.
  • a cost function is used to determine a cost value between sequences of string sections of the list and the sequence of the string sections of the first table row, on the basis of which data were extracted into the database fields in step S 16 .
  • this cost function low costs are assigned to the sequences of the string sections of the list whose string sections agree with, or are at least very similar to, the corresponding string sections of the first table row, with respect to their horizontal position and their width. This cost value is thus indirectly proportional to the degree of similarity between the sequences of string sections appearing in the list and the sequence of string sections contained in the first table row.
  • the cost function used here corresponds to the cost function described in Chapter 8.6.1 of String Matching Allowing Errors in Modern Information Retrieval (ISBN 0-201-39829-X), with which an individual cost value between a string section of the first table row and a string section of the further table rows is determined. Because each sequence comprises a plurality of string sections, the Viterbi algorithm is used to calculate an overall cost value or overall similarity value for each of the individual sequences of string sections, through summation of the individual cost values.
  • the sequences of string sections are determined as table rows whose similarity value lies beneath a predetermined threshold value (S 19 ). In this way, all table rows, and thus table entries, of the table are determined. They are marked in step S 20 ( FIG. 9 ) and in step S 21 they are extracted, i.e., automatically read out, converted into coded text if necessary, and stored in the corresponding database fields.
  • step S 22 this method is terminated.
  • the table entries i.e., to modify (move, enlarge, make smaller) the marked areas, or to remove or add individual rows.
  • the entries in the database fields are automatically updated correspondingly.
  • master documents are compared with a read document and their similarity is evaluated.
  • the method applied here can also be used for reading out from a table, the sequence of the selected string sections corresponding to the first table row of the master document, and the combinations of string sections corresponding to the further table rows of the read documents.
  • a user need merely move the pointer to the table entries in the first table row and confirm the transferring of the then automatically selected and marked string sections as data for the corresponding database field. After the user has done this for all table entries of the first table row, he need merely initiate the complete reading out of the further table entries by making an input. The method then automatically determines the further table entries, marks them, and extracts the data into the database.
  • Method segment S 17 to S 21 therefore represents an independent preferred embodiment in its own right, which is however preferably applied in combination with the method represented in FIG. 1 , to which step S 16 relates.
  • FIG. 10 schematically shows a system for executing the method according to the preferred embodiment.
  • This system 11 comprises a computer 12 having a storage device 13 , having a central processor device (CPU) 14 , and having an interface device 15 .
  • a scanner 16 , a display screen 2 , and an input device 17 are connected to computer 12 .
  • Input device 17 includes a keyboard 7 and/or a mouse 6 .
  • a software product is stored for executing the method according to the preferred embodiment, this software product being executed at CPU 14 .
  • Scanner 16 is used to acquire documents and to convert them into an electronic data file. These electronic data files are read by computer 12 and are preprocessed if necessary, using an OCR routine and/or a method for recognizing particular syntax or semantics in the data file. Subsequently, the documents contained in the data files are processed in a manner corresponding to the method described above, using system 11 .
  • the corresponding inputs can be carried out, these being limited to movements of pointer 5 and a few keyboard inputs. If necessary, the marked fields can be moved using the keyboard or the mouse, or can be adapted by enlargement or by being made smaller, or the coded text can be edited.
  • the documents are scanned in and are then present in a graphic format.
  • the method according to the preferred embodiment can also be used for reading information from documents that are already present in coded text, such as for example e-mails.
  • coded text such as for example e-mails.
  • OCR routine it is not necessary for the documents to be converted into coded text using an OCR routine.
  • the preferred embodiment relates to a method for acquiring data from machine-readable documents, the data being assigned to a database.
  • string sections located in the vicinity of a pointer that can be moved by the user are automatically selected and marked, and their content is proposed for transfer into a database.
  • the content of a table can be read out in a fully automatic manner if the table entries in a first table row have already been read out according to the above method.
  • the preferred embodiment can be realized both by means of electronic components (hardware) and through computer program elements (software or software modules).
  • the preferred embodiment is realized here as a combination of electronic hardware elements and software elements.
  • the preferred embodiment also includes computer program products, such as for example electronic data carriers (CDs, DVDs, diskettes, tape drives), or components that are distributed via computer networks (Internet) and/or on computers, and in particular are loaded into intermediate storage units and are kept ready there and/or are run from there.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Quality & Reliability (AREA)
  • Artificial Intelligence (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Character Discrimination (AREA)
  • Document Processing Apparatus (AREA)
  • Devices For Checking Fares Or Tickets At Control Points (AREA)
US10/573,429 2003-09-30 2004-08-26 Method And System For Acquiring Data From Machine-Readable Documents Abandoned US20070201768A1 (en)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
DE10345526.4 2003-09-30
DE10345526A DE10345526A1 (de) 2003-09-30 2003-09-30 Verfahren und System zum Erfassen von Daten aus maschinell lesbaren Dokumenten
PCT/EP2004/009539 WO2005043452A1 (fr) 2003-09-30 2004-08-26 Procede et systeme permettant de detecter des donnees dans des documents lisibles par machine

Related Parent Applications (1)

Application Number Title Priority Date Filing Date
PCT/EP2004/009539 A-371-Of-International WO2005043452A1 (fr) 2003-09-30 2004-08-26 Procede et systeme permettant de detecter des donnees dans des documents lisibles par machine

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US12/495,755 Continuation US8270721B2 (en) 2003-09-30 2009-06-30 Method and system for acquiring data from machine-readable documents

Publications (1)

Publication Number Publication Date
US20070201768A1 true US20070201768A1 (en) 2007-08-30

Family

ID=34484693

Family Applications (2)

Application Number Title Priority Date Filing Date
US10/573,429 Abandoned US20070201768A1 (en) 2003-09-30 2004-08-26 Method And System For Acquiring Data From Machine-Readable Documents
US12/495,755 Active 2024-12-11 US8270721B2 (en) 2003-09-30 2009-06-30 Method and system for acquiring data from machine-readable documents

Family Applications After (1)

Application Number Title Priority Date Filing Date
US12/495,755 Active 2024-12-11 US8270721B2 (en) 2003-09-30 2009-06-30 Method and system for acquiring data from machine-readable documents

Country Status (5)

Country Link
US (2) US20070201768A1 (fr)
EP (1) EP1671262B1 (fr)
AT (1) ATE409923T1 (fr)
DE (2) DE10345526A1 (fr)
WO (1) WO2005043452A1 (fr)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060136450A1 (en) * 2004-12-22 2006-06-22 John Barrus Document separator pages
US20070065011A1 (en) * 2003-09-15 2007-03-22 Matthias Schiehlen Method and system for collecting data from a plurality of machine readable documents
US20070168382A1 (en) * 2006-01-03 2007-07-19 Michael Tillberg Document analysis system for integration of paper records into a searchable electronic database
US20100272362A1 (en) * 2009-04-24 2010-10-28 Kazuyuki Ohnishi Image forming apparatus for extracting hand-written image

Families Citing this family (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE102005032046A1 (de) 2005-07-08 2007-01-11 Océ Document Technologies GmbH Verfahren, System und Computerprogramm-Produkt zum Übertragen von Daten aus einer Dokumentenanwendung in eine Datenanwendung
EP1946233A4 (fr) * 2005-10-25 2013-02-27 Charactell Ltd Extraction de donnees de formulaire sans personnalisation
EP1956517A1 (fr) * 2007-02-07 2008-08-13 WinBooks s.a. Procédé informatique pour traiter des opérations comptables et produit logiciel pour mettre en oeuvre un tel procédé
US20110090253A1 (en) * 2009-10-19 2011-04-21 Quest Visual, Inc. Augmented reality language translation system and method
US9613267B2 (en) 2012-05-31 2017-04-04 Xerox Corporation Method and system of extracting label:value data from a document
US9798711B2 (en) 2012-05-31 2017-10-24 Xerox Corporation Method and system for generating a graphical organization of a page
EP2883193A4 (fr) * 2012-08-10 2016-07-13 Transaxy Inc Système pour entrer des données dans un système de traitement de données
HRP20130700B1 (hr) * 2013-07-23 2016-03-11 Microblink D.O.O. Sustav za adaptivnu detekciju i ekstrakciju struktura iz strojno generiranih dokumenata
US9984287B2 (en) * 2015-03-05 2018-05-29 Wipro Limited Method and image processing apparatus for performing optical character recognition (OCR) of an article
US9766969B2 (en) 2015-06-18 2017-09-19 Xerox Corporation Assessing and improving quality of event logs including prioritizing and classifying errors into error-perspective and error-type classifications
JP6739937B2 (ja) * 2015-12-28 2020-08-12 キヤノン株式会社 情報処理装置、情報処理装置の制御方法、及びプログラム
US9965809B2 (en) 2016-07-25 2018-05-08 Xerox Corporation Method and system for extracting mathematical structures in tables
US10217086B2 (en) 2016-12-13 2019-02-26 Golbal Healthcare Exchange, Llc Highly scalable event brokering and audit traceability system
US10217158B2 (en) 2016-12-13 2019-02-26 Global Healthcare Exchange, Llc Multi-factor routing system for exchanging business transactions
US10241992B1 (en) * 2018-04-27 2019-03-26 Open Text Sa Ulc Table item information extraction with continuous machine learning through local and global models
CN110889310B (zh) * 2018-09-07 2023-05-09 深圳市赢时胜信息技术股份有限公司 金融文档信息智能提取系统及方法
US11543943B2 (en) 2019-04-30 2023-01-03 Open Text Sa Ulc Systems and methods for on-image navigation and direct image-to-data storage table data capture

Citations (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3611291A (en) * 1969-10-30 1971-10-05 Scan Data Corp Character recognition system for reading a document edited with handwritten symbols
US3925760A (en) * 1971-05-06 1975-12-09 Ecrm Method of and apparatus for optical character recognition, reading and reproduction
US4272756A (en) * 1978-05-22 1981-06-09 Hitachi, Ltd. Method of pattern recognition and apparatus thereof
US4933979A (en) * 1986-12-19 1990-06-12 Ricoh Company, Ltd. Data reading apparatus for reading data from form sheet
US5317646A (en) * 1992-03-24 1994-05-31 Xerox Corporation Automated method for creating templates in a forms recognition and processing system
US5448375A (en) * 1992-03-20 1995-09-05 Xerox Corporation Method and system for labeling a document for storage, manipulation, and retrieval
US5594809A (en) * 1995-04-28 1997-01-14 Xerox Corporation Automatic training of character templates using a text line image, a text line transcription and a line image source model
US5666549A (en) * 1992-03-10 1997-09-09 Hitachi, Ltd. Method and system for processing a document transmitted via facsimile in an initially input form stored in a knowledge base
US5689620A (en) * 1995-04-28 1997-11-18 Xerox Corporation Automatic training of character templates using a transcription and a two-dimensional image source model
US5835712A (en) * 1996-05-03 1998-11-10 Webmate Technologies, Inc. Client-server system using embedded hypertext tags for application and database development
US5923792A (en) * 1996-02-07 1999-07-13 Industrial Technology Research Institute Screen display methods for computer-aided data entry
US5963966A (en) * 1995-11-08 1999-10-05 Cybernet Systems Corporation Automated capture of technical documents for electronic review and distribution
US5966473A (en) * 1996-11-28 1999-10-12 International Business Machines Corporation Method and apparatus for recognizing a quadrilateral object contained in an input bitmap image
US6028970A (en) * 1997-10-14 2000-02-22 At&T Corp Method and apparatus for enhancing optical character recognition
US6131102A (en) * 1998-06-15 2000-10-10 Microsoft Corporation Method and system for cost computation of spelling suggestions and automatic replacement
US6353840B2 (en) * 1997-08-15 2002-03-05 Ricoh Company, Ltd. User-defined search template for extracting information from documents
US20020141660A1 (en) * 2001-03-12 2002-10-03 Multiscan Corp. Document scanner, system and method
US20030115189A1 (en) * 2001-12-19 2003-06-19 Narayan Srinivasa Method and apparatus for electronically extracting application specific multidimensional information from documents selected from a set of documents electronically extracted from a library of electronically searchable documents
US20040243552A1 (en) * 2003-05-30 2004-12-02 Dictaphone Corporation Method, system, and apparatus for viewing data
US20070065011A1 (en) * 2003-09-15 2007-03-22 Matthias Schiehlen Method and system for collecting data from a plurality of machine readable documents

Family Cites Families (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5339421A (en) * 1991-03-22 1994-08-16 International Business Machines Corporation General data stream parser for encoding and decoding data and program interface for same
DE69326364T2 (de) * 1992-02-04 2000-02-24 Ricoh Kk Informationsverarbeitungsapparat und -verfahren unter Verwendung eines zusätzlichen zweckmässigen Informationspakets
JPH06274680A (ja) * 1993-03-17 1994-09-30 Hitachi Ltd 文書認識方法およびシステム
US5659791A (en) * 1993-04-26 1997-08-19 Microsoft Corporation Encapsulation of extracted portions of documents into objects
US6108674A (en) * 1994-06-28 2000-08-22 Casio Computer Co., Ltd. Image output devices which automatically selects and outputs a stored selected image in correspondence with input data
JP2729356B2 (ja) * 1994-09-01 1998-03-18 日本アイ・ビー・エム株式会社 情報検索システム及び方法
US6457004B1 (en) * 1997-07-03 2002-09-24 Hitachi, Ltd. Document retrieval assisting method, system and service using closely displayed areas for titles and topics
SE511242C2 (sv) * 1997-04-01 1999-08-30 Readsoft Ab Förfarande och anordning för automatisk datafångst hos formulär
JP3756719B2 (ja) * 2000-01-20 2006-03-15 理想科学工業株式会社 文書修飾装置及び画像処理装置
US6928449B2 (en) * 2001-10-18 2005-08-09 Sun Microsystems, Inc. Mechanism for facilitating backtracking
US7221796B2 (en) * 2002-03-08 2007-05-22 Nec Corporation Character input device, character input method and character input program
US6941521B2 (en) * 2002-03-29 2005-09-06 Intel Corporation Method for dynamically generating a user interface from XML-based documents
US7707039B2 (en) * 2004-02-15 2010-04-27 Exbiblio B.V. Automatic modification of web pages

Patent Citations (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3611291A (en) * 1969-10-30 1971-10-05 Scan Data Corp Character recognition system for reading a document edited with handwritten symbols
US3925760A (en) * 1971-05-06 1975-12-09 Ecrm Method of and apparatus for optical character recognition, reading and reproduction
US4272756A (en) * 1978-05-22 1981-06-09 Hitachi, Ltd. Method of pattern recognition and apparatus thereof
US4933979A (en) * 1986-12-19 1990-06-12 Ricoh Company, Ltd. Data reading apparatus for reading data from form sheet
US5666549A (en) * 1992-03-10 1997-09-09 Hitachi, Ltd. Method and system for processing a document transmitted via facsimile in an initially input form stored in a knowledge base
US5448375A (en) * 1992-03-20 1995-09-05 Xerox Corporation Method and system for labeling a document for storage, manipulation, and retrieval
US5680223A (en) * 1992-03-20 1997-10-21 Xerox Corporation Method and system for labeling a document for storage, manipulation, and retrieval
US5317646A (en) * 1992-03-24 1994-05-31 Xerox Corporation Automated method for creating templates in a forms recognition and processing system
US5594809A (en) * 1995-04-28 1997-01-14 Xerox Corporation Automatic training of character templates using a text line image, a text line transcription and a line image source model
US5689620A (en) * 1995-04-28 1997-11-18 Xerox Corporation Automatic training of character templates using a transcription and a two-dimensional image source model
US5963966A (en) * 1995-11-08 1999-10-05 Cybernet Systems Corporation Automated capture of technical documents for electronic review and distribution
US5923792A (en) * 1996-02-07 1999-07-13 Industrial Technology Research Institute Screen display methods for computer-aided data entry
US5835712A (en) * 1996-05-03 1998-11-10 Webmate Technologies, Inc. Client-server system using embedded hypertext tags for application and database development
US5966473A (en) * 1996-11-28 1999-10-12 International Business Machines Corporation Method and apparatus for recognizing a quadrilateral object contained in an input bitmap image
US6353840B2 (en) * 1997-08-15 2002-03-05 Ricoh Company, Ltd. User-defined search template for extracting information from documents
US6028970A (en) * 1997-10-14 2000-02-22 At&T Corp Method and apparatus for enhancing optical character recognition
US6131102A (en) * 1998-06-15 2000-10-10 Microsoft Corporation Method and system for cost computation of spelling suggestions and automatic replacement
US20020141660A1 (en) * 2001-03-12 2002-10-03 Multiscan Corp. Document scanner, system and method
US20030115189A1 (en) * 2001-12-19 2003-06-19 Narayan Srinivasa Method and apparatus for electronically extracting application specific multidimensional information from documents selected from a set of documents electronically extracted from a library of electronically searchable documents
US20040243552A1 (en) * 2003-05-30 2004-12-02 Dictaphone Corporation Method, system, and apparatus for viewing data
US20070065011A1 (en) * 2003-09-15 2007-03-22 Matthias Schiehlen Method and system for collecting data from a plurality of machine readable documents

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070065011A1 (en) * 2003-09-15 2007-03-22 Matthias Schiehlen Method and system for collecting data from a plurality of machine readable documents
US7668372B2 (en) 2003-09-15 2010-02-23 Open Text Corporation Method and system for collecting data from a plurality of machine readable documents
US20060136450A1 (en) * 2004-12-22 2006-06-22 John Barrus Document separator pages
US7580164B2 (en) * 2004-12-22 2009-08-25 Ricoh Co., Ltd. Document separator pages
US20070168382A1 (en) * 2006-01-03 2007-07-19 Michael Tillberg Document analysis system for integration of paper records into a searchable electronic database
US20100272362A1 (en) * 2009-04-24 2010-10-28 Kazuyuki Ohnishi Image forming apparatus for extracting hand-written image
US8483483B2 (en) * 2009-04-24 2013-07-09 Sharp Kabushiki Kaisha Image forming apparatus for extracting hand-written image

Also Published As

Publication number Publication date
DE502004008170D1 (de) 2008-11-13
ATE409923T1 (de) 2008-10-15
WO2005043452A1 (fr) 2005-05-12
DE10345526A1 (de) 2005-05-25
US8270721B2 (en) 2012-09-18
EP1671262A1 (fr) 2006-06-21
US20100094888A1 (en) 2010-04-15
EP1671262B1 (fr) 2008-10-01

Similar Documents

Publication Publication Date Title
US8270721B2 (en) Method and system for acquiring data from machine-readable documents
US7668372B2 (en) Method and system for collecting data from a plurality of machine readable documents
US6782144B2 (en) Document scanner, system and method
US6567546B1 (en) Data medium handling apparatus medium handling method
JP3427692B2 (ja) 文字認識方法および文字認識装置
KR100570224B1 (ko) 전표정의데이터 작성방법 및 전표처리장치
JPH0668298A (ja) 文書書式の文字認識及びデータ修復処理のためのデータ処理システム及び方法
US20060045340A1 (en) Character recognition apparatus and character recognition method
WO2007049270A2 (fr) Extraction de donnees de formulaire sans personnalisation
US20120039536A1 (en) Optical character recognition with two-pass zoning
JPH03161891A (ja) 表形式文書読取装置
JP4649512B2 (ja) 文字列検索方法およびその装置
JPH11282955A (ja) 文字認識装置、文字認識方法およびその方法をコンピュータに実行させるプログラムを記録したコンピュータ読み取り可能な記録媒体
CN110688863A (zh) 一种文档翻译系统及文档翻译方法
EP1202213B1 (fr) Appareil et procédé d'identifcation de la mise en page des documents
Hamzah et al. Data capturing: Methods, issues and concern
EP3477547B1 (fr) Systèmes de reconnaissance optique de caractères et procédés
JP3394694B2 (ja) フォーマット情報登録方法及びｏｃｒシステム
WO2019194052A1 (fr) Dispositif de traitement d'image, procédé de traitement d'image, et support de stockage stockant un programme
Tomaschek Evaluation of off-the-shelf OCR technologies
JP7111143B2 (ja) 画像処理装置、画像処理方法およびプログラム
JPH103483A (ja) 情報検索装置
JPH11219409A (ja) 文書読取装置
JP3435374B2 (ja) 文字読取装置および文字認識方法
JPH0696288A (ja) 文字認識装置及び機械翻訳装置

Legal Events

Date Code Title Description
AS Assignment

Owner name: OCE DOCUMENT TECHNOLOGIES GMBH, GERMANY

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:SCHIEHLEN, MATTHIAS;REEL/FRAME:018721/0169

Effective date: 20060412

AS Assignment

Owner name: CAPTARIS, INC., WASHINGTON

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:OCE DOCUMENT TECHNOLOGIES GMBH;REEL/FRAME:021570/0520

Effective date: 20080902

Owner name: CAPTARIS, INC.,WASHINGTON

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:OCE DOCUMENT TECHNOLOGIES GMBH;REEL/FRAME:021570/0520

Effective date: 20080902

STCB Information on status: application discontinuation

Free format text: ABANDONED -- FAILURE TO RESPOND TO AN OFFICE ACTION