JP2005173686A - Document detection method and system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To perform a detection in defiance of the reading order of document blocks while improving the speed of detection by performing a concept vector retrieval of a document for retrieval of an original computerized document from a scanned paper document, and detecting a similar document. <P>SOLUTION: This system comprises a document storage means for storing the original computerized document; a document reading means for reading a printed document from the original computerized document as image data; a document recognition means for recognizing the read image data as a character code; a document analysis means for analyzing the recognized document data to extract layout information of the document; a document detection means for performing a concept vector retrieval of document for detection of the stored original computerized document from the analyzed document data to detect a similar document, and a detection result output means for outputting a detected result. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a document detection method and system for detecting an original digitized document stored in a hard disk of a PC from a scanned paper document, and particularly to document detection using the similarity of concept vector search for a document. It is.

  When an original digitized document is detected from a scanned paper document, the printed document is read as image data, the read image data is recognized as a character code, the recognized document data is analyzed, and the analyzed document is analyzed. A document detection method and system capable of detecting a document at high speed by performing a concept vector search on the document for detecting an original digitized document stored from data and detecting a similar document.

  In the conventional document detection method and system, as in Patent Document 1, when an original digitized document accumulated from a scanned paper document is detected, the scanned paper document is subjected to OCR (Optical Character Recognition, optical character recognition). Text information is extracted by recognition), the text information is analyzed, the analyzed document data is matched with the character string of the stored original electronic document, and a matching document is detected.

Also, text information is extracted from the scanned paper document by OCR. However, since OCR cannot be guaranteed 100%, there is a recognition error, and the analyzed document data and the stored original digitized character string are ambiguous. Matching, that is, matching between character strings was determined in consideration of ambiguity.
JP-A-8-263512

  However, in the case of the prior art as described above, when detecting an original digitized document stored from a scanned paper document, the scanned paper document is not guaranteed even though the OCR recognition rate cannot be guaranteed 100%. The text information is extracted by OCR, the text information is analyzed, and the character string of the analyzed document data is not detected unless the character string of the stored original digitized document is an exact match. .

  In the case of ambiguous matching of character strings, for example, when character string data and image data are mixed in a document, character string blocks and image blocks are extracted separately, and a plurality of character string blocks are extracted. The character string may be divided into a plurality of blocks. In this case, the document cannot be detected even in the ambiguous matching.

  In addition, since it is necessary to perform matching of character strings for all character strings in the document, it takes a long time for detection processing.

  The present invention has been made to solve the above-described problems of the prior art, and an object of the present invention is to detect an original digitized document accumulated from a scanned paper document at a high speed. . Further, it is possible to perform a search ignoring the reading order which is the order of the document block.

  In order to achieve the above object, a document detection method and system according to the present invention includes a document storage means for storing an original electronically created document in the document detection method and system for detecting a document, and the original A document reading unit that reads a document printed from an electronic document as image data, a document recognition unit that recognizes image data read by the document reading unit as a character code, and a document data recognized by the document recognition unit A document detection unit that performs a concept search for detection of the original electronic document stored in the document storage unit and detects a corresponding document, and a detection result output unit that outputs a result detected by the document detection unit And.

  Further, in a document detection method and system for detecting a document, document storage means for storing an electronically generated original electronic document, and document reading for reading a document printed from the original electronic document as image data Means, document recognition means for recognizing image data read by the document reading means as a character code, document analysis means for analyzing document data recognized by the document recognition means, and extracting document layout information, A concept search is performed on the detection of the original digitized document stored in the document storage unit from the document data analyzed by the document analysis unit, and the document detection unit detects the corresponding document, and the document detection unit detects the document. And a detection result output means for outputting the result.

  In addition, in a document detection method and system for detecting a document, document storage means for storing an electronically created electronic document, document reading means for reading a printed or handwritten document as image data, and the document reading means A document recognition means for recognizing the image data read as a character code, and a concept search for detecting a document similar to an electronic document stored in the document storage means from the document data recognized by the document recognition means. And a document detection means for detecting a similar document, and a detection result output means for outputting a result detected by the document detection means.

  In addition, in a document detection method and system for detecting a document, document storage means for storing an electronically created electronic document, document reading means for reading a printed or handwritten document as image data, and the document reading means The document recognition means for recognizing the image data read as a character code, the document analysis means for analyzing the document data recognized by the document recognition means and extracting the layout information of the document, and the document analysis means A concept search is performed for detection of similar documents in the digitized document stored in the document storage means from the document data, and a document detection means for detecting similar documents and a result detected by the document detection means are output. And a detection result output means.

  In the document detection means, the concept search is a concept vector search for a document.

  In the document detection means, the object of the document to be detected is a document unit.

  In the document detection means, the target of the document to be detected is a document layout unit extracted by the document analysis means.

  As described above, according to the present invention, when an original digitized document is detected from a scanned paper document, the printed document is read as image data, and the read image data is recognized as a character code. Analyzing the analyzed document data, performing a concept vector search on the document for the detection of the original digitized document stored from the analyzed document data, and detecting a similar document, fast document detection It is possible to perform a search that ignores the reading order that is the order of the document block.

(First embodiment)
Hereinafter, a first embodiment of the present invention will be described with reference to FIGS.

  FIG. 1 is a block diagram showing the configuration of a character processing apparatus to which the present invention is applied.

  In the configuration shown in the figure, the CPU is a microprocessor that performs calculation, logic determination, and the like for document detection processing, and controls each component connected to the bus via the bus. The microprocessor CPU also operates as document detection display means.

  BUS is a bus and transfers an address signal and a control signal instructing each component to be controlled by the microprocessor CPU. In addition, data transfer between each component is performed.

  The RAM is a writable random access memory and is used for primary storage of various data from each component.

  ROM is a read-only fixed memory. The boot program by the microprocessor CPU is stored. The boot program loads a control program stored in the hard disk into the RAM when the system is started, and causes the microprocessor CPU to execute it. The control program will be described in detail later with reference to a flowchart.

  The input device is a keyboard and a mouse.

  The display device is a CRT or a liquid crystal display.

  The scanner is a scanner for reading and digitizing a paper document.

  HD is a hard disk, a control program executed by the CPU, a morphological analysis dictionary for performing morphological analysis, a document database storing electronically created electronic documents, and a word vector obtained by conceptualizing each word of the document A dictionary and a concept search index used as an index when performing concept search are stored.

  The removable external storage device is a drive or the like for accessing an external storage such as a floppy disk, CD, or DVD. It is an apparatus that can be used in the same manner as the HD and exchanges data with other document processing apparatuses through these recording media. Note that the control program stored in the hard disk can be copied from these external storage devices to HD as necessary.

  The communication device is a network controller. A device for exchanging data with the outside via a communication line.

  In the document detection processing apparatus of the present invention composed of such components, it operates in response to various inputs from the input device. When an input from the input device is supplied, an interrupt signal is first sent to the microprocessor CPU. Along with this, the CPU reads various instructions stored in the ROM or RAM, and various controls are performed by executing the instructions.

  FIG. 2 is a diagram showing a processing flow of the first embodiment of the present invention.

  The first embodiment of the present invention is an electronic document 1, a document storage unit 2, a document printing unit 3, a paper document 4, a document reading unit 5, a document recognition unit 6, a document analysis unit 7, a document detection unit 8, and a detection result. The output unit 9 is configured.

  Reference numeral 1 denotes an electronic document, which is an original electronic document created electronically.

  Reference numeral 2 denotes a document storage unit that stores an electronically generated original electronic document of the electronic document 1.

  Reference numeral 3 denotes a document printing unit which prints an electronically generated original electronic document of the electronic document 1.

  Reference numeral 4 denotes a paper document, which is a paper document output by printing an electronically created original electronic document of the electronic document 1 by the electronic printing unit 3.

  A document reading unit 5 reads a printed document from the original electronically created document of the paper document 4 as image data.

  A document recognition unit 6 recognizes image data of a document printed from an electronically created original electronic document read by the document reading unit 5 as a character code.

  A document analysis unit 7 analyzes the document data recognized by the document recognition unit 6 and extracts document layout information.

  A document detection unit 8 performs a concept search on the detection of the original digitized document stored in the document storage unit 2 from the document data analyzed by the document analysis unit 7, and detects the corresponding document.

  A detection result output unit 9 outputs the result of the document detected by the document detection unit 8.

  3 to 5 are flow charts showing the flow of processing of the first embodiment of the present invention.

  FIG. 3 is a flowchart showing a flow of processing to be stored in the document storage unit 2 of the original digitized document 1 created electronically.

  In step S01, a list of electronically created original documents 1 created electronically is displayed. For example, the digitized documents existing on the hard disk of the PC are displayed in a digitized document list as shown in FIG.

  In step S02, the electronically created original digitized document 1 stored in the document storage unit 2 is selected from the digitized document list displayed in step S01. For example, in the first embodiment of the present invention, “economic 001.doc” is selected as the electronically created original electronic document 1 stored in the document storage unit 2.

  In step S03, the electronic computerized original document 1 selected in step S02 is stored in the document storage unit 2. For example, when the document storage unit 2 for storing the original electronic document 1 created electronically is a document database, it is stored in the document database as shown in FIG.

  In step S04, if the electronically created original document 1 is to be stored in the document storage unit 2, the process returns to step S01, and the processes of steps S01 to 03 are repeated. If the storage is to be terminated, the process is terminated.

  FIG. 4 is a flowchart showing a flow of processing for printing the original electronically created electronic document 1 stored in the document storage unit 2 by the document printing unit 3.

  In step S11, a list of electronically created original documents 1 created electronically is displayed. For example, as in the example of step S01, an electronic document list existing in the hard disk or the like of the PC is displayed as shown in FIG.

  In step S12, an electronically created original electronic document 1 to be printed by the document printing unit 3 is selected from the electronic document list displayed in step S11. For example, in the first embodiment of the present invention, “economic 001.doc” is selected as the electronically generated original electronic document 1 to be printed by the document printing unit 3.

  In step S13, the original electronically created document 1 selected in step S12 is printed by the document printing unit 2. For example, when the document printing unit 3 that prints the electronically created original electronic document 1 is a printer, a paper document 4 is printed as shown in FIG.

  In step S14, when the original electronic document 1 created electronically is continued to be printed by the document printing unit 2, the process returns to step S11, and the processes of steps S11 to S13 are repeated. If the printing is to be terminated, the process is terminated.

  FIG. 5 shows an original electronic document 1 created electronically from a paper document 4 printed by the document printing unit 3 from the electronic computerized original document 1 stored in the document storage unit 2. 4 is a flowchart showing a flow of processing for detecting a document stored in a document storage unit 2.

  In step S21, the paper document 4 of the document printed in step S14 is set in the document reading unit 5. For example, when the document reading unit 5 is a scanner device, the paper document 4 of the printed document is set so that the paper document 4 can be read.

  In step S22, the paper document 4 set in step S21 is read by the document reading unit 5. For example, a paper document 4 of a printed document is read as image data when the document reading unit 5 is a scanner device.

  In step S23, the image data of the paper document 4 of the printed document read in step S22 is recognized as a character code by the document recognition unit 6. For example, when the document recognition unit 6 that recognizes a character code is OCR (Optical Character Recognition), only the character code recognized as a character is output.

  In step S24, the document analysis unit 7 analyzes the document data recognized in step S23, and extracts document layout information. For example, when the recognized document data is analyzed by the document analysis unit 7 and the layout information of the document is extracted and cannot be handled as the same document block in the layout as shown in FIG. 9, it is output as a plurality of document blocks. .

  In step S25, the electronically created original electronic document stored in step S03 is read from the document storage unit 2. For example, when the document storage unit 2 for storing the original electronic document 1 created electronically is a document database, it reads from the document database as shown in FIG.

  In step S26, a concept vector search process is performed on the document analyzed in step S24 (a document obtained by scanning a paper document) and the document read in step S25 (an electronically created original document). Do.

  In step S27, the document detection unit 8 detects the corresponding document from the result of the concept search process in step S26.

  In step S28, the corresponding document detected in step S27 is output by the detection result output unit 9.

  FIG. 6 is an example of a computerized document list. For example, an electronic document existing on a hard disk of a PC is a case where a document name list is displayed and selected.

  FIG. 7 shows a case of a document database as an example of the document storage unit 3. For example, this document database is for storing an electronically created original electronic document that is a target of document detection, and the document database column is composed of “document name” and “document data”. Yes. Further, “document name” and “document data” are document databases that correspond one-to-one.

  FIG. 8 is an example of a paper document 4 printed by the document printing unit 3. For example, it is a paper document when a document to be printed is selected from an electronic document list of an original electronic document created electronically and printed by a printer or the like.

  FIG. 9 shows an example of a document block configuration after text information is extracted from a paper document scanned by the document reading unit 5 by the OCR which is the document recognition unit 6 and the document layout is analyzed by the document analysis unit 7. For example, in the paper document of FIG. 8, character string data and image data are mixed. When OCR is performed on this document, only the character code recognized as a character is output. When the output document data is analyzed and the document layout is extracted, it cannot be handled as the same document book in terms of layout, and thus is output as two document blocks.

  FIG. 10 shows an example of detection result output. For example, as a result of detecting an original digitized document accumulated from a scanned paper document, the corresponding document is output. In the first embodiment of the present invention, the detection result output is displayed, and “Economic 001.doc” and “Economic 001 copy.doc” are displayed as corresponding document names.

  11 to 15 will explain the concept vector search processing for a document.

  FIG. 11 shows the configuration of the word vector dictionary. A word vector is a semantic vector (a list of feature quantities for each semantic classification), and the word vector dictionary is a set of meanings of each word based on meaning. Each dimension represents one semantic classification. Each word (meaning) stores a value (= feature value) indicating how much the semantic classification of each dimension is implied. For example, dimension 3 represents “space / sky”, dimension 4 represents “transaction / trading”, and dimension 7 represents “gesture / motion”. The word 7 represents the word “form”. The dimension 3 of the word 7 is 0, which means that the word “form” has no meaning related to the meaning classification of “space / sky”. The value of dimension 4 of word 7 is large and the value of dimension 7 is small. This means that “form (form)” has a strong meaning of “dealing / trading” but “gesturing / motion” is weak. On the other hand, the dimension 4 of the word 8 is small and the dimension 7 is large. This means that “form (posture)” has a strong meaning of “gesture / motion”, but “transaction / trading” has a weak meaning. In this way, the word vector dictionary shows what each word means.

  FIG. 12 shows a concept search index. The concept search index stores a concept vector (= document vector) corresponding to each document for each document. Each document vector indicates what semantic content each document represents. For example, it can be seen that the feature amounts of dimension 4 and dimension 7 of document ID = 6949 are 0.009 and 0.425, respectively, and the feature amounts of dimension 4 and dimension 8 of document ID = 6953 are 0.362 and 0.008, respectively. Thus, it can be seen that the document ID = 6949 is a sentence that hardly includes the semantic classification of “transaction / trading”, and the document ID = 6953 is a sentence that hardly includes the semantic classification of “gesture / motion”.

  FIG. 13 is a diagram showing a method for calculating the concept similarity when determining the conceptual similarity between documents. Concept search processing is processing for searching a document database for documents that are conceptually similar to a search query. In this embodiment, the search query is regarded as one document and its document vector is obtained (= query vector = document vector of the scanned document), and the document vector of the document in the search target document database (= document vector of the original digitized document) ) And the query vector are obtained as the concept similarity.

  The document vector X is an n-dimensional vector having values x1 to xn in each dimension. Similarly, the query vector Q is an n-dimensional vector. The similarity based on the cosine measure is expressed as SD (X, Q). The cosine measure SD (X, Q) is the value obtained by dividing the inner product of both vectors by the product of the norms of both vectors. In this embodiment in which both vectors are normalized by norm = 1, SD (X, P) corresponds to the inner product itself. Therefore, the sum of the products of the same dimension feature values can be obtained for both vectors.

  FIG. 14 is a flowchart showing the flow of a similar document detection process in which a concept vector search process is performed on a document.

  In step S31, the document analyzed in step S24 (a document obtained by scanning a paper document) is input.

  In step S32, document concept vector generation processing is performed on the document (scanned document) input in step S31.

  In step S33, the document vector generated in step S32 is processed into a search index to create a concept search index.

  In step S34, the document read in step S25 (original electronically created document) is input.

  In step S35, document concept vector generation processing is performed on the document (original document) input in step S33.

  In step S36, the document concept vector of the scanned document generated in step S32 is compared with the document concept vector of the original document generated in step S35, and the concept similarity of the document concept vector is generated. The result is output to the sort list.

  In the first embodiment of the present invention, a document unit, that is, a concept vector of the entire document is a comparison target.

  In step S37, when an unprocessed document exists in the document read in step S25 (original electronically created document), the process returns to step S34, and the processes from step S34 to S36 are repeated. If there is no unprocessed document, the process proceeds to S38.

  In step S38, a similar document is detected from the result of the concept similarity of the document concept vector in step S35.

  FIG. 15 is a flowchart showing a flow of processing of document concept vector generation.

  In step S41, a word is extracted from the document, and morphological analysis is performed using a morphological analysis dictionary.

  In step S42, after the morphological analysis, ambiguity of each word is resolved based on the analysis result. Various methods have been proposed to solve the ambiguity, so follow it. For example, ambiguity elimination based on matching between the dependency analysis result and the co-occurrence database, ambiguity elimination based on concept matching with the user profile, and the like can be considered. A plurality of meanings are output for words that are not sufficiently ambiguous.

  In step S43, a document concept vector is generated. A word vector dictionary is searched from the words extracted in steps S41 and S42 and the specified meaning, a feature quantity for each dimension is obtained for each word, and a document vector is generated from the sum. For words whose meaning cannot be specified, weights according to frequency are added to all meaning word vectors having the notation.

(Second embodiment)
The configuration and the flow of processing are the same as in the first embodiment of the present invention,
In step S36, the document concept vector of the scanned document generated in step S32 is compared with the document concept vector of the original document generated in step S35, and the concept similarity of the document concept vector is generated. The result is output to the sort list.

  In the second embodiment of the present invention, the document layout unit extracted in step S24, that is, the concept vector of each block of each document block constituting the entire document is compared for each document block.

(Third embodiment)
Hereinafter, a third embodiment of the present invention will be described with reference to FIGS.

  The block diagram showing the configuration of the character processing apparatus to which the third embodiment of the present invention is applied is the same as the block diagram showing the configuration of the character processing apparatus to which the first embodiment of the present invention is applied in FIG.

  FIG. 16 is a diagram showing a flow of processing in the third embodiment of the present invention.

  The third embodiment of the present invention includes an electronic document 2-1, a document storage unit 2-2, a paper document 2-3, a document reading unit 2-4, a document recognition unit 2-5, a document detection unit 2-6, It consists of a detection result output unit 2-7.

  Reference numeral 2-1 denotes an electronic document, which is an electronic document created electronically.

  Reference numeral 2-2 denotes a document storage unit that stores electronically created electronic documents of the electronic document 1-1.

  2-3 is a paper document, which is a printed or handwritten paper document.

  Reference numeral 2-4 denotes a document reading unit which reads a printed or handwritten paper document 2-3 as image data.

  A document recognizing unit 2-5 recognizes a printed or handwritten paper document read by the document reading unit 2-4 as a character code.

  A document detection unit 2-6 performs a concept search on the detection of the digitized document stored in the document storage unit 2 from the document data recognized by the document recognition unit 2-5, and detects similar documents. Do.

  A detection result output unit 2-7 outputs a result of the document detected by the document detection unit 2-6.

  The flowchart showing the flow of processing to be stored in the document storage unit of the electronically created electronic document according to the third embodiment of the present invention is created electronically according to the first embodiment of the present invention shown in FIG. It is the same as the flowchart showing the flow of processing stored in the document storage unit of the original electronic document.

  FIG. 17 is a flowchart showing a flow of processing for detecting a document stored in the document storage unit 2-2 of an electronically created document 2-1 from a printed or handwritten paper document 2-3. .

  In step S2-11, the printed or handwritten paper document 2-3 is set in the document reading unit 2-4. For example, when the document reading unit 2-4 is a scanner device, the paper document 2-3 that is printed or handwritten is set so that the paper document 2-3 can be read.

  In step S2-12, the document reading unit 2-4 reads the paper document 2-3 set in step S2-11. For example, a printed or handwritten paper document 2-3 is read as image data when the document reading unit 2-4 is a scanner device.

  In step S2-13, the image data of the paper document 2-3 of the printed or handwritten document read in step S2-12 is recognized as a character code by the document recognition unit 2-5. For example, when the document recognition unit 2-5 that recognizes a character code is OCR (Optical Character Recognition), the character code recognized as a character is output.

  In step S2-14, the electronically created digitized document stored in step S2-03 is read from the document storage unit 2-2. For example, when the document storage unit 2-2 that stores the electronically created electronic document 2-1 is a document database, it reads from the document database.

  The document database of the third embodiment of the present invention is the same as the document database of the first embodiment of the present invention shown in FIG.

  In step S2-15, the document concept for the document recognized in step S2-13 (a document obtained by scanning a paper document) and the document read in step S2-14 (electronically created document) is described. Perform vector search processing.

  In step S2-16, the document detection unit 2-6 detects the corresponding document from the result of the concept search process in step S2-15.

  In step S2-17, the corresponding document detected in step S2-16 is output by the detection result output unit 2-7.

  FIG. 18 is an example of an electronically created electronic document stored in the document storage unit 2-3. For example, an electronically created electronic document including “economy 001.doc” and “weather.doc” is stored.

  FIG. 19 shows an example of the document data after the text information is extracted from the paper document scanned by the document reading unit 2-4 by the OCR which is the document recognition unit 2-5. For example, when OCR is performed on a printed or handwritten paper document, a character code recognized as a character is output. In the third embodiment of the present invention, output related to consumption expenditure is made.

  FIG. 20 shows an example of detection result output. For example, a similar document is output as a result of detecting an electronic document stored from a scanned paper document. In the third embodiment of the present invention, the detection result output is a display case, and since the content of the scanned document is document data related to consumption expenditure, it is determined that the similarity of “economy” is high and is similar. “Economy 001.doc” is displayed as the document name.

  The description of the concept vector search process for the document according to the third embodiment of the present invention is the same as the description of the concept vector search process for the document according to the first embodiment of the present invention shown in FIGS.

It is a block diagram which shows the whole structure of the 1st Example of this invention. It is a figure which shows the flow of a process of 1st Example of this invention. It is a flowchart which shows the flow of a process of the document storage of 1st Example of this invention. It is a flowchart which shows the flow of a document printing process of 1st Example of this invention. It is a flowchart which shows the flow of a document detection process of 1st Example of this invention. It is an example of the electronic document list of 1st Example of this invention. This is a case of a document database as an example of the document storage unit 2 of the first embodiment of the present invention. It is an example of the paper document 4 printed by the document printing part 3 of 1st Example of this invention. Document block configuration after text information is extracted from a paper document scanned by the document reading unit 5 of the first embodiment of the present invention by the OCR which is the document recognition unit 6 and the document layout is analyzed by the document analysis unit 7 It is an example. It is an example of the detection result output of 1st Example of this invention. 1 shows the configuration of a word vector dictionary according to a first embodiment of the present invention. 2 shows a concept search index according to the first embodiment of the present invention. It is the figure which showed the calculation method of the concept similarity degree at the time of determining the conceptual similarity of the document of 1st Example of this invention, and a document. It is a flowchart which performs the concept vector search process with respect to the document of 1st Example of this invention, and shows the flow of a similar document detection process. It is a flowchart which shows the flow of a process of the document concept vector production | generation of 1st Example of this invention. It is a figure which shows the flow of a process of the 3rd Example of this invention. It is a flowchart which shows the flow of a document detection process of the 3rd Example of this invention. It is an example of the printed or handwritten paper document 2-3 of the 3rd Example of this invention. It is an example of the document data after extracting the text information from the paper document scanned by the document reading part 2-4 of the 3rd Example of this invention by OCR which is the document recognition part 2-5. It is an example of the detection result output of 3rd Example of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Electronic document 2 Document storage part 3 Document printing part 4 Paper document 5 Document reading part 6 Document recognition part 7 Document analysis part 8 Document detection part 9 Detection result output part 2-1 Electronic document 2-2 Document storage part 2- 3 Paper Document 2-4 Document Reading Unit 2-5 Document Recognition Unit 2-6 Document Detection Unit 2-7 Detection Result Output Unit

Claims (12)

In a document detection method and system for detecting a document,
Document storage means for storing electronically created original electronic documents;
Document reading means for reading a document printed from the original electronic document as image data;
Document recognition means for recognizing image data read by the document reading means as a character code;
A document detection unit that performs a concept search on the detection of the original digitized document stored by the document storage unit from the document data recognized by the document recognition unit, and detects a corresponding document;
Detection result output means for outputting the result detected by the document detection means;
A document detection method and system.   2. The document detection method and system according to claim 1, wherein in the document detection means, the concept search is a concept vector search for a document. In a document detection method and system for detecting a document,
Document storage means for storing electronically created original electronic documents;
Document reading means for reading a document printed from the original electronic document as image data;
Document recognition means for recognizing image data read by the document reading means as a character code;
Document analysis means for analyzing document data recognized by the document recognition means, and extracting document layout information;
A document detection unit that performs a concept search on the detection of the original digitized document stored in the document storage unit from the document data analyzed by the document analysis unit, and detects a corresponding document;
Detection result output means for outputting the result detected by the document detection means;
A document detection method and system.   4. The document detection method and system according to claim 3, wherein in the document detection means, the concept search is a concept vector search for a document.   4. The document detection method and system according to claim 3, wherein the document detection unit detects a document to be detected in units of documents.   4. The document detection method and system according to claim 3, wherein in the document detection means, the target of the document to be detected is a document layout unit extracted by the document analysis means. In a document detection method and system for detecting a document,
Document storage means for storing electronically created electronic documents;
Document reading means for reading a printed or handwritten document as image data;
Document recognition means for recognizing image data read by the document reading means as a character code;
A document detection means for detecting a similar document by performing a concept search on the detection of a similar document of the digitized document stored by the document storage means from the document data recognized by the document recognition means;
Detection result output means for outputting the result detected by the document detection means;
A document detection method and system.   8. The document detection method and system according to claim 7, wherein in the document detection means, the concept search is a concept vector search for a document. In a document detection method and system for detecting a document,
Document storage means for storing electronically created electronic documents;
Document reading means for reading a printed or handwritten document as image data;
Document recognition means for recognizing image data read by the document reading means as a character code;
Document analysis means for analyzing document data recognized by the document recognition means, and extracting document layout information;
A document detection means for detecting a similar document by performing a concept search on the detection of a similar document of the digitized document stored by the document storage means from the document data analyzed by the document analysis means;
Detection result output means for outputting the result detected by the document detection means;
A document detection method and system.   10. The document detection method and system according to claim 9, wherein in the document detection means, the concept search is a concept vector search for a document.   10. The document detection method and system according to claim 9, wherein the document detection means detects a document to be detected in units of documents. 10. The document detection method and system according to claim 9, wherein in the document detection unit, a target of the document to be detected is a document layout unit extracted by the document analysis unit.

Images