Document retrieving method and apparatus

• US 20040243601A1
• Filed: 04/26/2004
• Published: 12/02/2004
• Est. Priority Date: 04/30/2003
• Status: Active Grant

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Toshima, Eiichiro
• Granted Patent: US 7,257,567 B2
• Time in Patent Office: Days
• US Class Current: 707/100
• CPC Class Codes: G06F 16/93 Document management systems
• CPC Class Codes: G06V 30/40 Document-oriented image-based pattern recognition
• CPC Class Codes: G06V 30/418 Document matching, e.g. of document images
• CPC Class Codes: Y10S 707/99931 Database or file accessing
• CPC Class Codes: Y10S 707/99936 Pattern matching access

Abstract

In the proposed document retrieving apparatus, text feature data that bases upon text data included in a document and image feature data that bases upon a document image are stored in a memory. Image data of a search document is subjected to character recognition processing, text feature data is acquired based on the obtained text data, and image feature data (layout data) is acquired based on the image data of the search document. Using the text feature data and image feature data acquired with respect to the search document, a memory is searched, and a document corresponding to the search document is retrieved from plural documents.

Claims

1. A document retrieving method comprising:

a first acquisition step of acquiring text data based on image data of a document and acquiring text feature data based on the acquired text data;

a second acquisition step of acquiring layout feature data based on the image data of the document;

a storing step of storing in storage means the text feature data and the layout feature data, respectively acquired in said first and second acquisition steps, in association with the document; and

a retrieving step of retrieving a document stored in the storage means using the text feature data and the layout feature data acquired by executing said first and second acquisition steps on a search document.

2. The method according to claim 1, wherein in said retrieving step, a similarity level between the search document and each of the plurality of documents is obtained with respect to the text feature data and the layout feature data, and a search result is determined based on the two similarity levels obtained.

3. The method according to claim 2, wherein in said retrieving step, a search procedure is switched in accordance with an analysis result of image data of the search document and/or text data obtained from the image data.

4. The method according to claim 3, wherein in said retrieving step, a weight added to a calculation result of the similarity level obtained with respect to the text feature data and a calculation result of the similarity level obtained with respect to the layout feature data is changed in accordance with the analysis result.

5. The method according to claim 1, wherein in said retrieving step,

a first narrowing-down step is performed for obtaining a candidate document from the documents stored in the storage means using either the text feature data or the layout feature data acquired in said first and second acquisition steps;

and

a second narrowing-down step is performed on the candidate document, obtained in said first narrowing-down step, using the text feature data and the layout feature data for obtaining a search result.

6. The method according to claim 5, wherein in said retrieving step, whether the text feature data or the layout feature data is to be used for said first narrowing-down step is determined based on the analysis result of the image data of the search document, and/or the analysis result of the text data obtained from the image data.

7. The method according to claim 6, wherein in said retrieving step, whether the text feature data or the layout feature data is to be used for said first narrowing-down step is determined based on an amount of text image data included in the image data of the search document and a precision evaluation of a result of character recognition performed on the text image data.

8. The method according to claim 5, wherein in said second narrowing-down step, a similarity level between the search document and each of the documents stored in the storage means is obtained with respect to the text feature data and the layout feature data, and a search result is determined based on the similarity levels obtained.

9. The method according to claim 8, wherein in said second narrowing-down step, a weight added to the calculation result of the similarity level obtained with respect to the text feature data and the calculation result of the similarity level obtained with respect to the layout feature data is changed in accordance with the analysis result of the image data of the search document, and/or the analysis result of the text data obtained from the image data.

10. The method according to claim 1, wherein the layout feature data is acquired based on average brightness data and/or color data of n×

m rectangles into which the image data of the document is divided.

11. The method according to claim 1, wherein the layout feature data is acquired based on position and/or size of blocks which are obtained by block analysis of the image data of the document.

23. A storage medium storing a control program which causes a computer to execute the document retrieving method described in claim 1.

24. A control program which causes a computer to execute the document retrieving method according to claim 1.

12. A document retrieving apparatus comprising:

a first acquisition unit configured to acquire text data based on image data of a document and acquire text feature data based on the acquired text data;

a second acquisition unit configured to acquire layout feature data based on the image data of the document;

a storage unit configured to store the text feature data and the layout feature data, respectively acquired by said first and second acquisition units, in association with the document; and

a retrieving unit configured to retrieve a document stored in said storage unit using the text feature data and the layout feature data acquired by executing said first and second acquisition units on a search document.

13. The apparatus according to claim 12, wherein said retrieving unit obtains a similarity level between the search document and each of the plurality of documents with respect to the text feature data and the layout feature data, and determines a search result based on the two similarity levels obtained.

14. The apparatus according to claim 13, wherein said retrieving unit switches a search procedure in accordance with an analysis result of image data of the search document and/or text data obtained from the image data.

15. The apparatus according to claim 14, wherein said retrieving unit changes a weight added to a calculation result of the similarity level obtained with respect to the text feature data and a calculation result of the similarity level obtained with respect to the layout feature data, in accordance with the analysis result.

16. The apparatus according to claim 12, wherein said retrieving unit performs a first narrowing-down process for obtaining a candidate document from the documents stored in said storage unit using either the text feature data or the layout feature data acquired by said first and second acquisition units, and performs a second narrowing-down process on the candidate document, obtained in the first narrowing-down process, using the text feature data and the layout feature data for obtaining a search result.

17. The apparatus according to claim 16, wherein said retrieving unit determines whether the text feature data or the layout feature data is to be used for the first narrowing-down process based on the analysis result of the image data of the search document, and/or the analysis result of the text data obtained from the image data.

18. The apparatus according to claim 17, wherein said retrieving unit determines whether the text feature data or the layout feature data is to be used for the first narrowing-down process based on an amount of text image data included in the image data of the search document and a precision evaluation of a result of character recognition performed on the text image data.

19. The apparatus according to claim 16, wherein in the second narrowing-down process, a similarity level between the search document and each of the documents stored in said storage unit is obtained with respect to the text feature data and the layout feature data, and a search result is determined based on the similarity levels obtained.

20. The apparatus according to claim 19, wherein in the second narrowing-down process, a weight added to the calculation result of the similarity level obtained with respect to the text feature data and the calculation result of the similarity level obtained with respect to the layout feature data is changed in accordance with the analysis result of the image data of the search document, and/or the analysis result of the text data obtained from the image data.

21. The apparatus according to claim 12, wherein the layout feature data is acquired based on average brightness data and/or color data of n×

m rectangles into which the image data of the document is divided.

22. The apparatus according to claim 12, wherein the layout feature data is acquired based on position and/or size of blocks which are obtained by block analysis of the image data of the document.