Method of recognizing and indexing documents

• US 7,319,799 B2
• Filed: 03/22/2001
• Issued: 01/15/2008
• Est. Priority Date: 03/22/2000
• Status: Expired due to Fees

• Current Assignee: Oce-Industries Incorporated
• Original Assignee: Oce-Industries Incorporated
• Inventors: Berche, Ste, Najman, Laurent
• Primary Examiner(s): Mehta; Bhavesh M
• Assistant Examiner(s): KIM, CHONG R
• Application Number:
  US09/813,955
• Publication Number:
  US 20010033690A1
• Time in Patent Office: 2,490 Days
• Field of Search: 382174-181, 382/199, 382/200, 382/229, 382/321
• US Class Current: 382/306
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/1444 Selective acquisition, locating or processing of specific regions, e.g. highlighted text, fiducial marks or predetermined fields
• CPC Class Codes: Y10S 707/99932 Access augmentation or optimizing

Abstract

A method of recognizing and indexing documents, using a scanner connected to a computer, the method including scanning the documents, then using a pointing device or member of the computer to designate an arbitrary point P in at least one box of the documents, and finally recognizing by OCR the characters in the box so as to store them in a first database connected to the computer to enable documents scanned in this way to be indexed. The designation step comprises a step of searching for and identifying the box of the document containing the point P designated by the user.

Claims

1. A method of recognizing and indexing documents in a system having a scanner connected to a computer, the method comprising:

scanning a document;

designating, by a user, an arbitrary point P in at least one box of the scanned document using a pointing device or member of the computer, wherein no extracting or analyzing process takes place previous to the user'"'"'s designation;

searching for and identifying said box by applying a shape search algorithm over a determined search zone surrounding said point P previously designated by the user, wherein if the determined search zone does not include the entire box, using at least one additional search zone of increasing size until the entire box is within the search zone;

recognizing by OCR the characters in said identified box of the scanned document;

storing the recognized characters in a first database connected to the computer to enable documents scanned in this way to be indexed; and

storing, in a second database connected to the computer, characterization data of said box of the scanned document, such that another box subsequently can be identified automatically without any point P within said another box being designated, for next documents of a same type.

2. The method according to claim 1, wherein said shape search algorithm is a projection algorithm which counts the number of pixels present in each vertical or horizontal line of said determined search zone and which, on the basis of these count numbers, finds the horizontal and vertical lines present in said search zone by examining the peaks in the X and Y projection profiles.

3. The method according to claim 1, wherein said shape search algorithm is an algorithm based on the Hough transform.

4. The method according to claim 1, wherein said OCR step is preceded by a step in which a user defines the type of character to be recognized in said box of the scanned document.

5. The method according to claim 1, wherein said scanning step is performed initially for a set of documents to be processed, with said steps of identifying the box and performing OCR being performed subsequently in succession for each of the documents.

6. The method according to claim 1, wherein said scanning step is initially performed for a first document, with said steps of identifying the box and performing OCR subsequently being performed on that document so as to define a sequence, with said sequence then being repeated in succession for each of the documents to be processed.

7. The method according to claim 1, wherein said documents to be recognized and indexed are a set of technical drawings of the same or different types.

8. The method according to claim 1, wherein said documents to be recognized and indexed are a set of forms, of the same or different types.

9. An apparatus for recognizing and indexing documents, the apparatus comprising:

a scanner for scanning a document and delivering an image of the scanned document;

a computer connected to the scanner to receive said scanned image;

a first database connected to said computer for storing said scanned image;

first software for using a pointing member to designate, by a user, an arbitrary point P in at least one box of the scanned image, wherein no extracting or analyzing process takes place previous to the user'"'"'s designation, for searching for and identifing the box containing said point P designated by the user, for recognizing by OCR the characters in said box of the scanned image, and for storing the recognized characters so as to enable images scanned in this way to be indexed, wherein the searching for and identifying said box is performed by applying a shape search algorithm over a determined search zone surrounding said point P previously designated by the user, wherein if the determined search zone does not include the entire box, using at least one additional search zone of increasing size until the entire box is within the search zone;

a second database connected to the computer to store characterization data of said box of the scanned image, such that another box subsequently can be identified automatically by said first software without any point P within said another box being designated, for next documents of a same type.

10. The apparatus according to claim 9, further comprising second software for defining the type of data to be recognized in said box of the scanned image.

11. The apparatus according to claim 9, wherein the first and second databases are integrated in a memory of the computer.

12. The apparatus according to claim 9, wherein said pointing member includes a keyboard of the computer or a finger of the user.

13. A computer-readable medium having embodied thereon software to be processed by a computer connected so as to receive a scanned image, the software being operable to cause said computer to perform the functions of the first software of claim 9.

14. The apparatus of claim 9, wherein the first software includes a shape search algorithm.

15. The apparatus of claim 9, wherein the first software includes a projection algorithm for counting the number of pixel present in each vertical or horizontal line of a determined search zone surrounding said previously designated point P.

16. The apparatus of claim 15, wherein the projection algorithm locates, based on the counted numbers, the horizontal or vertical lines present in the search zone by analyzing peaks in X or Y projection profiles.

17. A method of recognizing and indexing documents in a system having a scanner connected to a computer, the method comprising:

scanning a document;

manually designating, by a user an arbitrary point P in a predetermined area of the scanned document, if a type of the scanned document is not known, wherein no extracting or analyzing process takes place previous to the user'"'"'s designation;

searching for and identifying a box around the arbitrary point P of the scanned document by applying a shape search algorithm over a determined search zone surrounding the arbitrary point P designated by the user, wherein if the determined search zone does not include the entire box, using at least one additional search zone of increasing size until the entire box is within the search zone;

storing, in a database connected to the computer, characterization data of the identified box of the scanned document, such that boxes in next documents of a same type can be identified automatically without designation of an arbitrary point P on the next documents;

recognizing characters in the identified box of the scanned document; and

storing the recognized characters to index the scanned document.