Certifiable optical character recognition

• US 5,625,721 A
• Filed: 04/06/1994
• Issued: 04/29/1997
• Est. Priority Date: 10/09/1992
• Status: Expired due to Fees

• Current Assignee: Matsushita Information Technology Laboratory
• Original Assignee: Matsushita Information Technology Laboratory
• Inventors: Lipton, Richard J., Korth, Henry F., Lopresti, Daniel P., Sandberg, Jonathan S.
• Primary Examiner(s): Mancuso, Joseph
• Application Number:
  US08/223,830
• Time in Patent Office: 1,119 Days
• Field of Search: 382/57, 382/61, 382/9, 382/8, 382/176, 382/180, 382/309, 382/310, 382/311, 358/403, 358/405, 364/737, 371/53, 371/52, 371/57.2
• US Class Current: 382/309
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/12 Detection or correction of errors, e.g. by rescanning the pattern
• CPC Class Codes: G06V 30/1444 Selective acquisition, locating or processing of specific regions, e.g. highlighted text, fiducial marks or predetermined fields

Abstract

A certificate, calculated on the text of a document to be reproduced, is included in the document in a form which is machine-readable. Upon optical character recognition scanning, or other digitizing reproduction, the certificate is also scanned and stored for comparison. A new certificate is calculated on the contents of the reproduced document and compared to the scanned certificate. If the two certificates are different, it is known that at least one error exists in the reproduced document. More than one certificate can be provided for a document in order to facilitate error location on the document and error correction therein.

Claims

1. A method for ensuring the accuracy of lines of reproduced data electronically reproduced from lines of original data comprising:

first deriving one or more data-dependent first values from said original data;

second deriving one or more data-dependent second values from said reproduced data;

comparing said first and second values and identifying errors and error locations within said lines of reproduced data based on the difference between said first and second values; and

effecting one or more methods of automatic correction of said identified errors on said reproduced data.

2. The method of claim 1 wherein said deriving comprises calculating said data-dependent values.

3. The method of claim 1 wherein said automatic correction produces error-corrected reproduced data and further comprising the steps of:

deriving one or more data-dependent third values to said error-corrected reproduced data; and

comparing said first and third values.

4. A method for electronically reproducing lines of original data of computer-preprocessed printed documents comprising the steps of:

deriving one or more first data-dependent values from said original data;

printing said document including said one or more first values;

creating an electronic document having lines of reproduced data by optically scanning said printed document and said first values and saving said electronic document and first values;

deriving one or more second data-dependent values from said reproduced data of said electronic document;

comparing said first and second values and identifying errors and error locations within said lines of reproduced data based on the difference between said first and second values; and

automatically correcting said identified errors by applying one or more error correction methods to said reproduced data.

5. The method of claim 4 wherein said deriving of first and second values comprises applying the same algorithm to the data of each of said computer-generated and electronic documents.

6. The method of claim 4 wherein said creating and said deriving of said second values are conducted simultaneously.

7. The method of claim 4 wherein said automatic correcting of said identified errors produces a second electronic document and further comprising the steps of:

deriving one or more third data-dependent values to said second electronic document; and

comparing said first and third data-dependent values.

8. The method of claim 4 wherein said printing of said one or more first values comprises providing said one or more first values in a non-textual machine-readable symbology.

9. A method for providing an electronic reproduction having lines of reproduced data from a printed document having lines of original data using optical character recognition comprising the steps of:

creating a first computer version of said document in a first computer;

deriving at least one first digital value for said first version of said document from said original data;

creating a printed version of said document by outputting said first version and said at least one first digital value to a printer associated with said first computer;

creating a second computer version of said document comprising reproduced data by optically scanning said printed version and said at least one first digital value with a scanner associated with a second computer having software for recognizing said optical characters and said digital value;

deriving at least one second digital value for said second computer version of said document from said reproduced data;

comparing said at least one first digital value to said at least one second digital value and identifying errors and error locations within said lines of reproduced data based on the difference between said at least one first and second digital values; and

effecting automatic error correction of said second computer version by applying one or more error correction methods to said reproduced data.

10. The method of claim 9 wherein said printed document comprises a plurality of blocks comprising at least multiple text blocks of multiple lines of original data and wherein said deriving of said at least one first digital value includes providing at least one first digital characterization of the spatial relationship of said plurality of blocks in said document.

11. The method of claim 9 wherein said deriving of said at least one first and second digital values comprises calculating said values on the data in each of said first and second computer versions of said documents, respectively.

12. The method of claim 9 further comprising the steps of deriving at least one third digital value to said second computer version of said document after said error correction and comparing said at least one first digital value to said at least one third digital value.

13. The method of claim 9 whereby said effecting automatic error correction produces a third computer version and further comprising the steps of:

deriving at least one third digital value to said third computer version; and

comparing said first and third digital values.

14. The method of claim 9 wherein said at least one first digital value is printed on said document in non-textual machine-readable symbology.

15. A method for electronically reproducing a computer-preprocessed printed document having a layout comprising a plurality of original blocks comprising at least multiple text blocks of multiple lines of characters comprising the steps of:

deriving at least one first layout-dependent value from the spatial relationship of said plurality of original blocks on said document;

printing said multiple text blocks of multiple lines and said at least one first layout-dependent value;

creating an electronic document with layout comprising a plurality of reproduced blocks comprising at least multiple text blocks of multiple lines of reproduced characters by optically scanning said multiple lines and said at least one first layout-dependent value;

deriving at least one second layout-dependent value from the spatial relationship of said plurality of reproduced blocks on said electronic document;

comparing said at least one first layout-dependent value to said at least one second layout-dependent value; and

detecting and locating layout errors based upon the difference between said at least one first and second layout-dependent values.

16. The method of claim 15 further comprising the steps of:

deriving at least one first character-dependent value from said characters or said document; and

printing said at least one first character-dependent value upon printing of said multiple lines and said at least one first layout dependent value.

17. The method of claim 16 further comprising the steps of:

deriving at least one second character-dependent value from said reproduced characters of said electronic document;

comparing said at least one first character-dependent value to said at least one second character-dependent value and identifying character errors and error locations in said lines based upon the differences between said first and second character-dependent values; and

effecting one or more methods of automatic correction of said identified character errors on said reproduced characters.

18. A method for insuring the accuracy of a second document comprising reproduced data sequences reproduced by optically scanning a first document, comprising original data sequences and at least one encoded, machine-readable, first data-dependent value assigned to and scannable on the face of said first document, comprising the steps of:

decoding said at least one encoded, machine-readable, first data-dependent value;

deriving at least one second data-dependent value for said second document from said reproduced data sequences;

comparing said at least one first data-dependent value to said at least one second data-dependent value and identifying errors and error locations within said reproduced data sequences based upon the differences between said first and said second data-dependent values; and

effecting automatic error correction on said reproduced data sequences when said first and said second data-dependent values are different.