Method and apparatus for identification and correction of document skew

• US 5,187,753 A
• Filed: 12/08/1989
• Issued: 02/16/1993
• Est. Priority Date: 12/08/1989
• Status: Expired due to Term

• Current Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Original Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Inventors: Bloomberg, Dan S., Kopec, Gary E.
• Primary Examiner(s): Boudreau, Leo H.
• Assistant Examiner(s): CAMMARATA, MICHAEL
• Application Number:
  US07/448,774
• Time in Patent Office: 1,166 Days
• Field of Search: 382/30, 382/46, 382/51, 382/18, 358/488, 340/727, 395/137
• US Class Current: 382/289
• CPC Class Codes: G06T 7/70 Determining position or orientation of objects or cameras camera calibration G06T7/80
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/1478 of characters or characters lines
• CPC Class Codes: H04N 1/3878 Skew detection or correction

Abstract

A method and apparatus for identifying and correcting for document skew. Lines of a bitmap are scanned and a variance in the number of ON pixels as a function of skew angle is calculated. Skew of the original document occurs when the variance is a maximum. Once the skew has been identified, the document is deskewed accordingly.

Claims

1. A method of determining skew of an image in an optical character recognition device (OCR) comprising the steps of optically scanning a document to produce an image of said document, said image in the form of a digital data structure;

reducing said image using a thresholded reduction, said thresholded reduction mapping N×

N squares of pixels into a destination image, said destination image having an ON pixel when a number of ON pixels in said square is equal to or greater than a threshold value;

determining a variance in a number of ON pixels in at least one scanline of said image as a function of document rotation angle, said variance including the sum of the number of ON pixels in said at least one scan line raised to a power greater than one, said skew located at a document rotation angle at which said variance is substantially a maximum;

rotating said image about an angle of about said skew to produce a deskewed image of said document; and

identifying characters in said deskewed image of said document.

3. The method as recited in claims 1, or 2, wherein substantially all of the scanlines in a document are included the calculation of said variance.

4. The method as recited in claims 1, or 2, wherein document rotation angles are simulated by a vertical shear of said image.

5. The method as recited in claims 1, or 2, wherein variance is calculated through a range of about +5 to -5 degrees of rotation.

6. The method as recited in claims 1, or 2, wherein variance is calculated at increments of about 0.2 degrees of rotation.

7. The method as recited in claim 1 or 2 wherein said step of rotating using at least about two orthoganal shears.

8. The method as recited in claim 1 or 2 further comprising the step of identifying characters in said deskewed document image using an optical character recognition system.

9. The method as recited in claim 1 or 2 further comprising the step of reproducing said deskewed document image.

2. A method of determining skew of an image in an optical character recognition device (OCR) comprising the steps of optically scanning a document to produce an image of said document, said image in the form of a digital data structure;

reducing said image using a thresholded reduction, said thresholded reduction mapping N×

N squares of pixels into a destination image, said destination image having an OFF pixel when the number of OFF pixels in said square is equal to or greater than a threshold value;

determining a variance in a number of OFF pixels in at least one scanline of said image as a function of document rotation angle, said variance including the sum of the number of OFF pixels in said at least one scanline raised to a power greater than one, said skew located at a document rotation angle at which said variance is substantially a maximum;

rotating said image about an angle of about said skew to produce a deskewed image of said document; and

identifying characters in a deskewed image of said document.

10. An optical character recognition device (OCR) programmed to determine skew of an image of a document, said OCR comprising:

an optical scanner producing an image of said document, said image in the form of a digital data structure;

a digital processor programmed to input said image of said document;

reduce said image using a thresholded reduction, said thresholded reduction mapping N×

N squares of pixels into a destination image, said destination image having an ON pixel when a number of ON pixels in said square is equal to or greater than a threshold value; and

determine a variance in a number of ON pixels in at least one scanline of said image as a function of document rotation angle, said variance including the sum of the number of ON pixels in said at least one scan line raised to a power greater than one, said skew located at a document rotation angle at which said variance is substantially a maximum;

rotating said image about an angle of about said skew to produce a deskewed image of said document; and

means for identifying characters in said deskewed image of said document.

12. The digital processing device as recited in claims 17, or 18, programmed to input substantially all of the scanlines in a document are to calculate said variance.

13. The digital processing device as recited in claims 10, or 11, programmed to simulate said rotation angle by a vertical shear of said image.

14. The digital processing device as recited in claims 10, or 11, programmed to calculate said variance is over a range of about ±

5 degrees of rotation.

15. The digital processing device as recited in claims 10, or 11, programmed to calculate said variance is at increments of about 0.2 degrees of rotation.

16. The digital processing device as recited in claim 10 or 11 programmed to rotate said image using at least about two orthogonal shears.

17. The digital processing device as recited in claim 10 or 11 further comprising an optical character recognition system for identifying characters in said image.

18. The digital processing device as recited in claim 10 or 11 further comprising reproduction means for reproducing said deskewed document image.

11. An optical character recognition device (OCR) programmed to determine skew of an image of a document, said OCR comprising:

an optical scanner producing an image of said document, said image in the form of a digital data structure;

a digital processor programmed to input said image of said document;

reduce said image using a thresholded reduction, said thresholded reduction mapping N×

N squares of pixels into a destination image, said destination image having an OFF pixel when a number of OFF pixels in said square is equal to or greater than a threshold value; and

determine a variance in a number of OFF pixels in at least one scanline of said image as a function of document rotation angle, said variance including the sum of the number of ON pixels in said at least one scan line raised to a power greater than one, said skew located at a document rotation angle at which said variance is substantially a maximum;

rotating said image about an angle of about said skew to produce a deskewed image of said document; and

means for identifying characters in said deskewed image of said document.

19. An electroreprographic (ER) copier programmed to determine skew of an image of a document, said ER copier comprising

an optical scanner producing an image of said document, said image in the form of a digital data structure;

a digital processor programmed to input said image of said document;

reduce said image using a thresholded reduction, said thresholded reduction mapping N×

N squares of pixels into a destination image, said destination image having an ON pixel when a number of ON pixels in said square is equal to or greater than a threshold value; and

determine a variance in a number of ON pixels in at least one scanline of said image as a function of document rotation angle, said variance including the sum of the number of ON pixels in said at least one scan line raised to a power greater than one, said skew located at a document rotation angle at which said variance is substantially a maximum;

means for rotating said image about an angle of about said skew to produce a deskewed image of said document; and

means for outputting said deskewed image of said document.

20. An electroreprographic (ER) copier programmed to determine skew of an image of a document, said ER copier comprising:

an optical scanner producing an image of said document, said image in the form of a digital data structure;

a digital processor programmed to input said image of said document;

reduce said image using a thresholded reduction, said thresholded reduction mapping N×

N squares of pixels into a destination image, said destination image having an OFF pixel when a number of OFF pixels in said square is equal to or greater than a threshold value; and

determine a variance in a number of OFF pixels in at least one scanline of said image as a function of document rotation angle, said variance including the sum of the number of ON pixels in said at least one scan line raised to a power greater than one, said skew located at a document rotation angle at which said variance is substantially a maximum;

means for rotating said image about an angle of about said skew to produce a deskewed image of said document; and

means for outputting said deskewed image of said document.