Black white image scaling for optical character recognition

• US 20050196071A1
• Filed: 03/04/2004
• Published: 09/08/2005
• Est. Priority Date: 03/04/2004
• Status: Active Grant

• Current Assignee: Toshiba Global Commerce Solutions Holdings Corporation (Toshiba Corporation)
• Original Assignee: International Business Machines Corporation
• Inventors: Prakash, Ravinder, Thompson, Gerhard Robert
• Granted Patent: US 7,593,600 B2
• Time in Patent Office: Days
• US Class Current: 382/298
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/166 Normalisation of pattern dimensions

Abstract

A system and method for capturing and scaling images includes a scaling engine with the ability to employ a first scaling factor in a first direction and a second scaling factor in a second direction. In addition, the preferred scaling engine manipulates the scaling process so that the scaling calculations are performed using fixed point arithmetic. The preferred scaling engine preserves isolated features such as a single white pixel in a field of black pixels and vice versa. Improved readability is achieved in one embodiment by performing the scaling process multiple times using different degrees of “padding” where padding refers to the technique of surrounding an image with a perimeter of one or more blank (white) pixel elements.

Claims

1. A method for scaling a bi-tonal image, comprising:

responsive to receiving a scanned image, determining a first-scale factor;

processing each line of the image oriented parallel to a first axis by;

determining the pixel position of each pixel transition in the line;

dividing the pixel position by the first-scale factor to determine a corresponding scaled pixel transition; and

if the scaled pixel transition equals the scaled pixel transition of a previously processed pixel transition, adjusting the scaled pixel position of the current pixel transition to prevent the previous pixel transition from being overwritten in the scaled image.

2. The method of claim 1, wherein determining the scaled pixel position includes determining the pixel position using fixed point arithmetic exclusively.

3. The method of claim 2, wherein determining the pixel position using fixed point arithmetic exclusively includes:

multiplying each of the determined pixel locations by 10^N, where N is an integer between 2 and 5; and

dividing the result of the multiplication by the first-scale factor multiplied by 10^N.

4. The method of claim 2, further, comprising:

determining a second-scale factor, wherein the second scale is different than the first scale factor; and

upon processing each of the scanned image lines oriented parallel to the first axis to obtain an interim scaled image, processing each line of the interim scaled image parallel to a second axis including, for each line;

determining the pixel position of each pixel transition in the line;

dividing the pixel position by the second scale factor to determine a corresponding scaled pixel transition; and

if the scaled pixel transition equals the scaled pixel transition of a previously processed pixel transition, adjusting the scaled pixel position of the current pixel transition to prevent the previous pixel transition from being overwritten in the scaled image.

5. The method of claim 4, further comprising, processing a first image using first x and y scale factors and scaling a second image using second x and y scale factors wherein the first and second scale factors differ and generating a document containing the first and second images.

6. The method of claim 5, wherein generating a document comprises generating a bank statement and wherein the first and second scanned images comprise first and second check copies.

7. The method of claim 1, further comprising, prior to processing each of the image lines, padding the image with a perimeter of empty pixels.

8. The method of claim 7, further comprising:

performing multiple iterations of the method of claim 7, wherein the number of empty cells in the perimeter varies for each iteration; and

selecting the scaled image from the set of scaled images produced by the multiple iterations.

9. A computer program product comprising computer executable instructions, stored on a computer readable medium, for scaling an image, comprising:

computer code means for scaling an image in a first direction using a first scaling factor;

computer code means for scaling an image in a second direction according to a second scaling factor;

wherein at least one of a set of image lines in the image includes an isolated pixel and further wherein the code means for scaling the image includes code means for preserving the isolated pixel in the scaled image.

10. The computer program product of claim 9, wherein the code means for scaling the image in the first direction includes:

code means for determining, for each line oriented in a first direction of the image, the pixel value of each pixel transition in the line; and

code means for dividing each pixel transition value by the first scale factor to determine a value for a corresponding pixel in the scaled image.

11. The computer program product of claim 10, wherein the code means for dividing each pixel transition includes codes means, responsive to determining that a determined pixel transition value equals a previously determined pixel transition value, for altering the value of the current pixel transition value to preserve the previously recorded pixel transition values.

12. The computer program product of claim 11, wherein the code means for dividing each pixel transition value include code means for multiplying each pixel transition value by an integer value N and then dividing the resulting product by N.

13. The computer program product of claim 12, wherein the code means for dividing and multiplying each pixel transition value includes code means for multiplying and dividing each pixel transition value using integer arithmetic exclusively.

14. The computer program product of claim 13, further comprising:

code means for determining, for each line oriented in a second direction of the image, the pixel value of each pixel transition in the line; and

code means for dividing each pixel transition value by the second scale factor to determine a value for a corresponding pixel in the scaled image.

15. The computer program product of claim 14, further comprising code means for dividing each pixel transition including code means, responsive to determining that a determined pixel transition value equals a previously determined pixel transition value, for altering the value of the current pixel transition value to preserve the previously recorded pixel transition values.

16. The computer program product of claim 15, further comprising, code means for processing a first image using a first set of first and second scale factors and scaling a second image using a second set of scale factors wherein the first and second sets of scale factors differ.

17. The computer program product of claim 16, further comprising, generating a document including the first and second scaled images, wherein the document comprises a bank statement and wherein the first and second scanned images comprise first and second check copies.

18. The computer program product of claim 10, further comprising computer code means for padding the image prior to scaling by appending a perimeter of blank pixels to the image.

19. The computer program product of claim 18, further comprising, computer code means for performing multiple iterations of the image scaling of claim 18, wherein the degree of padding varies during each iteration to produce a set of scaled images, and computer code means for selecting a single scaled image from among the set of scaled images.

20. An image scaling system, comprising:

a capture device to scan an original and to produce a bi-tonal image therefrom; and

a scaling engine to scale the bi-tonal image using independent x and y scale factors to produce a scaled image, wherein isolated pixels in the bi-tonal image are preserved in the scaled image.

21. The system of claim 20, wherein the scaling engine produces the scaled image using fixed point arithmetic calculations exclusively.

22. The system of claim 20, wherein scaling engine pads the image prior to scaling by appending a perimeter of blank pixels around the image.