Correction of distortions in form processing

• US 6,658,166 B1
• Filed: 03/08/2000
• Issued: 12/02/2003
• Est. Priority Date: 03/08/2000
• Status: Expired due to Term

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Krutyolkin, Sergey, Zlotnick, Aviad, Shevach, Ronen
• Primary Examiner(s): Boudreau, Leo
• Assistant Examiner(s): Dang, Duy M.
• Application Number:
  US09/521,399
• Time in Patent Office: 1,364 Days
• Field of Search: 382/276, 382/277, 382/280-283, 382/293, 382/295, 382/296, 382/173, 382/198, 382/209, 358/452-453, 707/500
• US Class Current: 382/282
• CPC Class Codes: G06T 2207/30176 Document
• CPC Class Codes: G06T 3/00 Geometric image transformations in the plane of the image
• CPC Class Codes: G06T 5/80 Geometric correction
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/1478 of characters or characters lines

Abstract

A method for processing an input image includes applying one or more transformations to the input image, whereby different shifts are applied to different pixels in the input image, so as to generate an output image. A region containing content of interest is selected in the output image, and the region corresponding to the selected region in the output image is located in the input image. The content of the located region in the input image is substituted for the content of the selected region in the output image.

Claims

1. A method for processing an input image, comprising:

applying one or more transformations to the input image, whereby different shifts are applied to different pixels in the input image, so as to generate an output image;

selecting in the output image a region containing content of interest;

locating in the input image the region corresponding to the selected region in the output image; and

substituting the content of the located region in the input image for the content of the selected region in the output image.

2. A method according to claim 1, wherein applying the one or more transformations comprises dividing the input image into segments, and determining a transformation to be applied to each segment.

3. A method according to claim 2, wherein the input image includes a template delineating the region, which is filled in with the content of interest, and wherein determining the transformation to be applied to each segment comprises finding one or more translations of the segment that approximately compensate for a distortion of the input image relative to a reference template, whereby the output image is registered with the reference template.

4. A method according to claim 3, wherein the one or more translations compensate for a rotation of the input image relative to the reference template.

5. A method according to claim 2, wherein applying the one or more transformations comprises applying different transformations to two or more mutually-adjoining segments.

6. A method according to claim 5, wherein selecting the region comprises selecting a region that was divided between the two or more segments to which different transformations were applied.

7. A method according to claim 1, wherein the input image includes a template delineating the region, which is filled in with the content of interest, and wherein selecting the region comprises identifying a field of the template that is intended to receive the content of interest.

8. A method according to claim 1, wherein the input image includes a template delineating the region, which is filled in with the content of interest, and wherein selecting the region comprises removing the template from the output image and selecting a portion of the image remaining after the template is removed.

9. A method according to claim 1, wherein selecting the region comprises selecting a region responsive to the one or more transformations applied to the input image.

10. A method according to claim 9, wherein the content of interest comprises alphanumeric characters, and wherein selecting the region comprises selecting a region in which it is likely that the one or more transformations have adversely affected the readability of the characters in the region.

11. A method according to claim 1, wherein locating the region comprises finding the region of the input image that was transformed into the selected region of the output image by the one or more transformations.

12. A method according to claim 1, wherein substituting the content of the located region comprises finding connected components in the located region and copying the connected components to the selected region in the output image.

13. A method according to claim 12, wherein copying the connected components comprises finding, for each of the connected components in the located region, a translation operation to be applied to all of the points in the connected component.

14. A method according to claim 13, wherein finding the translation operation comprises, for each of the connected components, choosing a point on or in a vicinity of the connected component and determining a translation that was applied to that point by the one or more transformations applied to the input image.

15. A method according to claim 12, wherein finding the connected components comprises finding characters in the image.

16. A method according to claim 1, wherein the content of interest comprises alphanumeric characters, and wherein the method comprises applying optical character recognition to the substituted content in the selected region.

17. Apparatus for processing an input image of a document including a template having one or more regions that are filled in with content, the apparatus comprising a form processor, which is adapted to apply one or more transformations to the input image so as to generate an output image in which the template is registered with a reference template, to select at least one of the filled-in regions in the output image, to locate in the input image at least one filled-in region corresponding to the at least one selected region in the output image, and to substitute the content of the at least one located region in the input image for the content of the at least one selected region in the output image.

18. Apparatus according to claim 17, wherein the content filled into the one or more region comprises alphanumeric characters, and wherein the processor applies optical character recognition to the substituted content in the at least one selected region so as to extract the content from the document.

19. Apparatus according to claim 18, and comprising an imaging device, which is adapted to scan the document so as to generate the input image.

20. A computer software product for processing an input image, the product comprising a computer-readable medium in which program instructions are stored, which instructions, when read by a computer, cause the computer to apply one or more transformations to the input image, whereby different shifts are applied to different pixels in the input image, so as to generate an output image, to select in the output image a region containing content of interest, to locate in the input image the region corresponding to the selected region in the output image, and to substitute the content of the located region in the input image for the content of the selected region in the output image.

21. A product according to claim 20, wherein the instructions cause the computer to divide the input image into segments, and to determine a transformation to be applied to each segment.

22. A product according to claim 20, wherein the content of interest comprises alphanumeric characters, and wherein the instructions cause the computer to select a region for substitution of the content thereof when it is likely that the one or more transformations have adversely affected the readability of the characters in the selected region.

23. A product according to claim 20, wherein the instructions cause the computer to find connected components in the located region and to copy the connected components to the at least one selected region in the output image.

24. A product according to claim 23, wherein the instructions cause the computer to find, for each of the connected components in the located region, a translation operation to be applied to all of the points in the connected component.

25. A product according to claim 20, wherein the content of interest comprises alphanumeric characters, and wherein the instructions further cause the computer to apply optical character recognition to the substituted content in the selected region.

26. A product according to claim 20, wherein the input image includes a template delineating the region, which is filled in with the content of interest, and wherein the instructions cause the computer to apply the one or more transformations so as to register the output image with a reference template.