Image pre-processor for character image recognition system

• US 5,818,978 A
• Filed: 01/27/1997
• Issued: 10/06/1998
• Est. Priority Date: 04/15/1994
• Status: Expired due to Term

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Al-Hussein, Hussein
• Primary Examiner(s): Couso, Jose L.
• Application Number:
  US08/791,458
• Time in Patent Office: 617 Days
• Field of Search: 382/289, 382/296, 382/306
• US Class Current: 382/296
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/1478 of characters or characters lines
• CPC Class Codes: G06V 30/148 Segmentation of character regions
• CPC Class Codes: G06V 30/155 Removing patterns interfering with the pattern to be recognised, such as ruled lines or underlines
• CPC Class Codes: G06V 30/162 Quantising the image signal
• CPC Class Codes: G06V 30/413 Classification of content, e.g. text, photographs or tables

Abstract

A personal imaging computer system, which is connectable to and operable with a computerized local or wide area network, identifies characters in a document on which the characters are formed. The system scans the document to obtain a gray-scale image of the document, de-skews the gray-scale image, generates a binary image from the de-skewed gray-scale image by comparing the gray-scale image with the threshold, segments the binary image to locate individual characters within the binary image and to determine the shape of the individual characters, extracts gray-scale image information from the gray-scale image for each such individual character based on the location and shape of the character in the binary image, recognition-processes the extracted gray scale image information to determine the identity of the character, and stores the identity of the character. Image pre-processing for the personal imaging computer system includes pre-processing for de-skewing the image, for obtaining and applying a global threshold which converts a gray-scale image to a binary image, for removing underlines from underlined characters in the image, for obtaining connected components within a binary image, and for applying plural sets of rules to the connected components so as to filter text-type connected components from non-text type connected components, whereby only text-type connected components are subjected to character recognition processing.

Claims

1. An image processing apparatus for de-skewing an image of a document, the image comprised by pixels arranged into rows and columns, said apparatus comprising:

measuring means for measuring skew angle of the image;

rotation-transforming means for rotation-transforming pixels of the image by a mathematical rotation transformation so as to reduce the skew angle to approximately zero;

pixel-shifting means for pixel-shifting columns of the image so as to reduce the skew angle to approximately zero; and

determination means for causing said rotation-transforming means to reduce skew angle in a case where the measured skew angle is greater than a predetermined limit, and for causing said pixel-shifting means to reduce skew angle in a case where the measured skew angle is less than the predetermined limit.

2. An image processing apparatus according to claim 1, wherein said measuring means performs a rough measurement to obtain a rough determination of skew angle, and a fine measurement to obtain a fine measure of skew angle in a neighborhood around the rough skew angle.

3. An image processing apparatus according to claim 1, wherein said measuring means sub-samples the image, binarizes the image, performs rough Hough transformation to obtain a rough skew angle and performs fine Hough transformation in a neighborhood around the rough skew angle to determine the skew angle.

4. An image processing apparatus according to claim 1, wherein said measuring means omits measuring skew angle of pictures.

5. An image processing apparatus according to claim 1, wherein said rotation-transforming means is comprised by matrix multiplication.

6. An image processing apparatus according to claim 1, wherein in said pixel shifting means pixel-shifts columns of the image in blank areas between characters in the image.

7. An image processing apparatus according to claim 6, wherein said pixel shifting means includes determining means for determining a vertical shift factor, and wherein the vertical shift factor is accumulated until a blank area between characters is reached.

8. An image processing apparatus according to claim 1, wherein said pixel shifting means includes determining means for determining a vertical shift factor by which all columns of the image are successively shifted.

33. An image processing apparatus according to claim 1, wherein the mathematical transformation is such that each pixel in the reduced skew image is a mathematical combination of several pixels in the skewed image.

34. An image process apparatus according to claim 1 or 33, wherein pixel-shifting is such that each pixel in the reduced skew image has a value found identically in a single pixel of the skewed image.

9. An image processing apparatus for de-skewing an image of a document, the image comprised by pixels arranged into rows and columns, said apparatus comprising:

a memory for storing the image and for storing instruction sequences by which the image is processed; and

a processor for executing the stored instruction sequences;

wherein the stored instruction sequences include process steps adapted to cause said processor to (a) measure skew angle of the image, (b) rotation-transform pixels of the image by applying a mathematical rotation transformation so as to reduce the skew angle to approximately zero in a case where the measured skew angle is greater than a predetermined limit, and (c) pixel-shift columns of the image so as to reduce the skew angle to approximately zero in a case where the measured skew angle is less than the predetermined limit.

10. An apparatus according to claim 9, wherein said stored instruction sequences include process steps adapted to cause said processor to perform a rough measurement to obtain a rough determination of skew angle, and a fine measurement to obtain a fine measure of skew angle in a neighborhood around the rough skew angle.

11. An apparatus according to claim 9, wherein said stored instruction sequences include process steps adapted to cause said processor to sub-sample the image, binarize the image, perform rough Hough transformation to obtain a rough skew angle and perform fine Hough transformation in a neighborhood around the rough skew angle to determine the skew angle.

12. An apparatus according to claim 9, wherein said stored instruction sequences include process steps adapted to cause said processor to omit measurement of skew angle of pictures.

13. An apparatus according to claim 9, wherein said rotation-transform is comprised by matrix multiplication.

14. An apparatus according to claim 9, wherein said stored instruction sequences include process steps adapted to cause said processor to pixel-shift columns of the image in blank areas between characters in the image.

15. An apparatus according to claim 14, wherein said stored instruction sequences include process steps adapted to cause said processor to determine a vertical shift factor, and wherein the vertical shift factor is accumulated until a blank area between characters is reached.

16. An apparatus according to claim 9, wherein said stored instruction sequences include process steps adapted to cause said processor to determine a vertical shift factor by which all columns of the image are successively shifted.

35. An image processing apparatus according to claim 9, wherein the mathematical transformation is such that each pixel in the reduced skew image is a mathematical combination of several pixels in the skewed image.

36. An image process apparatus according to claim 9 or 35, wherein pixel-shifting is such that each pixel in the reduced skew image has a value found identically in a single pixel of the skewed image.

17. An image processing method for de-skewing an image of a document, the image comprised by pixels arranged into rows and columns, said method comprising the steps of:

measuring skew angle of the image;

rotation-transforming pixels of the image by applying a mathematical rotation transformation so as to reduce the skew angle to approximately zero in a case where the measured skew angle is greater than a predetermined limit; and

pixel-shifting columns of the image so as to reduce the skew angle in a case where the measured skew angle is less than the predetermined limit.

18. A method according to claim 17, wherein said measuring step includes a rough measuring step for obtaining a rough determination of skew angle, and a fine measuring step for obtaining a fine measure of skew angle in a neighborhood around the rough skew angle.

19. A method according to claim 17, wherein said measuring step includes the steps of sub-sampling the image, binarizing the image, performing rough Hough transformation to obtain a rough skew angle and performing fine Hough transformation in a neighborhood around the rough skew angle to determine the skew angle.

20. A method according to claim 17, wherein said measuring step omits measuring skew angle of pictures.

21. A method according to claim 17, wherein said rotation-transforming step is comprised by matrix multiplication.

22. A method according to claim 17, wherein in said pixel shifting step, columns of the image are pixel-shifted in blank areas between characters in the image.

23. A method according to claim 22, wherein said pixel shifting step includes a step of determining a vertical shift factor, and wherein the vertical shift factor is accumulated until a blank area between characters is reached.

24. A method according to claim 17, wherein said pixel shifting step includes the step of determining a vertical shift factor by which all columns of the image are successively shifted.

37. An image processing method according to claim 17, wherein the mathematical transformation is such that each pixel in the reduced skew image is a mathematical combination of several pixels in the skewed image.

38. An image process method according to claim 17 or 37, wherein pixel-shifting is such that each pixel in the reduced skew image has a value found identically in a single pixel of the skewed image.

25. Computer-executable process steps stored on a computer readable medium, the computer-executable process steps to de-skew an image of a document which contains character, the image comprised by pixels arranged into rows and columns, said process steps including:

a measuring step to measure skew angle of the image;

a comparing step to compare the skew angle to a predetermined limit;

a rotation step to rotation-transform pixels of the image by applying a mathematical rotation transformation so as to reduce the skew angle to approximately zero in a case where the skew angle is greater than the predetermined limit; and

a shifting step to pixel-shift columns of the image so as to reduce the skew angle to approximately zero in a case where the skew angle is less than a predetermined limit.

26. Computer-executable process steps according to claim 25, wherein said measuring step includes a rough measuring step which obtains a rough determination of skew angle, and a fine measuring step which obtains a fine measure of skew angle in a neighborhood around the rough skew angle.

27. Computer-executable process steps according to claim 25, wherein said measuring step includes a sub-sampling step which sub-samples the image, a binarizing step which binarizes the image, a first performing step which performs rough Hough transformation to obtain a rough skew angle and a second performing step which performs fine Hough transformation in a neighborhood around the rough skew angle to determine the skew angle.

28. Computer-executable process steps according to claim 25, wherein said measuring step omits measuring skew angle of pictures.

29. Computer-executable process steps according to claim 25, wherein said rotation-transforming step is comprised by matrix multiplication.

30. Computer-executable process steps according to claim 25, wherein in said pixel shifting step, columns of the image are pixel-shifted in blank areas between characters in the image.

31. Computer-executable process steps according to claim 30, wherein said pixel shifting step includes a determining step which determines a vertical shift factor, and wherein the vertical shift factor is accumulated until a blank area between characters is reached.

32. Computer-executable process steps according to claim 25, wherein said pixel shifting step includes a determining step which determines a vertical shift factor by which all columns of the image are successively shifted.

39. Computer-executable process steps according to claim 25, wherein the mathematical transformation is such that each pixel in the reduced skew image is a mathematical combination of several pixels in the skewed image.

40. Computer-executable process steps according to claim 25 or 39, wherein pixel-shifting is such that each pixel in the reduced skew image has a value found identically in a single pixel of the skewed image.