Device, method and storage medium for recognizing a document image
First Claim
1. A document image recognizing device, comprising:
- image converting means for converting an input document image into a gray-scale image if the input document image is a color image, and for newly outputting a gray-scale image if the input image is a gray-scale image;
variable resolution binarizing means for converting the input document image into a binary image having a higher resolution according to a resolution of the gray-scale image;
wherein said variable resolution binarizing means performs a sub-pixel generation process for increasing a number of pixels included in an image by interpolating pixel values ofa gray-scale image, sets a local threshold value within a local area centering around a particular pixel, and obtains a binary image by using the local threshold value; and
recognizing means for recognizing the binarized image.
1 Assignment
0 Petitions
Accused Products
Abstract
A color image input from a document image inputting unit is converted into a gray-scale image by a brightness image extracting unit. The gray-scale image is then converted into an image having a higher resolution according to the resolution of the original gray-scale image. When this conversion is performed, subpixels are generated between the original pixels, and the values of the subpixels are obtained with an interpolation method. Furthermore, a threshold value for a binarization process is generated by using an original pixel value and a subpixel value. The characters included in the binarized image are recognized by a binary image recognizing unit, and a recognition result is output from a recognition result outputting unit.
-
Citations
54 Claims
-
1. A document image recognizing device, comprising:
-
image converting means for converting an input document image into a gray-scale image if the input document image is a color image, and for newly outputting a gray-scale image if the input image is a gray-scale image;
variable resolution binarizing means for converting the input document image into a binary image having a higher resolution according to a resolution of the gray-scale image;
wherein said variable resolution binarizing means performs a sub-pixel generation process for increasing a number of pixels included in an image by interpolating pixel values ofa gray-scale image, sets a local threshold value within a local area centering around a particular pixel, and obtains a binary image by using the local threshold value; and
recognizing means for recognizing the binarized image. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18)
said recognizing means recognizes a converted or input binary image, and converts the binary image into electronic codes.
-
-
3. The device according to claim 1, further comprising:
-
drawing area roughly extracting means for roughly extracting a drawing area according to a global threshold value for pixel values of a gray-scale image, wherein said recognizing means recognizes a binary image in an area extracted by said drawing area roughly extracting means.
-
-
4. The device according to claim 3, wherein:
the global threshold value uses a linear combination of an average pixel value, a standard deviation value, and a variance.
-
5. The device according to claim 3, wherein
said variable resolution binarizing means performs the subpixel generation process for an entire input gray-scale image; - and
said drawing area roughly extracting means roughly extracts a drawing area from gray-scale image data for which the subpixel generation process is performed.
- and
-
6. The device according to claim 1, wherein:
the subpixel generation process is performed by linearly interpolating original pixel values of the gray-scale image.
-
7. The device according to claim 1, wherein:
-
said variable resolution binarizing means obtains a binary image by binarizing pixel values by using the local threshold value obtained from a distribution of the pixel values within a local area including pixels; and
the local threshold value uses a linear combination of an average pixel value, a standard deviation value, and a variance.
-
-
8. The device according to claim 1, further comprising:
-
drawing area roughly extracting means for performing a global process which roughly extracts a drawing area according to a global threshold value for pixel values of a gray-scale image, wherein said variable resolution binarizing means performs the subpixel generation process for increasing a number of pixels included in an image by interpolating pixel values of a gray-scale image, for a drawing area which is roughly extracted with the global process, and performs a local binarization process by using the local threshold value for each pixel included in the roughly extracted drawing area.
-
-
9. The device according to claim 1, wherein the local threshold value at a pixel point after the subpixel generation process is based on obtaining a value of a subpixel generated with the subpixel generation process from original pixel values, and obtaining a local threshold value from subpixel values.
-
10. The device according to claim 1, wherein a local threshold value after the subpixel generation process is obtained at a pixel point of a gray-scale image, and a local threshold value of a subpixel is obtained by interpolating local threshold values at pixel points of the gray-scale image.
-
11. The device according to claim 10, wherein the local threshold value at the pixel point of the gray-scale image is obtained by using a value of a subpixel obtained by performing the subpixel generation process for pixel points of the gray-scale image.
-
12. The device according to claim 10, wherein the local threshold value at the pixel point of the gray-scale image after the subpixel generation process is based on obtaining a subpixel value from original pixel values, and obtaining a local threshold value from subpixel values.
-
13. The device according to claim 12, wherein the local threshold value is obtained, using subpixel values, from a local threshold value from a local area that recognizes original pixels as its boundary.
-
14. The device according to claim 10, wherein an interpolation value at a subpixel point is obtained by interpolating a difference between a value at a pixel point of a gray-scale image and a local threshold value after the subpixel generation process, and a value of a binary image at the subpixel point is determined with a sign of the interpolation value.
-
15. The device according to claim 1, wherein a local threshold value at an original pixel point of a gray-scale image after the subpixel generation process is obtained from a linear combination of an average value, a standard deviation value, and a variance, by obtaining the standard deviation value and the variance after obtaining the average value and a square average value of values at pixel points of the gray-scale image, for which the subpixel generation process is performed, based on obtaining a subpixel value from original pixel values, and obtaining the average value and the square average value from subpixel values.
-
16. The device according to claim 1, wherein:
-
the local area recognizes original pixels as its boundary; and
a local threshold value at an original pixel point of a gray-scale image after the subpixel generation process is obtained from a linear combination of an average value, a standard deviation value, and a variance, by obtaining the standard deviation value and the variance after obtaining the average value and a square average value of values at pixel points of the gray-scale image, for which the subpixel generation process is performed, based on obtaining a subpixel value from original pixel values, and obtaining the average value and the square average value from subpixel values.
-
-
17. The device according to claim 1, further comprising:
-
specifying means for specifying a range of a local area, wherein whether or not the local area recognizes original pixels as its boundary is determined by using as specification data a number of subpixels generated with the subpixel generation process and a size of the local area.
-
-
18. The device according to claim 1, wherein:
-
a local threshold value after the subpixel generation process is obtained at a pixel point of a gray-scale image;
a table to which pixel values and local threshold values at four pixel points of a gray-scale are input and from which a binary image enclosed by the four pixel points is output as data, is included; and
a binary image is obtained by using said table.
-
-
19. A document image recognizing method, comprising:
-
converting an input document image into a gray-scale image if the input document image is a color image, and newly outputting a gray-scale image if the input image is a gray-scale image;
converting the input document image into a binary image having a higher resolution according to a resolution of the gray-scale image;
wherein said converting the input document image into a binary image performs a sub-pixel generation process for increasing a number of pixels included in an image by interpolating pixel values of a gray-scale image, sets a local threshold value within a local area centering around a particular pixel, and obtains a binary image by using the local threshold value; and
recognizing the binary image. - View Dependent Claims (20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36)
said recognizing the binary image recognizes a converted or input binary image, and converts the binary image into electronic codes.
-
-
21. The method according to claim 20, further comprising:
-
roughly extracting a drawing area according to a global threshold value for pixel values of a gray-scale image, wherein said recognizing the binary image recognizes a binary image in an area extracted by said roughly extracting a drawing area.
-
-
22. The method according to claim 21, wherein the global threshold value uses a linear combination of an average pixel value, a standard deviation value, and a variance.
-
23. The method according to claim 21, wherein:
-
said converting the input document image into a binary image performs the subpixel generation process for an entire input gray-scale image; and
said roughly extracting a drawing area roughly extracts a drawing area from gray-scale image data for which the subpixel generation process is performed.
-
-
24. The method according to claim 19, wherein:
the subpixel generation process is performed by linearly interpolating original pixel values of the gray-scale image.
-
25. The method according to claim 19, wherein:
-
said converting the input document image into a binary image obtains a binary image by binarizing pixel values by using the local threshold value obtained from a distribution of the pixel values within a local area including pixels; and
the local threshold value uses a linear combination of an average pixel value, a standard deviation value, and a variance.
-
-
26. The method according to claim 19, further comprising:
-
performing a global process which roughly extracts a drawing area according to a global threshold value for pixel values of a gray-scale image, wherein said converting the input document image into a binary image performs the subpixel generation process for increasing a number of pixels included in an image by interpolating pixel values of a gray-scale image, for a drawing area which is roughly extracted with the global process, and performs a local binarization process by using the local threshold value for each pixel included in the roughly extracted drawing area.
-
-
27. The method according to claim 19, wherein the local threshold value at a pixel point after the subpixel generation process is based on obtaining a value of a subpixel generated with the subpixel generation process from original pixel values, and obtaining a local threshold value from subpixel values.
-
28. The method according to claim 19, wherein a local threshold value after the subpixel generation process is obtained at a pixel point of a gray-scale image, and a local threshold value of a subpixel is obtained by interpolating local threshold values at pixel points of the gray-scale image.
-
29. The method according to claim 28, wherein the local threshold value at the pixel point of the gray-scale image after the subpixel generation process is based on obtaining a subpixel value from original pixel values, and obtaining a local threshold value from subpixel values.
-
30. The method according to claim 29, wherein the local threshold value is obtained, using subpixel values, from a local threshold value within a local area that recognizes original pixels as its boundary.
-
31. The method according to claim 28, wherein an interpolation value at a subpixel point is obtained by interpolating a difference between a value at a pixel point of a gray-scale image and a local threshold value after the subpixel generation process, and a value of a binary image at the subpixel point is determined by a sign of the interpolation value.
-
32. The method according to claim 19, wherein the local threshold value at the pixel point of the gray-scale image is obtained by using a value of a subpixel obtained by performing the subpixel generation process for pixel points of the gray-scale image.
-
33. The method according to claim 19, wherein a local threshold value at an original pixel point of a gray-scale image after the subpixel generation process is obtained from a linear combination of an average value, a standard deviation value, and a variance, by obtaining the standard deviation value and the variance after obtaining the average value and a square average value of values at pixel points of the gray-scale image, for which the subpixel generation process is performed, based on obtaining a subpixel value from original pixel values, and obtaining the average value and the square average value from subpixel values.
-
34. The method according to claim 19, wherein:
-
the local area recognizes original pixels as its boundary; and
a local threshold value at an original pixel point of a gray-scale image after the subpixel generation process is obtained from a linear combination of an average value, a standard deviation value, and a variance, by obtaining the standard deviation value and the variance after obtaining the average value and a square average value of values at pixel points of the gray-scale image, for which the subpixel generation process is performed, based on obtaining a subpixel value from original pixel values, and obtaining the average value and the square average value from subpixel values.
-
-
35. The method according to claim 19, further comprising:
specifying a range of a local area, wherein whether or not the local area recognizes original pixels as its boundary is determined by using as specification data a number of subpixels generated with the subpixel generation process and a size of the local area.
-
36. The method according to claim 19, wherein:
-
a local threshold value after the subpixel generation process is obtained at a pixel point of a gray-scale image;
a table to which pixel values and local threshold values at four pixel points of a gray-scale are input and from which a binary image enclosed by the four pixel points is output as data, is included; and
a binary image is obtained by using said table.
-
-
37. A computer-readable storage medium for directing a computer to execute a process comprising:
-
converting an input document image into a gray-scale image if the input document image is a color image, and newly outputting a gray-scale image if the input image is a gray-scale image;
converting the input document image into a binary image having a higher resolution according to a resolution of a gray-scale image;
wherein said converting the input document image into a binary image performs a sub-pixel generation process for increasing a number of pixels included in an image by interpolating pixel values of a gray-scale image, sets a local threshold value within a local area centering around a particular pixel, and obtains a binary image by using the local threshold value; and
recognizing the binary image. - View Dependent Claims (38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54)
said recognizing the binary image recognizes a converted or input binary image, and converts the binary image into electronic codes.
-
-
39. The storage medium according to claim 37, wherein the process further comprises:
-
roughly extracting a drawing area according to a global threshold value for pixel values of a gray-scale image, and wherein said recognizing the binary image recognizes a binary image in an area extracted by said roughly extracting a drawing area.
-
-
40. The storage medium according to claim 39, wherein the global threshold value uses a linear combination of an average pixel value, a standard deviation value, and a variance.
-
41. The storage medium according to claim 39, wherein:
-
said converting the input document image into a binary image performs the subpixel generation process for an entire input gray-scale image; and
said roughly extracting a drawing area roughly extracts a drawing area from gray-scale image data for which the subpixel generation process is performed.
-
-
42. The storage medium according to claim 39, wherein the subpixel generation process is performed by linearly interpolating original pixel values of the gray-scale image.
-
43. The storage medium according to claim 37, wherein:
-
said converting the input document image into a binary image obtains a binary image by binarizing pixel values by using the local threshold value obtained from a distribution of the pixel values within a local area including pixels; and
the local threshold value uses a linear combination of an average pixel value, a standard deviation value, and a variance.
-
-
44. The storage medium according to claim 37, further comprising:
-
performing a global process which roughly extracts a drawing area according to a global threshold value for pixel values of a gray-scale image, and wherein said converting the input document image into a binary image performs the subpixel generation process for increasing a number of pixels included in an image by interpolating pixel values of a gray-scale image, for a drawing area which is roughly extracted with the global process, and performs a local binarization process by using the local threshold value for each pixel included in the roughly extracted drawing area.
-
-
45. The storage medium according to claim 37, wherein the local threshold value at a pixel point after the subpixel generation process is based on obtaining a value of a subpixel generated with the subpixel generation process from original pixel values, and obtaining a local threshold value from subpixel values.
-
46. The storage medium according to claim 37, wherein a local threshold value after the subpixel generation process is obtained at a pixel point of a gray-scale image, and a local threshold value of a subpixel is obtained by interpolating local threshold values at pixel points of the gray-scale image.
-
47. The storage medium according to claim 46, wherein the local threshold value at the pixel point of the gray-scale image is obtained by using a value of a subpixel obtained by performing the subpixel generation process for pixel points of the gray-scale image.
-
48. The storage medium according to claim 46, wherein the local threshold value at the pixel point of the gray-scale image after the subpixel generation process is based on obtaining a subpixel value from original pixel values, and obtaining a local threshold value from subpixel values.
-
49. The storage medium according to claim 48, wherein the local threshold value is obtained, using subpixel values, from a local threshold value within a local area that recognizes original pixels as its boundary.
-
50. The storage medium according to claim 46, wherein an interpolation value at a subpixel point is obtained by interpolating a difference between a value at a pixel point of a gray-scale image and a local threshold value after the subpixel generation process, and a value of a binary image at the subpixel point is determined with a sign of the interpolation value.
-
51. The storage medium according to claim 37, wherein a local threshold value at an original pixel point of a gray-scale image after the subpixel generation process is obtained from a linear combination of an average value, a standard deviation value, and a variance, by obtaining the standard deviation value and the variance after obtaining the average value and a square average value of values at pixel points of the gray-scale image, for which the subpixel generation process is performed, based on obtaining a subpixel value from original pixel values, and obtaining the average value and the square average value from subpixel values.
-
52. The storage medium according to claim 37, wherein:
-
the local area recognizes original pixels as its boundary; and
a local threshold value at an original pixel point of a gray-scale image after the subpixel generation process is obtained from a linear combination of an average value, a standard deviation value, and a variance, by obtaining the standard deviation value and the variance after obtaining the average value and a square average value of values at pixel points of the gray-scale image, for which the subpixel generation process is performed, based on obtaining a subpixel value from original pixel values, and obtaining the average value and the square average value from subpixel values.
-
-
53. The storage medium according to claim 37, further comprising:
-
specifying a range of a local area, and wherein whether or not the local area recognizes original pixels as its boundary is determined by using as specification data a number of subpixels generated with the subpixel generation process and a size of the local area.
-
-
54. The storage medium according to claim 37, wherein:
-
a local threshold value after the subpixel generation process is obtained at a pixel point of a gray-scale image;
a table to which pixel values and local threshold values at four pixel points of a gray-scale are input and from which a binary image enclosed by the four pixel points is output as data, is included; and
a binary image is obtained by using said table.
-
Specification