Sequential product code quantization of digital color image

• US 5,544,284 A
• Filed: 02/11/1992
• Issued: 08/06/1996
• Est. Priority Date: 02/11/1992
• Status: Expired due to Term

• Current Assignee: Eastman Kodak Company
• Original Assignee: Eastman Kodak Company
• Inventors: Bouman, Charles A., Allebach, Jan P., Balasubramanian, Thyagarajan
• Primary Examiner(s): Nguyen, Phu K.
• Application Number:
  US07/833,529
• Time in Patent Office: 1,638 Days
• Field of Search: 395/131, 395/128, 395/127, 364/200 MS File, 364/900 MS File, 382/164, 382/169, 382/234, 382/235, 382/255, 382/258
• US Class Current: 345/603
• CPC Class Codes: H04N 1/644 using a reduced set of representative colours, e.g. each representing a particular range in a colour space

Abstract

A digital color image quantization mechanism employs sequential product code vector quantization, to sequentially extract chrominance and luminance values from the vectors and quantizes chrominance and luminance features based upon a conditional distribution of these features within partitioned regions of chrominance/luminance color space. The mechanism sequentially partitions a histogram of the original digital color image in luminance, chrominance (Y,Cb,Cr) space coordinates into a plurality of sub-regions or color space cells, such that each partitioned color cell is associated with a color of the output palette through which the color composition of a reproduced color image is defined. A splitting criterion determines the sequential order of partitioning of an axis. Because of the increased sensitivity of the human visual system to contouring artifacts in regions of an image to low spatial activity, the splitting criterion along the luminance axis is scaled or weighted in inverse proportion to the average spatial activity of the luminance-chrominance region subject to be split. A map of chrominance and luminance output codes is generated for the respective pixels of the output color image in accordance with the axial splitting or quantization of the chrominance and luminance components of the histogram.

Claims

1. A method of converting a first digital color image, having a color composition encoded into first luminance representative codes and first chrominance representative codes of a first encoded color resolution, to a second digital color image having a color composition encoded at a second encoded color resolution different from said first encoded color resolution and color codes of which are selected to reproduce a color image having a quality that is pleasing to a human visual system, said method comprising the steps of:

(a) subdividing said first digital color image into a plurality of spatially adjacent groups of pixels;

(b) for each group of the pixels subdivided in step (a), deriving luminance activity values from the first luminance representative codes associated therewith;

(c) generating a color space histogram of said first luminance representative codes and said first chrominance representative codes;

(d) associating the luminance activity values derived in step (b) with the color space histogram generated in step (c);

(e) partitioning said color space histogram, along an axis associated with a chrominance component of said first digital color image, into histogram chrominance regions, based upon a chrominance composition of said color space histogram;

(f) partitioning said color space histogram chrominance regions, along an axis associated with a luminance component of said first digital color image, into histogram luminance-chrominance regions, based upon a distribution of the luminance activity values within said color spaced histogram chrominance regions;

(g) or each of said color space histogram chrominance and luminance regions, deriving a respective set of chrominance and luminance output codes based upon a distribution of the chrominance composition and luminance activity within said color space histogram chrominance and luminance regions, respectively; and

(h) defining the second color composition of a respective pixel of said second digital color image in accordance with one of the respective sets of chrominance and luminance output codes derived in step (g).

2. A method according to claim 1, wherein step (h) comprises generating a map of chrominance and luminance output codes for the respective pixels of said second digital color image in accordance with a quantization of the chrominance component of said first digital color image into chrominance values associated with said color space histogram chrominance regions.

3. A method according to claim 2, wherein step (h) comprises generating said map of chrominance and luminance output codes in accordance with the luminance representative codes through which the color composition of each of the pixels of said first digital color image is encoded.

4. A method according to claim 1, further comprising the step of

(i) transforming said map of chrominance and luminance output codes of said second digital color image into sets of output color codes for application to a color image reproduction device through which a color output image is provided.

5. A method of quantizing a color composition of a digital color image, comprising the steps of:

(a) generating a color space histogram of the color composition of pixels of said digital color image in terms of prescribed color coordinates through which the color composition of said digital image is definable;

(b) performing sequential product code partitioning of said color space histogram along respective chrominance and luminance coordinate axes thereof, so as to subdivide said color space histogram into a plurality of color space regions each comprising respective color space contents; and

(c) for each of the color space regions into which said color space histogram has been subdivided in step (b), deriving a respective output code representative of the respective color space contents of the plurality of color space regions.

6. A method according to claim 5, further including the step of (d) associating respective ones of the output codes derived in step (c) with pixels of an output image, which correspond spatially to the pixels of said digital color image.

7. A method according to claim 6, further including the step of (e) transforming the respective codes of said output image into output color codes for application to a color image reproduction device through which a color output image is provided.

8. A method according to claim 5, wherein at least one of the prescribed color coordinates of the color space histogram generated in step (a) is prequantized to a digital resolution less than the digital resolution to which said at least one of the prescribed color coordinates of said digital color image is encoded.

9. A method according to claim 8, wherein the prescribed color coordinates of said digital color image correspond to respective chrominance and luminance coordinates of the color space into which said digital color image has been encoded.

10. A method according to claim 9, wherein said at least one prescribed color coordinate of the color space histogram generated in step (a) includes a chrominance coordinate.

11. A method according to claim 10, wherein the prescribed color coordinates of said digital color image correspond to a luminance and a pair of chrominance coordinates of the color space into which said digital color image has been encoded, and wherein said at least one prescribed color coordinate of the color space histogram generated in step (a) includes each of said chrominance coordinates.

12. A method according to claim 5, wherein the prescribed color coordinates of said digital color image correspond to a luminance coordinate and a pair of chrominance coordinates of the color space into which said digital color image has been encoded, and wherein each of the chrominance coordinates of the color space histogram generated in step (a) is prequantized to a digital resolution less than the digital resolution to which said digital color image is encoded, and said luminance coordinate of the color space histogram generated in step (a) has the same digital resolution as said digital color image.

13. A method according to claim 5, wherein the prescribed color coordinates of the color space of said digital color image correspond to a luminance coordinate and a pair of chrominance coordinates, and wherein said step (b) of performing sequential product code partitioning os said color space histogram along the respective luminance and chrominance coordinate axes thereof comprises sequentially splitting said chrominance coordinates and thereafter splitting said luminance coordinate.

14. A method according to claim 13, wherein step (b) comprises sequentially partitioning said luminance coordinate using a splitting criterion that depends upon spatial activity characteristics of said digital color image.

15. A method according to claim 5, wherein the coordinates of the color space of said digital color image correspond to a luminance coordinate and a pair of chrominance coordinates, and wherein said step (b) of performing sequential product code partitioning of said color space histogram comprises performing sequential binary splitting of the respective luminance and chrominance coordinate axes in accordance with a splitting criterion that depends upon the variance of the contents of said color space histogram for that portion of a respective axis to be split.

16. A method according to claim 5, including the preliminary step of subsampling said digital color image to provide a sub-array of digital color pixels and steps (a)-(c) are carried out with respect to the subsampled digital color image.

17. A method according to claim 5, wherein step (c) comprises, for each of the coordinates of said color space histogram, generating a respective look-up table that associates the digitized color composition for that respective coordinate of said digital color image with a quantized color space region.

18. A method according to claim 5, wherein step (a) comprises generating said color space histogram in the form of one-dimensional color space histograms of the color space activity of said digital color image, the composition of each one-dimensional color space histogram being referenced to a respective one of the color coordinates through which the color composition of said digital image is definable.

19. A method according to claim 18, wherein step (b) comprises performing sequential product code partitioning of said one-dimensional color space histograms along their respectively associated luminance and chrominance coordinate axes, so as to subdivide said color space histograms into a plurality of color space regions.

20. A method according to claim 5, wherein step (a) comprises generating a two-dimensional array of color space coordinate data entries, respective ones of which are associated with a range of color coordinate values of a further color space coordinate.

21. A method of generating a plurality of color codes, respective ones of which represent a color composition of a digital color image comprising an array of pixels, the color compositions of which are digitized to a prescribed color code resolution, said method comprising the steps of:

(a) generating a color space histogram of the color composition of the pixels of said digital color image in terms of prescribed color coordinates through which the color composition of said digital image is definable;

(b) quantizing the color composition of the color space histogram by performing sequential product code partitioning of said color space histogram along respective luminance and chrominance coordinate axes thereof into a plurality of color cells; and

(c) for each of the color cells into which said color space histogram has been partitioned in step (b), generating a respective color code representative of the color space contents of that cell.

22. A method according to claim 21, wherein step (c) comprises generating a respective color code representative of the centroid of the color space contents of that cell.

23. A method according to claim 21, wherein step (c) comprises, for each of the coordinates of said color space histogram, generating a respective look-up table that associates the digitized color composition for that respective coordinate of said digital color image with a quantized coordinate value thereof.

24. A method according to claim 23, wherein step (c) comprises cascading the look-up tables together, and generating therefrom, in response to the color composition of a pixel of said digitized color image, a respective color code which is representative of the color composition of said digital color image.

25. A method according to claim 21, further including the step of (d) associating respective ones of the color codes generated in step (c) with pixels of an output image, which correspond spatially to the pixels of said digital color image.

26. A method according to claim 25, further including the step of (e) transforming the respective color codes of said output image into output color codes for application to a color image reproduction device through which a color output image is provided.

27. A method according to claim 21, wherein at least one of the coordinate axes of the color space histogram generated in step (a) is prequantized to a digital resolution less than the digital resolution to which said at least one of the color coordinates of said digital color image is encoded.

28. A method according to claim 27, wherein the coordinates of said digital color image correspond to respective chrominance and luminance coordinates of the color space into which said digital color image has been encoded.

29. A method according to claim 28, wherein the coordinates of said digital color image correspond to a luminance coordinate and a pair of chrominance coordinates of the color space into which said digital color image has been encoded, and wherein each of the chrominance coordinates of the color space histogram generated in step (a) is prequantized to a digital resolution less than the digital resolution to which said digital color image is encoded, and said luminance coordinate of the color space histogram generated in step (a) has the same digital resolution as said digital color image.

30. A method according to claim 21, wherein the coordinates of the color space of said digital color image correspond to a luminance coordinate and a pair of chrominance coordinates, and wherein said step (b) quantizing the color composition of the color space histogram by performing sequential product code partitioning of said color space histogram along the respective luminance and chrominance coordinate axes thereof comprises performing sequential binary splitting in accordance with a splitting criterion that depends upon the variance of the contents of said color space histogram for that portion of a respective axis to be split.

31. A method according to claim 30, wherein step (b) comprises sequentially partitioning said chrominance coordinates, and thereafter splitting said luminance coordinate.

32. A method according to claim 31, wherein step (b) comprises sequentially partitioning said luminance coordinate using a splitting criterion that depends upon spatial activity characteristics of said digital color image.

33. A method according to claim 21, wherein said step (b) of performing sequential product code partitioning of said color space histogram along respective coordinate axes thereof comprises performing sequential binary splitting of said axes along respective planes perpendicular to the respective axes being split.

34. A method according to claim 21, including the preliminary step of subsampling said digital color image for provide a sub-array of digital color pixels and steps (a)-(c) are carried out with respect to the subsampled digital color image.

35. A method according to claim 21, wherein step (a) comprises generating said color space histogram in the form of one-dimensional color space histograms of the color space activity of said digital color image, the composition of each one-dimensional color space histogram being referenced to a respective one of the color coordinates through which the color composition of said digital image is definable.

36. A method according to claim 35, wherein step (b) comprises performing sequential product code partitioning of said one-dimensional color space histograms along their respectively associated coordinate axes, so as to subdivide said one-dimensional color space histograms into a plurality of color cells.

37. A method according to claim 21, wherein step (a) comprises generating a two-dimensional array of color space coordinate data entries, respective ones of which are associated with a range of color coordinate values of a further color space coordinate.

38. An apparatus for quantizing a data vector structure having a multi-dimensional coordinate system, comprising a plurality of sequentially coupled look-up tables such that an output of an ith look-up table is a pointer to an entry in an (i+l)th look-up table, the respective look-up tables of said plurality being associated with the respective axes of said multi-dimensional coordinate system of said data vector structure, and wherein the output of the last look-up table in sequence is a multi-bit code representative of a quantized data structure value.

39. An apparatus according to claim 38, wherein said data vector structure is representative of a digital color image color space.

40. An apparatus according to claim 38, wherein said data vector structure is representative of respective luminance and chrominance coordinates of a digital color image color space.

41. An apparatus according to claim 38, wherein said data vector structure is representative of respective luminance and chrominance coordinates of a digital color image color space, and wherein an ith one of said look-up tables is associated with a chrominance coordinate of said digital color image color space and wherein an (i+j)th one of said look-up tables is associated with a luminance coordinate of said digital color image color space, and wherein the output of the last look-up table in sequence is a multi-bit code representative of a color of said digital image color space.

42. An apparatus according to claim 38, wherein said vector structure is representative of a digital color image color space, and wherein each look-up table is coupled to receive, as part of its entry pointer, a code representative of a prescribed quantization of a coordinate of said digital color image color space.

43. An apparatus according to claim 42, wherein said prescribed quantization of a coordinate of said digital color image color space corresponds to a sequential product code partitioning of said color space along that coordinate'"'"'s axis into a plurality of color space regions.

44. An apparatus according to claim 38, wherein each look-up table is coupled to receive, as part of the entry pointer, a code representative of a prescribed quantization of a coordinate of said data vector structure, and wherein said prescribed quantization of a coordinate of said data vector structure corresponds to a sequential product code partitioning of said data vector structure along that coordinate'"'"'s axis into a plurality of data vector structure regions.

45. A method of quantizing a data vector structure having a multi-dimensional coordinate system, comprising the steps of:

(a) providing a plurality of look-up tables, a respective look-up table of said plurality being associated with a respective axis of said multi-dimensional coordinate system of said data vector structure; and

(b) sequentially coupling said look-up tables together such that an output of an ith look-up table is a pointer to an entry in an (i+l)th look-up table, whereby output of the last look-up table in sequence is a multi-bit code representative of a quantized data structure value.

46. A method apparatus according to claim 45, wherein each look-up table is coupled to receive, as part of its entry pointer, a code representative of a prescribed quantization of a coordinate of said data vector structure, and wherein said prescribed quantization of a coordinate of said data vector structure corresponds to a sequential product code partitioning of said data vector structure along that coordinate'"'"'s axis into a plurality of data vector structure regions.

47. A method according to claim 45, wherein said data vector structure is representative of a digital color image color space.

48. A method according to claim 45, wherein said data vector structure is representative of respective luminance and chrominance coordinates of a digital color image color space.

49. A method according to claim 45, wherein said data vector structure is representative of respective luminance and chrominance coordinates of a digital color image color space, and wherein an ith one of said look-up tables is associated with a chrominance coordinate of said digital color image color space and wherein an (i+j)th one of said look-up tables is associated with a luminance coordinate of said digital color image color space, and wherein the output of the last look-up table in sequence is a multi-bit code representative of a color of said digital image color space.

50. A method according to claim 45, wherein said vector structure is representative of a digital color image color space, and wherein each look-up table is coupled to receive, as part of the entry pointer, a code representative of a prescribed quantization of a coordinate of said digital color image color space.

51. For use with a digital color image comprised of an array of pixels the color compositions of which are digitized to a prescribed color code resolution, a method of quantizing a multi-dimensional color space structure through which the color compositions of the pixels of said array are definable, comprising the steps of:

(a) providing a multi-dimensional data structure representative of the multi-dimensional color space of said digital color image; and

(b) performing sequential product code partitioning of said multi-dimensional data structure along at least one of respective coordinate axes thereof into a plurality of color cells.

52. A method according to claim 51, wherein step (b) comprises performing sequential product code partitioning of said multi-dimensional data structure along each of its respective coordinate axes into a plurality of multi-dimensional color cells.

53. A method according to claim 52, wherein step (a) comprises providing a luminance, chrominance data structure associated with a luminance, chrominance color space representative of said digital color image, and wherein step (b) comprises performing sequential product code partitioning of said luminance, chrominance data structure along each of the respective chrominance and luminance coordinate axes into a plurality of luminance-chrominance cells, each of which is associated with a quantized value of said digital color image.

54. A method according to claim 51, further including the step (c) of, for each of the cells into which said data structure has been partitioned in step (b), generating a respective color code representative of the color space contents of that cell.

55. A method according to claim 54, wherein step (c) comprises generating a respective color code representative of the centroid of the color space contents of that cell.

56. A method according to claim 54, wherein step (c) comprises, for each of the coordinates of said data structure, generating a respective look-up table that associates the digitized color composition for that respective coordinate of said digital color image with its quantized coordinate value.

57. A method according to claim 56, wherein step (c) comprises sequentially coupling the look-up tables together, and generating therefrom, in response to the color composition of a pixel of said digitized color image, a respective color code which is representative of the color composition of said digital color image.

58. A method according to claim 51, wherein at least one of the coordinate axes of the multi-dimensional data structure provided in step (a) is prequantized to a digital precision less than the digital precision to which at least one of the color coordinates of said digital color image is encoded.

59. A method according to claim 58, wherein a prequantized coordinate axis corresponds to a coordinate axis of said multi-dimensional data structure where the sensitivity of the human visual system is reduced in comparison to another coordinate axis of said multi-dimensional data structure.

60. A method according to claim 59, wherein a prequantized coordinate axis corresponds to a chrominance coordinate axis of a luminance-chrominance data structure where the sensitivity of the human visual system is reduced in comparison to a luminance coordinate axis of said data structure.

61. In a method of quantizing the color composition of a digital color image comprised of an array of pixels which are digitized to a prescribed color code precision, the improvement comprising providing a multi-dimensional data structure representative of the multi-dimensional color space of said digital color image, and wherein at least one of the coordinates of said multi-dimensional data structure is prequantized to a digital code precision less than the digital code precision to which at least one of the color coordinates of said digital color image is encoded.

62. The improvement according to claim 61, wherein a prequantized coordinate axis corresponds to a coordinate axis of said multi-dimensional data structure where the sensitivity of the human visual system is reduced in comparison to another coordinate axis of said multi-dimensional data structure.

63. The improvement according to claim 61, wherein a prequantized coordinate axis corresponds to a chrominance coordinate axis of a luminance-chrominance data structure where the sensitivity of the human visual system is reduced in comparison to a luminance coordinate axis of said data structure.

64. For use with a digital color image comprised of an array of pixels the color compositions of which are digitized to a prescribed color code resolution, a method of quantizing a multi-dimensional color space structure through which the color compositions of the pixels of said array are definable, comprising the steps of:

(a) providing a multi-dimensional data structure representative of the multi-dimensional color space of said digital color image; and

(b) partitioning said multi-dimensional data structure along one of its respective coordinate axes thereof into a plurality of color cells using an axis-splitting criterion that depends upon spatial activity characteristics of said digital color image along that axis.

65. A method according to claim 64, wherein step (a) comprises providing a luminance, chrominance data structure associated with a luminance, chrominance color space representative of said digital color image, and wherein step (b) comprises performing sequential product code partitioning of said luminance, chrominance data structure along each of its respective chrominance and luminance coordinate axes into a plurality of luminance-chrominance cells, each of which is associated with a quantized value of said digital color image, and wherein the partitioning of said multi-dimensional data structure along its luminance coordinate axes into a plurality of color cells employs an axis-splitting criterion that depends upon spatial activity characteristics of said digital color image along the luminance axis.

66. A method according to claim 64, wherein the coordinates of the color space of said digital color image correspond to a luminance coordinate and a pair of chrominance coordinates, and wherein step (b) comprises partitioning the corresponding luminance-chrominance data structure by sequential product code partitioning of said data structure along its respective coordinate luminance and chrominance axes, in accordance with a splitting criterion that depends upon the variance of the contents of said data structure for that portion of a respective axis to be split, and such that the partitioning of said luminance-chrominance data structure along its luminance coordinate axis into a plurality of color cells employs an axis-splitting criterion that depends upon spatial activity weighting of the variance of the contents of said data structure along the luminance axis.

67. A method of quantizing a multi-coordinate data vector structure comprising the steps of:

(a) generating a first one-dimensional data vector structure by projecting components of said multi-coordinate data structure onto a first coordinate axis thereof;

(b) iteratively partitioning said first one-dimensional data vector structure into a first plurality of one-dimensional data vector structure regions;

(c) for each of the first plurality of one-dimensional data vector structure regions partitioned in step (b), generating a respective associated second one-dimensional data vector structure by projecting components of said multi-coordinate data structure contained within said first plurality of one-dimensional data vector structure regions onto a second coordinate axis of said multi-dimensional data vector structure; and

(d) iteratively partitioning the second one-dimensional data vector structures generated in step (c) into a second plurality of one-dimensional data vector structure regions.

68. A method according to claim 67, further including the step of (e) generating a plurality of codes respectively associated with each of said second plurality of one-dimensional data vector structure regions iteratively partitioned in step (d).

69. A method according to claim 67, further including the steps of:

(e) for each of the second plurality of one-dimensional data vector structure regions partitioned in step (d), generating a respective associated third one-dimensional data vector structure by projecting components of said multi-coordinate data structure contained within said second plurality of one-dimensional data vector structure regions onto a third coordinate axis of said multi-dimensional data vector structure; and

(f) iteratively partitioning the third one-dimensional data vector structures generated in step (e) into a third plurality of one-dimensional data vector structure regions.

70. A method according to claim 69, further including the step of (g) generating a plurality of codes respectively associated with each of the third plurality of one-dimensional data vector structure regions iteratively partitioned in step (f).

71. A method according to claim 67, wherein said multi-coordinate data vector structure comprises a multi-dimensional color space structure through which the color compositions of the pixels of a digital color image comprised of an array of pixels the color compositions of which are digitized to a prescribed color code resolution.

72. A method according to claim 71, wherein said multi-coordinate data vector structure comprises a three-dimensional luminance-chrominance histogram through which the color compositions of the pixels of a digital color image comprised of an array of pixels the color compositions of which are digitized to a prescribed color code resolution.

73. A method according to claim 72, wherein step (a) comprises generating a first one-dimensional luminance-chrominance histogram by projecting components of said luminance-chrominance histogram onto a first chrominance axis thereof, step (b) comprises iteratively partitioning said first one-dimensional luminance-chrominance histogram into a first plurality of one-dimensional luminance-chrominance regions, step (c) comprises generating a respective associated second one-dimensional luminance-chrominance histogram by projecting components of said luminance-chrominance histogram contained within said first plurality of one-dimensional luminance-chrominance histogram regions onto a second chrominance axis of said three-dimensional luminance-chrominance histogram, for each of the first plurality of one-dimensional luminance-chrominance histogram regions partitioned in step (b), and wherein step (d) comprises iteratively partitioning the second one-dimensional luminance-chrominance histograms generated in step (c) into a second plurality of one-dimensional luminance-chrominance histogram regions.

74. A method according to claim 73, further including the steps of:

(e) for each of the second plurality of one-dimensional luminance-chrominance histogram regions partitioned in step (d), generating a respective associated third one-dimensional luminance-chrominance histogram by projecting components of said luminance-chrominance coordinate data structure contained within said second plurality of one-dimensional luminance-chrominance histogram regions onto the luminance axis of said three-dimensional luminance-chrominance histogram; and

(f) iteratively partitioning the third one-dimensional luminance-chrominance histograms generated in step (e) into a third plurality of one-dimensional luminance-chrominance histogram regions.

75. A method according to claim 74, wherein each of steps (b), (d) and (f) comprises partitioning a corresponding one-dimensional luminance-chrominance histogram by sequential product code partitioning of said one-dimensional luminance-chrominance histogram in accordance with a splitting criterion that depends upon the variance of the contents of said one-dimensional histogram for that portion of a respective axis to be split, and such that the partitioning of said third one-dimensional luminance-chrominance histogram along the luminance coordinate axis into a plurality of color cells employs an axis-splitting criterion that depends upon spatial activity weighting of the variance of the contents of said luminance-chrominance histogram along the luminance axis.

76. A method of quantizing a multi-coordinate data vector structure comprising the steps of:

(a) generating a two-dimensional array of data vector entries, respective ones of which are associated with data structures of a third coordinate of said multi-coordinate data vector structure; and

(b) for a respective data entry of said multi-coordinate data structure accessing said two-dimensional array of data vector entries and locating, along said third coordinate, a corresponding value of said data entry.

77. A method according to claim 76, wherein step (a) comprises storing an entry value of said third coordinate of said data structure as a binary tree structure, and wherein step (b) comprises locating said corresponding value of said data entry by conducting a binary search along said third coordinate of said two-dimensional array.

78. A method according to claim 76, wherein said multi-coordinate data vector structure comprises a digital color image color space, and each entry of said two-dimensional array is a look-up table of color data covering the range of luminance values for a respective chrominance coordinate location of said color space.

79. A method according to claim 76, wherein at least one of the color coordinates of the two-dimensional array is prequantized to a digital precision less than the digital precision to which color coordinates of said digital color image are encoded.

80. A method according to claim 79, wherein step (a) comprises storing an entry value of said third coordinate of said data structure as a binary tree structure, and wherein step (b) comprises locating said corresponding value of said data entry by conducting a binary search along said third coordinate of said two-dimensional array.