System and method for authentication of a workpiece using three dimensional shape recovery

• US 7,251,347 B2
• Filed: 04/09/2003
• Issued: 07/31/2007
• Est. Priority Date: 04/09/2002
• Status: Active Grant

• Current Assignee: The Escher Group, Ltd.
• Original Assignee: The Escher Group, Ltd.
• Inventors: Smith, Joshua R.
• Primary Examiner(s): Johns; Andrew W.
• Application Number:
  US10/410,487
• Publication Number:
  US 20030219145A1
• Time in Patent Office: 1,574 Days
• Field of Search: 382/100, 382/108, 382/135, 382/154, 356/71, 340/5.8, 340/5.86, 283/72, 283/74, 283/79, 283/81
• US Class Current: 382/108
• CPC Class Codes: G06V 10/10 Image acquisition document image scanning and transmission H04N1/00; control of digital cameras H04N23/60
• CPC Class Codes: G06V 10/225 based on a marking or identifier characterising the area
• CPC Class Codes: G06V 10/245 by locating a pattern; Special marks for positioning
• CPC Class Codes: G06V 10/60 relating to illumination properties, e.g. using a reflectance or lighting model
• CPC Class Codes: G06V 20/80 Recognising image objects characterised by unique random patterns
• CPC Class Codes: G06V 2201/12 Acquisition of 3D measurements of objects
• CPC Class Codes: G07B 2017/00443 Verification of mailpieces, e.g. by checking databases
• CPC Class Codes: G07D 7/004 using digital security elements, e.g. information coded on a magnetic thread or strip
• CPC Class Codes: G07D 7/0043 using barcodes
• CPC Class Codes: G07D 7/12 Visible light, infrared or ultraviolet radiation

Abstract

A workpiece authentication system uses shape recovery techniques to extract explicit three dimensional (“3-D”) features of the surface geometry of the designated portion of a workpiece from images produced using different lighting conditions. The system then bases authentication on the 3-D surface features. The system recovers surface normals, or equivalently gradients, for selected locations within a designated portion of the workpiece from multiple enrollment images produced under different illumination conditions. The system then encodes the surface normal information into authentication indicia that is placed on the workpiece and/or stores the surface normals or related information. Thereafter, the system determines that a given workpiece is authentic if the surface normals recovered from various verification images correspond to the stored surface normal information or the surface normal information encoded into the indicia. Alternatively, the system may use the surface normals to predict what an image should contain when the workpiece is subjected to a particular lighting condition. The system then determines that the workpiece is authentic if the predicted image and the image produced using the workpiece correspond. The system may instead encode brightness patterns associated with one or more enrollment images into the indicia. The system then recovers surface normals from images produced during verification operations, predicts what the brightness image should contain and compares the enrollment image to the prediction.

Claims

1. Apparatus for use in generating a string for use in determining whether a workpiece is valid, comprising:

an electromagnetic radiation source for illuminating at least one portion of the workpiece with electromagnetic radiation from one or more illumination positions relative to said at least one portion;

an imaging device for generating one or more images of optical characteristics of the at least one portion when the at least one portion is illuminated with the radiation by the source from respective illumination positions; and

a string generator that generates the string based upon, at least in part, recovered shapes of 3-D features of the illuminated surface of the at least one portion, the generator recovering the shapes from the one or more images.

2. The apparatus according to claim 1 wherein the string generator recovers the shapes as surface normals or gradients.

3. Apparatus according to claim 2, wherein the one or more illumination positions include one or more orientations relative to the at least one portion of the workpiece.

4. Apparatus according to claim 2, wherein a portion of the radiation is reflected from the at least one portion at an angle that is normal to a surface of the at least one portion, and the one or more images are generated from said portion of the radiation.

5. Apparatus according to claim 1, further comprising a mechanism that marks the workpiece with a set of one or more markings corresponding to at least one of a signature, the string, and an encrypted representation of the string, the signature being based upon the string and an encrypted representation of the string and a cryptographic key of a certifying authority (CA) or an assigned codeword.

6. Apparatus according to claim 5, wherein said one or more markings comprise at least one of the following on said workpiece:

a barcode, a sequence of numeric or alphanumeric characters, a spread-spectrum marking, and angular symbology.

7. Apparatus according to claim 5, wherein said workpiece comprises a postal mailpiece, a security label, a certificate of authentication, a credit/debit card, an identity card, and a product.

8. Apparatus according to claim 5, wherein said workpiece comprises a postal mailpiece, and said string is also representative of a postage value associated with said mailpiece, if said workpiece is valid.

9. Apparatus according to claim 8, wherein said one or more markings are imprinted on said mailpiece, and said string is also identifies the apparatus, if said workpiece is valid.

10. Apparatus according to claim 9, wherein said string is based upon respective numerical values representative of:

one or more hash values representative of said appearance, said postage value, and an identification value identifying said apparatus.

11. Apparatus according to claim 1, wherein the at least one portion comprises a plurality of portions of the workpiece.

12. Apparatus according to claim 1, wherein said string is based upon, at least in part, a concatenation of a plurality of numerical hash values derived from said one or more images.

13. Apparatus according to claim 1, wherein the string is based upon, at least in part, a numerical hash value derived from surface normals values that correspond to one or more images, the hash value being generated by a process that includes extracting from the one or more images one or more image portions, scaling the image portions to generate scaled image portions, determining surface normals values for selected pixels, averaging the surface normal values to generate filtered surface normal values, and subtracting the filtered surface normal values for corresponding pixels.

14. Apparatus according to claim 1, wherein the string uniquely identifies the workpiece.

15. Apparatus of claim 1 wherein the workpiece is a security label or certificate of authenticity that is associated within a given product or document and the string is further based, at least in part, on information that identifies the given product or document.

16. Apparatus according to claim 1, wherein the string is based upon, at least in part, a numerical hash value derived from one or more images, the value being generated by a process that includes determining surface normal values for workpiece locations that correspond to pixels in the one or more images, and filtering the surface normal values to produce filtered values.

17. Apparatus according to claim 16, wherein the process further includes scaling the surface normal values before the values are filtered.

18. Method for generating a string for use in determining whether a workpiece is valid, comprising:

illuminating at least one portion of the workpiece with electromagnetic radiation from one or more illumination positions relative to the at least one portion;

generating one or more images of optical characteristics of 3-D features of a surface of the at least one portion when the at least one portion is illuminated with the radiation at the one or more illumination positions;

recovering the shapes of one or more of the 3-D features from the one or more images; and

generating the string based upon, at least in part, the recovered one or more shapes.

19. Method according to claim 18 wherein the step of recovering the one or more shapes includes recovering surface normals or gradients.

20. Method according to claim 19 wherein the step of recovering the one or more shapes further includes recovering the surface normals or gradients based on directional derivatives along one or more coordinate axes in the workpiece coordinate frame.

21. Method according to claim 20 wherein the step of recovering the surface normals includes determining in the workpiece coordinate frame illumination direction vectors that correspond to the illumination positions.

22. Method according to claim 18, further comprising marking the workpiece with a set of one or more markings corresponding to at least one of a signature, the string, and an encrypted representation of the string, the signature being based upon the string and a cryptographic key of a certifying authority or an assigned codeword.

23. Method according to claim 22, wherein said workpiece comprises a postal mailpiece, a security label, a certificate of authentication, a credit/debit card, an identity card, and a product that includes the optically active medium as a coating or layer.

24. Method according to claim 22, wherein said workpiece comprises a postal mailpiece, and said string is also representative of a postage value associated with said mailpiece, if said workpiece is valid.

25. Method according to claim 24, wherein said one or more markings are imprinted on said mailpiece by an apparatus, and said string also identifies the apparatus, if said workpiece is valid.

26. Method according to claim 25, wherein said string is based upon respective numerical values representative of:

one or more hash values representative of said appearance, said postage value, and an identification value identifying said apparatus.

27. Method according to claim 18, wherein the one or more illumination positions include one or more orientations relative to the at least one portion of the workpiece.

28. Method according to claim 27, wherein the at least one portion comprises a plurality of portions of the workpiece.

29. Method according to claim 18, wherein a portion of the radiation is reflected from the surfaces of the 3-D features of the at least one portion at respective angles that are normal to the surface, and the one or more images are generated from said portion of the radiation.

30. Method according to claim 18, wherein said string is based upon, at least in part, a concatenation of a plurality of numerical hash values derived from said one or more images.

31. Method according to claim 18, wherein the sting is based upon, at least in part, a numerical hash value derived from surface normals values that correspond to one or more images, the hash value being generated by a process that includes extracting from the one or more images one or more image portions, scaling the image portions to generate scaled image portions, determining surface normals values for selected pixels, averaging the surface normal values to generate filtered surface normal values, and subtracting the filtered surface normal values for corresponding pixels.

32. Method according to claim 18, wherein the string uniquely identifies the workpiece.

33. The method of claim 18 wherein the workpiece is a security label or certificate of authenticity that is associated within a given product or document and the string is further based, at least in part, on information that identifies the given product or document.

34. Method according to claim 18, wherein the string is based upon, at least in part, a numerical hash value derived from one or more images, the value being generated by a process that includes determining surface normal values for workpiece locations that correspond to pixels in the one or more images, and filtering the surface normal values to produce filtered values.

35. Method according to claim 34, wherein the process further includes scaling the surface normal values before the values are filtered.

36. Computer-readable memory comprising computer-executable program instructions for use in generating a string for use in determining whether a workpiece is valid, the instructions when executed causing:

illumination of at least one portion of the workpiece with electromagnetic radiation from one or more illumination positions relative to said at least one portion;

generation of one or more images of optical characteristics of 3-D features of a surface of the at least one portion when the at least one portion is illuminated with the radiation at the one or more illumination positions;

recovery of the shapes of one or more of the 3-D features from the one or more images; and

generation of the string based upon, at least in part, the one or more shapes recovered from the one or more images.

37. Memory according to claim 36, wherein the instructions when executed also cause marking of the workpiece with a set of one or more markings corresponding to at least one of a signature, the string, and an encrypted representation of the string, the signature being based upon the string and a cryptographic key of a certifying authority or an assigned codeword.

38. Memory according to claim 37, wherein said one or more markings comprise at least one of the following on said workpiece:

a barcode, a sequence of numeric or alphanumeric characters, a spread-spectrum marking, and angular symbology.

39. Memory according to claim 37, wherein said workpiece comprises a postal mailpiece, and said string is also representative of a postage value associated with said mailpiece, if said workpiece is valid.

40. Memory according to claim 39, wherein said one or more markings are imprinted on said mailpiece by an apparatus, and said string also identifies the apparatus, if said workpiece is valid.

41. Memory according to claim 40, wherein said string is based upon respective numerical values representative of:

one or more hash values representative of said appearance, said postage value, and an identification value identifying said apparatus.

42. Memory according to claim 36 wherein the recovery of the one or more shapes includes the recovery of surface normals or gradients.

43. Memory according to claim 36, wherein said workpiece comprises a postal mailpiece, a security label, a certificate of authentication, a credit/debit card, an identity card, and a product.

44. Memory according to claim 36, wherein the one or more illumination positions are one or more orientations relative to the at least one portion of the workpiece.

45. Memory according to claim 36, wherein a portion of the radiation is reflected from the respective 3-D features of the surface of the at least one portion angles that are normal to the features and the one or more images are generated from said portion of the radiation.

46. Memory according to claim 36, wherein the at least one portion comprises a plurality of portions of the workpiece.

47. Memory according to claim 36, wherein said string is based upon, at least in part, a concatenation of a plurality of numerical hash values derived from said images.

48. Memory according to claim 36, wherein said string is based upon, at least in part, differences between the images.

49. Memory according to claim 36, wherein the sting is based upon, at least in part, a numerical hash value derived from surface normals values that correspond to one or more images, the hash value being generated by a process that includes extracting from the one or more images one or more image portions, scaling the image portions to generate scaled image portions, determining surface normals values for selected pixels, averaging the surface normal values to generate filtered surface normal values, and subtracting the filtered surface normal values for corresponding pixels.

50. Memory according to claim 36, wherein the string uniquely identifies the workpiece.

51. The memory of claim 36 wherein the workpiece is a security label or certificate of authenticity that is associated within a given product or document and the string is further based, at least in part, on information that identifies the given product or document.

52. Memory according to claim 36, wherein the string is based upon, at least in part, a numerical hash value derived from one or more images, the value being generated by a process that includes determining surface normal values for workpiece locations that correspond to pixels in the one or more images, and filtering the surface normal values to produce filtered values.

53. Memory according to claim 52, wherein the process further includes scaling the surface normal values before the values are filtered.

54. A method for authenticating a workpiece

illuminating at least one portion of the workpiece from a plurality of illumination positions;

producing one or more images of the optical characteristics of 3-D features of the surface of the at least one portion;

recovering from the one or more images the respective shapes of one or more of the 3-D features;

generating a string based upon, at least in part, the shapes recovered from the one or more images; and

determining if the string corresponds to a previously generated string that is associated with the workpiece.

55. A method for authenticating a workpiece

illuminating at least one portion of the workpiece from a plurality of illumination positions and producing one or more image of the optical characteristics of 3-D features of the surface of the at least one portion;

recovering from the one or more images the respective shapes of one or more of the 3-D features;

predicting an image produced under a different illumination condition based on the recovered shapes;

determining that the workpiece is authentic if the predicted image corresponds to an image produced during workpiece enrollment using the different illumination condition.

56. Method for generating a string for use in determining whether a workpiece is valid

illuminating at least one portion of the workpiece with electromagnetic radiation from a plurality of illumination positions relative to said at least one portion;

generating one or more images of optical characteristics of 3-D features of the at least one portion when the at least one portion is illuminated with the radiation at the respective illumination positions;

recovering from the one or more images surface normals or gradients that correspond to selected ones of the 3-D features; and

generating a string based upon, at least in part, the surface normals and information that identifies one or both of the workpiece and an associated product or document.

57. Method according to claim 56 further including a step of authenticating the workpiece, the step including

predicting an image that corresponds to a given illumination condition based on the surface normals or gradients,

illuminating at least one portion of the workpiece in accordance with the given illumination condition to produce an authentication image;

and

determining that the workpiece is authentic if the predicted image corresponds to the authentication image.

58. Method for authenticating a workpiece

illuminating at least one portion of the workpiece with electromagnetic radiation from a plurality of illumination positions relative to said at least one portion;

generating one or more images of optical characteristics of 3-D features of the at least one portion when the at least one portion is illuminated with the radiation at the respective illumination positions;

recovering from the one or more images surface normals or gradients that correspond to selected ones of the 3-D features;

generating a string based upon, at least in part, the surface normals and information that identifies one or both of the workpiece and an associated product or document; and

determining that the workpiece is authentic if the generated string corresponds to a string associated with the workpiece.

59. Method according to claim 58 further including the step of repeating the steps of generating the string and comparing the string to the string associated with the workpiece when the one or more images are made with the workpiece being in a predetermined relative position when the workpiece has not been determined to be authentic.

60. Method for authenticating a workpiece

illuminating at least one portion of the workpiece with electromagnetic radiation from a plurality of illumination positions relative to said at least one portion;

generating one or more images of optical characteristics of 3-D features of the at least one portion when the at least one portion is illuminated with the radiation at the respective illumination positions;

recovering from the one or more images surface normals or gradients that correspond to selected ones of the 3-D features;

predicting an image that corresponds to a predetermined illumination condition based on the recovered surface normals or gradients; and

determining if the predicted image corresponds to an image associated with the workpiece.

61. Method for identifying a workpiece

illuminating at least one portion of the workpiece with electromagnetic radiation from a plurality of illumination positions relative to said at least one portion;

generating one or more images of optical characteristics of 3-D features of the at least one portion when the at least one portion is illuminated with the radiation at the respective illumination positions;

recovering from the one or more images surface normals or gradients that correspond to selected ones of the 3-D features;

generating a string based upon, at least in part, the surface normals; and

identifying the workpiece based on a correspondence between the generated string and a stored string that is associated with the workpiece.

62. Method for authenticating a workpiece

illuminating at least one portion of the workpiece with electromagnetic radiation from a plurality of illumination positions relative to said at least one portion;

generating one or more images of optical characteristics of 3-D features of the at least one portion when the at least one portion is illuminated with the radiation at the respective illumination positions;

recovering from the one or more images surface normals or gradients that correspond to selected ones of the 3-D features;

generating a string based upon, at least in part, the surface normals and information that identifies one or both of the workpiece and an associated product or document;

determining that the workpiece is authentic if the generated string corresponds to a string associated with the workpiece;

if the generated string does not correspond to the string associated with the workpiece

determining if the workpiece is within a predetermined range of position and orientation to an enrollment orientation associated with the workpiece;

if the workpiece is not generating a new string with the workpiece is within the predetermined range and if the new string corresponds to the string associated with the workpiece determining that the workpiece is authentic; and

if the generated string and the string associated with the workpiece do not correspond when the workpiece is within the predetermined range, determining that the workpiece is counterfeit.