Method and apparatus for identifying an imaging device

• US 7,616,237 B2
• Filed: 05/19/2006
• Issued: 11/10/2009
• Est. Priority Date: 12/16/2005
• Status: Active Grant

• Current Assignee: The Research Foundation for The State University of New York (State University of New York)
• Original Assignee: The Research Foundation for The State University of New York (State University of New York)
• Inventors: Lukas, Jan, Fridrich, Jessica, Goljan, Miroslav
• Primary Examiner(s): HO, TUAN V
• Application Number:
  US11/437,176
• Publication Number:
  US 20090154809A1
• Time in Patent Office: 1,271 Days
• Field of Search: 348/241, 348/239, 348/207.99, 348/243, 348/231.3
• US Class Current: 348/241
• CPC Class Codes: G06T 7/44 using image operators, e.g. filters, edge density metrics or local histograms
• CPC Class Codes: G06V 10/30 Noise filtering
• CPC Class Codes: G06V 20/80 Recognising image objects characterised by unique random patterns
• CPC Class Codes: G06V 20/90 Identifying an image sensor based on its output data

Abstract

A new technique for identifying whether images are derived from a common imager, e.g., a camera, or other imaging device such as a scanner, based on the device'"'"'s measured or inferred reference noise pattern, a unique stochastic characteristic of all common digital imaging sensors, including CCD, CMOS (Foveon™ X3), and JFET. The measured or inferred noise pattern of two images may be extracted and then cross correlated, with a high correlation being consistent with a common imager. Various preprocessing techniques may be used to improve tolerance to various types of image transform. It is also possible to perform the analysis without explicit separation of inferred image and inferred noise.

Claims

1. A non-intrusive method for determining whether digital images are derived from the same electronic imager, comprising the steps of analyzing at least two-electronic images to determine whether an intrinsic imager noise pattern represented therein is consistent with an intrinsic imager noise pattern attributable to a common imager for the at least two digital images, and outputting an indicia of that determined consistency.

2. The method of claim 1, wherein said analyzing step infers a photoresponse non-uniformity noise of the electronic imager.

3. The method of claim 1, wherein said analyzing step infers a dark frame noise of the electronic imager.

4. The method of claim 1, wherein said analyzing step infers a white frame noise of the electronic imager.

5. The method of claim 1, wherein consistency is determined by correlating parameters of at least two images presumed to be related to the intrinsic imager noise pattern.

6. The method of claim 1, wherein said analyzing step comprises performing at least one transformation selected from the group consisting of a translation, a rotation, an isotropic scaling, an anisotropic scaling, a color scale transformation, and a brightness transformation.

7. The method of claim 1, wherein said analyzing step is tolerant to a processing of at least one digital image selected from the group consisting of irreversible image compression, spatial frequency-selective filtering, Gaussian blurring, subsampling, colorspace conversion, superimposition of noise from a source other than the electronic imager, and subtotal digital noise subtraction.

8. The method according to claim 1, wherein said analyzing step comprises extracting an intrinsic imager noise from a plurality of electronic images derived from one digital imager.

9. The method according to claim 1, further comprising the step of separating an inferred noise pattern from an inferred image pattern of the digital image.

10. The method according to claim 1, wherein said analyzing step comprises correlating a pattern of the digital images.

20. A computer readable medium containing a program for controlling a general purpose processor to perform the method of claim 1.

22. The method according to claim 1, wherein the at least two digital images are obtained under partially different illumination conditions of the electronic imager.

11. Apparatus for determining whether a plurality of different digital images are derived from the same electronic imager, comprising an input for receiving the plurality of different digital images to be analyzed, a processor for analyzing the plurality of different digital images to determine whether an intrinsic imager noise pattern represented therein is consistent with an intrinsic imager noise pattern attributable to a common imager, and an output for presenting an indication of that determined consistency.

12. The apparatus according to claim 11, wherein the processor processes the digital images to infer a photoresponse non-uniformity noise of the electronic imager.

13. The apparatus according to claim 11, wherein said processor processes the digital images to infer a dark frame noise of the electronic imager.

14. The apparatus according to claim 11, wherein said processor processes the digital images to infer a white frame noise of the electronic imager.

15. The apparatus according to claim 11, wherein said processor correlates parameters presumed to be related to the intrinsic imager noise pattern of at least two images as an indication of consistency of common digital imager.

16. The apparatus according to claim 11, wherein said processor processes at least one digital image using at least one transformation selected from the group consisting of a translation, a rotation, an isotropic scaling, an anisotropic scaling, a color scale transformation, and a brightness transformation.

17. The apparatus according to claim 11, wherein said processor employs an algorithm tolerant to at least one image transformation selected from the group consisting of irreversible image compression, spatial frequency-selective filtering, Gaussian blurring, subsampling, colorspace conversion, superimposition of noise from a source other than the electronic imager, and subtotal digital noise subtraction.

18. The apparatus according to claim 11, wherein said processor extracts an inferred noise from a plurality of electronic images derived from a common digital imager.

19. The apparatus according to claim 11, wherein the processor separates an inferred noise pattern from an inferred image pattern, and compares the inferred noise pattern with a second inferred noise pattern or measured intrinsic imager noise pattern to determine a consistency with a common digital imager.

21. The apparatus according to claim 11, wherein the at least two digital images are obtained under at least partially different illumination conditions of the electronic imager.

23. A method for determining whether a plurality of different digital images are derived from the same electronic imager, comprising analyzing a plurality of different digital images to determine whether an intrinsic imager noise pattern comprising imager noise under varying states of illumination represented therein is consistent with an intrinsic imager noise pattern attributable to the same electronic imager, and outputting an indicia of that determined consistency.