Geometric parameter measurement of an imaging device

• US 8,818,130 B2
• Filed: 11/21/2008
• Issued: 08/26/2014
• Est. Priority Date: 12/21/2007
• Status: Active Grant

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Morgan-Mar, David, Hardy, Stephen James, Larkin, Kieran Gerard, Deller, Christine Anne, Arnison, Matthew R
• Primary Examiner(s): Chu, Randolph I
• Assistant Examiner(s): BLOOM, NATHAN J
• Application Number:
  US12/275,963
• Publication Number:
  US 20090161945A1
• Time in Patent Office: 2,104 Days
• Field of Search: 382/100, 382/103, 382/106, 382/141, 382153-154, 382/181, 382/190, 382/195, 382/199, 382/201, 382203-206, 382/209, 382217-218, 382/276, 348/42, 348 86- 95, 348135-141, 348/143, 348175-176, 348187-188
• US Class Current: 382/276
• CPC Class Codes: G06T 5/80 Geometric correction
• CPC Class Codes: G06V 10/245 by locating a pattern; Special marks for positioning
• CPC Class Codes: H04N 1/00827 Arrangements for reading an image from an unusual original, e.g. 3-dimensional objects
• CPC Class Codes: H04N 17/002 for television cameras

Abstract

Disclosed is a method of determining at least one three-dimensional (3D) geometric parameter of an imaging device. A two-dimensional (2D) target image is provided having a plurality of alignment patterns. The target image is imaged with an imaging device to form a captured image. At least one pattern of the captured image is compared with a corresponding pattern of the target image. From the comparison, the geometric parameter of the imaging device is then determined. The alignment patterns include at least one of (i) one or more patterns comprising a 2D scale and rotation invariant basis function, (ii) one or more patterns comprising a 1D scale invariant basis function, and (iii) one or more patterns having a plurality of grey levels and comprising a plurality of superimposed sinusoidal patterns, the plurality of sinusoidal patterns having a plurality of predetermined discrete orientations. Also disclosed is a two-dimensional test chart for use in testing an imaging device, the test chart comprising a plurality of alignment patterns, at least one of said alignment patterns including one of those patterns mentioned above.

Claims

1. A method of determining three-dimensional (3D) geometric parameters of an imaging device, the method comprising the steps of:

(a) providing a two-dimensional (2D) target image comprising a plurality of alignment patterns;

(b) imaging said target image with an imaging device to form a captured image;

(c) comparing at least one said pattern of said captured image with a corresponding pattern of said target image; and

(d) determining from said comparison said geometric parameters of the imaging device,

wherein the geometric parameters comprise at least one geometric parameter describing a geometric lens distortion transformation between said target image and said captured image and at least one geometric parameter describing a perspective transformation between said target image and said captured image,

wherein the geometric parameter describing the perspective transformation is determined based on said geometric parameter describing the geometric lens distortion transformation.

2. A method according to claim 1, where said alignment patterns are represented as binary images using a half-toning algorithm.

3. A method according to claim 1, where said geometric parameters further comprise a parameter describing a translation of said target image from a reference position.

4. A method according to claim 1, where said geometric parameters further comprise a parameter describing a translation of said target image from a reference position.

5. A method according to claim 1, where said geometric parameters further comprise a parameter describing a scaling of said target image from a reference size.

6. A method according to claim 1, where said geometric parameters further comprise a parameter describing the focal length of the lens system of the said imaging device.

7. A method according to claim 1, where said geometric parameters further comprise a parameter describing the location of a principal point of the lens system of the said imaging device.

8. A method according to claim 1, further comprising the step of aligning said target image to said imaging device so that distortion of said captured image compared to said target image caused by perspective transformation is small when compared to the distortion of the said captured image compared to said target image caused by geometric lens distortion of the imaging device.

9. A method according to claim 1, where said geometric parameter describing the geometric lens distortion transformation between said target image and said captured image and said geometric parameter describing the perspective transformation between said target image and said captured image are determined by simultaneous fitting to a composite transformation function characterised by said plurality of geometric parameters.

10. A method according to claim 1, where said at least one geometric parameter that describes the perspective transformation from said target image to said captured image, and said at least one geometric parameter that describes the geometric lens distortion transformation from said target image to said captured image, are derived by the steps of:

(A) deriving said parameters characterising said geometric lens distortion by said comparison of at least one pattern of said captured image with a corresponding pattern of said target image;

(B) transforming said captured image according to an inverse of said geometric lens distortion parameters to produce a modified captured image; and

(C) deriving said parameters characterising the perspective transformation by comparing at least one pattern of said modified captured image with a corresponding pattern of said target image.

11. A method according to claim 10, wherein step (B) comprises modifying coordinates of said captured image according to said derived parameters.

12. A method according to claim 1, where said target image is an image fabricated on to a substantially planar surface material.

13. A method according to claim 1, where said target image is an image displayed on a display device.

14. A method according to claim 1, where said target image is an image displayed on a projection surface by an image projector.

15. A method according to claim 1, where said imaging device is a document scanner and said geometric parameters relate to the image-forming geometry of said scanner.

16. A method according to claim 1, where said imaging device is a camera and said geometric parameters relate to the image-forming geometry of said camera.

17. A method according to claim 1, where said geometric parameters are used to determine the alignment of an imaging plane within said imaging device, relative to a reference plane.

18. A method according to claim 17, where said imaging plane is coincident with an imaging sensor of said imaging device.

19. A method according to claim 17, where said imaging plane is coincident with a manual focusing screen of said imaging device.

20. A method according to claim 17, where said imaging plane is coincident with an autofocussing sensor of said imaging device.

21. A method according to claim 1, wherein step (c) comprises correlating a pattern of said captured image with a pattern of said target image.

22. A method according to claim 1, wherein at least one said alignment pattern of said target image comprises a composite of at least two of the patterns (i) and (ii).

23. A method according to claim 1, wherein said alignment patterns include one or more patterns comprising a 2D scale and rotation invariant basis function

24. A method according to claim 1, wherein said alignment patterns include one or more patterns having a plurality of grey levels and comprise a plurality of superimposed sinusoidal patterns, said plurality of sinusoidal patterns arranged in a plurality of predetermined discrete orientations.