Method and arrangement for multi-camera calibration

• US 9,303,525 B2
• Filed: 03/08/2011
• Issued: 04/05/2016
• Est. Priority Date: 03/26/2010
• Status: Active Grant

• Current Assignee: RPX Corporation
• Original Assignee: Alcatel-Lucent SA (Nokia Corporation)
• Inventors: Aerts, Maarten, Tytgat, Donny, Macq, Jean-Francois, Lievens, Sammy
• Primary Examiner(s): Harold, Jefferey
• Assistant Examiner(s): Haiem, Sean
• Application Number:
  US13/582,164
• Publication Number:
  US 20130070108A1
• Time in Patent Office: 1,855 Days
• US Class Current: 1/1
• CPC Class Codes: F01D 21/003 Arrangements for testing or measuring for measuring vibrations G01H
• CPC Class Codes: F01D 25/00 Component parts, details, or accessories, not provided for in, or of interest apart from, other groups
• CPC Class Codes: F05D 2260/80 Diagnostics
• CPC Class Codes: G01N 2021/151 Gas blown
• CPC Class Codes: G01N 21/15 Preventing contamination of the components of the optical system or obstruction of the light path
• CPC Class Codes: G01N 21/85 Investigating moving fluids or granular solids
• CPC Class Codes: G01N 21/954 Inspecting the inner surface of hollow bodies, e.g. bores
• CPC Class Codes: G01N 2201/08 Optical fibres; light guides
• CPC Class Codes: G02B 23/2492 Arrangements for use in a hostile environment, e.g. a very hot, cold or radioactive environment
• CPC Class Codes: G06T 2207/10012 Stereo images
• CPC Class Codes: G06T 7/85 Stereo camera calibration

Abstract

A method for determining calibration data for at least two cameras (camera1, camera2) in a multi view position, includes a step of determining respective parameters ((h¹₀₀, . . . , h¹₂₂), (h²₀₀, . . . , h²₂₂)) for identifying at least one respective homographic transformation on respective images (image1,image2) taken by said cameras of a same scene, by performing respective geometry analyses on said respective images (image1, image2), a step of performing at least one respective combined homographic transformation/feature detection step on said respective images thereby obtaining respective sets (feature set1, feature set2) of features on respective transformed images, such that said calibration data are obtained from matches (m_l, . . . , m_k) determined between said respective sets of features.

Claims

1. A method for determining calibration data for at least two cameras (camera1, camera2) in a multi view position, said method including the steps of:

determining respective parameters ((h¹₀₀ . . . h¹₂₂ )), (h²₀₀, . . . h²₂₂)) for identifying at least one respective homographic transformation on respective images (image1,image2) taken by said cameras of a same scene, by performing respective geometry analyses on said respective images (image1, image2); and

performing at least one respective homographic transformation on said respective images and a respective feature detection step on respective transformed images thereby obtaining respective sets (feature set1, feature set2) of features on said respective transformed images, such that said calibration data are obtained from matches (ml,..,mk) determined between said respective sets of features.

2. A method according to claim 1 wherein said calibration data comprise information with respect to the relative position between said cameras.

3. A method according to claim 1 wherein said calibration data comprise information with respect to the relative viewing direction differences between said cameras.

4. A method according to claim 1 wherein said at least one respective homographic transformation further comprises an iterative process of pixel coordinates transformation of said respective image, followed by feature detection of the partially transformed image until all pixel coordinates of said respective image (imagel) are transformed.

5. A method according to claim 1 wherein said respective geometry analysis steps comprise determining line segments on said respective images.

6. A method according to claim 1 wherein said respective geometry analysis steps comprise determining curved segments on said respective image.

7. A method according to claim 1 wherein said respective geometry analysis steps comprise performing object recognition on said respective images.

8. A method according to claim 5 wherein said respective line segments are further used for determining a viewpoint angle difference between said at least two cameras.

9. A method according to claim 1 wherein said calibration data are further determined from said respective sets of features.

10. A method according to claim 1 wherein said calibration data are further determined from said parameters for identifying said respective homographic transformations.

11. A method according to claim 1, wherein said calibration data further comprise internal camera parameters which values can be further optimized based on said matches.

12. A device for determining calibration data for at least two cameras (camera1, camera2) in a multi view position, said device adapted to perform the steps of:

determining respective parameters ((h¹₀₀ . . . h¹₂₂), (h²₀₀, . . . h²₂₂)) for identifying at least one respective homographic transformation on respective images (image1,image2) taken by said cameras of a same scene, by performing respective geometry analyses on said respective images (image1, image2) ; and

performing at least one respective homographic transformation on said respective images and a respective feature detection step on respective transformed images thereby obtaining respective sets (feature set1, feature set2) of features on said respective transformed images, such that said calibration data are obtained from matches (ml,..,mk) determined between said respective sets of features.

14. A device according to claim 12 wherein said calibration data comprise information with respect to the relative position between said cameras.

15. A device according to claim 12 wherein said calibration data comprise information with respect to the relative viewing direction differences between said cameras.

16. A device according to claim 12 wherein said at least one respective homographic transformation further comprises an iterative process of pixel coordinates transformation of said respective image, followed by feature detection of the partially transformed image until all pixel coordinates of said respective image (image1) are transformed.

17. A device according to claim 12 wherein said respective geometry analysis steps comprise determining line segments on said respective images.

18. A device according to claim 12 wherein said respective geometry analysis steps comprise determining curved segments on said respective image.

19. A device according to claim 12 wherein said respective geometry analysis steps comprise performing object recognition on said respective images.

20. A device according to claim 12 wherein said calibration data are further determined from said parameters for identifying said respective homographic transformations.

13. A computer program product comprising a non-transitory computer usable medium having computer usable program code stored therein, said computer program product when executed performs the following process:

determining respective parameters ((h¹₀₀ . . . h¹₂₂) (h²₀₀, . . . h²₂₂)) for identifying at least one respective homographic transformation on respective images (image1,image2) taken by said cameras of a same scene, by performing respective geometry analyses on said respective images (image1, image2) ; and

performing at least one respective homographic transformation on said respective images and a respective feature detection step on respective transformed images thereby obtaining respective sets (feature set1, feature set2) of features on said respective transformed images, such that said calibration data are obtained from matches (ml, . . . ,mk) determined between said respective sets of features.