Method for calibrating a depth camera

1. A method for calibrating a depth camera comprising:

S1) acquiring, using a depth camera to be calibrated, a plurality of depth maps of a physical three-dimensional reference object, corresponding to different positions of the depth camera relative to the physical three-dimensional reference object;

S2) computing a three-dimensional reconstruction of an outer surface of the reference object from said depth maps, depending on values assigned to a set of intrinsic calibration parameters of the depth camera;

S3) computing an error between the three-dimensional reconstruction computed at step S2) and a predetermined digital model of the outer surface of the reference object;

iterating said steps S2) and S3) by assigning different values to said set of intrinsic calibration parameters of the depth camera; and

S4) choosing values of said set of intrinsic calibration parameters which minimize the error computed at step S3), wherein

said steps S2) to S4) are performed by a computer.

2. The method of claim 1 , wherein said steps S2) and S3) are iterated a predetermined number of times by assigning respective predetermined values to said set of intrinsic calibration parameters of the depth camera, said predetermined values sampling a parameter space.

3. The method of claim 1 , wherein said steps S2) and S3) are iterated until the error computed at step S3) becomes lower, or lower of equal, to a predetermined threshold.

4. The method of claim 1 , wherein said set of intrinsic calibration parameters comprises at least a focal length and two coordinates of a principal point of the depth camera.

5. The method of claim 4 , wherein said set of intrinsic calibration parameters also comprises at least one distortion parameter.

6. The method of claim 1 , wherein said step S2) comprises jointly determining the position of the depth camera, relative to a reference position, corresponding to each of said depth maps, and fusing said depth maps into a single three-dimensional signed and truncated distance field; and extracting an isosurface of said distance field into a three-dimensional mesh constituting said reconstruction of the reference object.

7. The method of claim 1 , wherein said step S3) comprises registering the three-dimensional reconstruction computed at step S2) and said predetermined digital model of the reference object prior to computing said error.

8. The method of claim 1 , wherein said step S3) comprises computing an exact or approximate Hausdorff distance or symmetrical Hausdorff distance between the three-dimensional reconstruction computed at step S2) and said predetermined digital model of the reference object.

9. A computer program product, stored on a non-transitory computer-readable data-storage medium, comprising computer-executable instructions to cause a computer system to:

compute a three-dimensional reconstruction of an outer surface of a physical three-dimensional reference object from a plurality of depth maps of said physical three-dimensional reference object, acquired using a depth camera to be calibrated and corresponding to different positions of the depth camera relative to the reference object;

compute an error between the computed three-dimensional reconstruction and a predetermined digital model of the outer surface of the reference object;

iterate the previous operations by assigning different values to a set of intrinsic calibration parameters of the depth camera; and

choose values of said set of intrinsic calibration parameters which minimize the computed error.

10. A non-transitory computer-readable data-storage medium containing computer-executable instructions to cause a computer system to:

compute a three-dimensional reconstruction of an outer surface of a physical three-dimensional reference object from a plurality of depth maps of said physical three-dimensional reference object, acquired using a depth camera to be calibrated and corresponding to different positions of the depth camera relative to the reference object;

compute an error between the computed three-dimensional reconstruction and a predetermined digital model of the outer surface of the reference object;

iterate the previous operations by assigning different values to a set of intrinsic calibration parameters of the depth camera; and

choose values of said set of intrinsic calibration parameters which minimize the computed error.

11. A computer system comprising:

a processor coupled to a memory and a graphical user interface, the memory storing computer-executable instructions to cause the processor to be configured to:

compute a three-dimensional reconstruction of an outer surface of a physical three-dimensional reference object from a plurality of depth maps of said physical three-dimensional reference object, acquired using a depth camera to be calibrated and corresponding to different positions of the depth camera relative to the reference object;

compute an error between the computed three-dimensional reconstruction and a predetermined digital model of the outer surface of the reference object;

iterate the previous operations by assigning different values to a set of intrinsic calibration parameters of the depth camera; and

choose values of said set of intrinsic calibration parameters which minimize the computed error.