Calibration of a multi-camera capture system

Abstract

A calibration method is provided for a multi-camera capture system that images a scene for 3D reconstruction of a target display space. The method includes calibrating the cameras in the capture system by iteratively performing the steps of: acquiring a set of images from the cameras, wherein at least two images in the set of images correspond to different cameras; attempting to determine the extrinsic parameters of the cameras based on the set of images; and controlling a control system to adapt the poses of one or more cameras based on the attempt to determine the extrinsic parameters.

Claims

1. A method for calibrating a multi-camera capture system for capturing a scene for 3D reconstruction of a target display space, wherein the target display space defines a portion of the scene to be 3D reconstructed using images from the capture system, and the capture system comprises two or more cameras and a control system for adapting the poses of the cameras, and the method is A step of acquiring a set of images from the camera, wherein at least two images in the set of images correspond to different cameras. The steps include attempting to determine the camera's extrinsic parameters based on the aforementioned set of images, The steps include controlling the control system to mechanically adjust the pose of one or more cameras based on the trial that determines the exogenous parameters, A method including calibrating a camera by repeatedly performing the following.

2. The method according to claim 1, further comprising the step of controlling the control system to mechanically adjust the pose of a camera whose exogenous parameters have been determined based on the target display space.

3. The method according to claim 1 or 2, wherein the step of controlling the control system includes controlling the control system to mechanically adjust the pose of a camera whose extrinsic parameters have not been determined by predefined rotation and / or predefined translation.

4. The step of obtaining a set of images from the cameras includes selecting images obtained from the set of cameras, wherein the set of cameras includes two or more cameras or a subset of two or more cameras, and the subset includes at least two cameras. The aforementioned calibration is When an exogenous parameter is determined for at least one camera in the set of cameras, the control system is controlled to mechanically adjust the pose of one or more of the cameras. The method according to claim 1 or 2, wherein, if no exogenous parameters have been determined for any of the cameras in the set of cameras, one or more cameras are removed from the set of cameras for use in subsequent iterations.

5. The method according to claim 4, further comprising using the same set of cameras in subsequent iterations if an exogenous parameter is not determined for at least one camera in the set of cameras, and if an exogenous parameter is determined for at least one other camera in the set of cameras.

6. The method according to claim 4, further comprising adding one or more cameras to the set of cameras for use in subsequent iterations, once exogenous parameters have been determined for all cameras in the set of cameras.

7. The method according to claim 1 or 2, wherein the capture system has a focal length control device for controlling the focal length of the camera, and the method further comprises controlling the focal length control device to reduce the focal length of one or more cameras before calibrating the camera.

8. The method according to claim 1 or 2, wherein the step of attempting to determine the exogenous parameters of the camera includes inputting the set of images to an algorithm configured to attempt to determine the exogenous parameters of input images.

9. A method for calibrating a multi-camera capture system for capturing a scene for 3D reconstruction of a target display space, wherein the target display space defines a portion of the scene to be 3D reconstructed using images from the capture system, and the capture system comprises two or more cameras and a control system for adapting the poses of the cameras, and the method is The steps of controlling the control system to mechanically adjust the pose of one or more cameras through a plurality of predefined rotations and / or predefined translations, A step of obtaining a set of images from each pose-adjusted camera, wherein each set of images includes multiple images from various camera poses, The steps include determining the relative extrinsic parameters of a pose-adjusted camera using a corresponding set of images, The steps include determining the extrinsic parameters of all cameras using the aforementioned relative extrinsic parameters and images from all cameras, A method comprising calibrating the camera by the means described above.

10. The method according to claim 9, wherein the capture system has a focal length control device for controlling the focal length of the camera, and the method further comprises the step of controlling the focal length control device to reduce the focal length of one or more cameras before calibrating the camera.

11. The method according to claim 9 or 10, wherein the step of determining the exogenous parameters includes inputting the set of images to an algorithm configured to determine exogenous parameters for input images,

12. A computer program that, when executed on a computer, causes the computer to perform the method according to any one of claims 1 to 11.

13. A system for calibrating a multi-camera capture system that captures a scene for 3D reconstruction of a target display space, wherein the target display space defines a portion of the scene to be 3D reconstructed using images from the capture system, and comprises two or more cameras and a control system for adapting the poses of the cameras. The system in question is A step of acquiring a set of images from the camera, wherein at least two images in the set of images correspond to different cameras. The steps include attempting to determine the camera's extrinsic parameters based on the aforementioned set of images, The steps include controlling the control system to mechanically adjust the pose of one or more cameras based on the trial that determines the exogenous parameters, A system having a processor configured to calibrate the camera by iteratively performing the following.

14. The system according to claim 13, wherein the processor is configured to control the control system to mechanically adjust the pose of a camera by predefined rotation and / or predefined translation for a camera whose extrinsic parameters have not been determined.

15. A system for calibrating a multi-camera capture system that captures a scene for 3D reconstruction of a target display space, wherein the target display space defines a portion of the scene to be 3D reconstructed using images from the capture system, the capture system comprises two or more cameras and a control system for adapting the poses of the cameras, and the system The steps of controlling the control system to mechanically adjust the pose of one or more cameras through a plurality of predefined rotations and / or predefined translations, A step of obtaining a set of images from each pose-adjusted camera, wherein each set of images includes multiple images from various camera poses, The steps include determining the relative extrinsic parameters of a pose-adjusted camera using a corresponding set of images, The steps include determining the extrinsic parameters of all cameras using the aforementioned relative extrinsic parameters and images from all cameras, A system having a processor configured to calibrate the camera.

Images