Method and system of 3D image capture with dynamic cameras

Patent metadata

Field	Value
Publication number	US-10003786-B2
Application number	US-201615239528-A
Country	US
Kind code	B2
Filing date	Aug 17, 2016
Priority date	Sep 25, 2015
Publication date	Jun 19, 2018
Grant date	Jun 19, 2018

Abstract

Techniques related to 3D image capture with dynamic cameras.

First claim

What is claimed is:

1. A computer-implemented method of image capture comprising:
positioning a plurality of cameras about at least one scene to be captured and that are automatically movable from scene to scene;
obtaining image data of individual sequences of frames of the at least one scene and from individual cameras of the plurality of cameras; and
determining the location of the image data of the sequences of frames on a shared coordinate system of the plurality of the cameras by using the image data of multiple frames to determine the location of the image data on the coordinate system,
wherein the location of the image data on the shared coordinate system is determined without determining values of the positions of the cameras before the capture of the sequences of frames in order to use the values of the positions in computations to determine the location of the image data on the shared coordinate system.

2. The method of claim 1 comprising moving the cameras to positions to capture the image data wherein the positions are not planned before the capture of the image data, wherein the position comprises a coordinate location of the camera, an orientation of the optical axis of the camera, or both.

3. The method of claim 1 wherein the shared coordinate system is determined by using iterative closest point algorithms.

4. The method of claim 1 comprising using overlapping frames with overlapping fields of view to determine the shared coordinate system based at least in part on the overlapping frames.

5. The method of claim 4 comprising determining the location of at least one frame relative to a frame that overlaps the at least one frame and in the same sequence of frames as the at least one frame.

6. The method of claim 1 comprising locating image data of at least one frame relative to image data of at least one other frame along the same sequence of frames from an individual camera.

7. The method of claim 6 comprising performing the locating by using an iterative closest point algorithm.

8. The method of claim 1 comprising locating image data of at least one frame from one camera of the plurality of cameras relative to the image data of at least one other frame from at least one other camera of the plurality of cameras.

9. The method of claim 8 comprising performing the locating by using an iterative closest point algorithm.

10. The method of claim 8 wherein the at least one frame and another frame have images captured at substantially the same time.

11. The method of claim 1 comprising:
calculating image data locations on a frame in the same sequence of frames of an individual camera to determine coordinate axes for the sequence of frames;
finding aligned frames of at least two different sequences depending on whether the difference in the location of points in an image of a frame on one sequence to the points in an image of a frame of another sequence at the same time as the frame of the one sequence meet a criteria; and
converting the coordinate axes of frames in one of the sequences of the aligned frames to the other sequence of the aligned frames.

12. The method of claim 11 comprising using an iterative closest point transformation matrix to change coordinate axes from one sequence to the coordinate axes of another sequence of frames.

13. The method of claim 1 comprising:
moving the cameras to positions to capture the image data wherein the positions are not planned before the capture of the image data, wherein the position comprises a coordinate location of the camera, an orientation of the optical axis of the camera, or both, wherein the shared coordinate system is determined by using iterative closest point algorithms;
using overlapping frames with overlapping fields of view to determine the shared coordinate system based at least in part on the overlapping frames;
locating image data of at least one frame relative to image data of at least one other frame along the same sequence of frames from an individual camera, and performing the locating by using an iterative closest point algorithm;
locating image data of at least one frame from one camera of the plurality of cameras relative to the image data of at least one other frame from at least one other camera of the plurality of cameras, and performing the locating by using an iterative closest point algorithm, wherein the at least one frame and another frame have images captured at substantially the same time;
calculating image data locations on a frame in the same sequence of frames of an individual camera to determine coordinate axes for the sequence of frames;
finding aligned frames of at least two different sequences depending on whether the difference in the location of points in an image of a frame on one sequence to the points in an image of a frame of another sequence at the same time as the frame of the one sequence meet a criteria;
converting the coordinate axes of frames in one of the sequences of the aligned frames to the other sequence of the aligned frames; and
using an iterative closest point transformation matrix to change coordinate axes from one sequence to the coordinate axes of another sequence of frames.

14. A computer-implemented system comprising:
a plurality of movable cameras to position about at least one scene to be captured and automatically movable from scene to scene;
at least one memory to receive image data from the plurality of cameras;
at least one processor communicatively coupled to the at least one of the memory; and
a 3D space generation unit operated by the at least one processor and to:
obtain image data including depth data of sequences of frames of the at least one scene, wherein individual sequences are captured from the individual cameras of the plurality of cameras; and
determine the location of the image data of the sequences of frames on a shared three-dimensional coordinate system of the plurality of the cameras by using the image data of multiple frames to determine the location of the image data on the coordinate system comprising matching point clouds on overlapping frames,
wherein the location of the image data on the shared coordinate system is determined without using positions of the cameras determined before the capture of the sequences of frames.

15. The system of claim 14 wherein the 3D space generation unit is to select one camera of the plurality of cameras as a master camera, wherein the axes of the master camera is used as the axes of the shared coordinate system.

16. The system of claim 14 wherein the 3D space generation unit is to move the cameras to positions to capture the image data wherein the positions are not planned before the capture of the image data, wherein the position comprises a coordinate location of the camera, an orientation of the optical axis of the camera, or both.

17. The system of claim 14 wherein the shared coordinate system is determined by using iterative closest point algorithms that matches the point clouds.

18. The system of claim 14 wherein the point clouds are matched by using a mean square error type of algorithm.

19. The system of claim 18 wherein the 3D space generation unit is to determine the location of at least one frame relative to a frame that overlaps the at least one frame and in the same sequence of frames as the at least one frame.

20. The system of claim 14 wherein the 3D space generation unit is to locate image data of at least one frame relative to image data of at least one other frame along the same seque

Assignees

• Intel Corp

Inventors

• Kouperman Vladimir
• Kamhi Gila
• Zamir Nadav
• Hurwitz Barak

Classifications

• G06T2207/30244

  Camera pose · CPC title

• G06T2207/10021

  Stereoscopic video; Stereoscopic image sequence · CPC title

• G06T2207/10016

  Video; Image sequence · CPC title

• H04N13/0242

  Electricity · mapped topic

• H04N2013/0081

  Depth or disparity estimation from stereoscopic image signals · CPC title