Registration of range images using virtual gimbal information

Patent metadata

Field	Value
Publication number	US-10242458-B2
Application number	US-201715493859-A
Country	US
Kind code	B2
Filing date	Apr 21, 2017
Priority date	Apr 21, 2017
Publication date	Mar 26, 2019
Grant date	Mar 26, 2019

Abstract

Systems and methods configured to generate virtual gimbal information for range images produced from 3D depth scans are described. In operation according to embodiments, known and advantageous spatial geometries of features of a scanned volume are exploited to generate virtual gimbal information for a pose. The virtual gimbal information of embodiments may be used to align a range image of the pose with one or more other range images for the scanned volume, such as for combining the range images for use in indoor mapping, gesture recognition, object scanning, etc. Implementations of range image registration using virtual gimbal information provide a realtime one shot direct pose estimator by detecting and estimating the normal vectors for surfaces of features between successive scans which effectively imparts a coordinate system for each scan with an orthogonal set of gimbal axes and defines the relative camera attitude.

First claim

The invention claimed is:

1. A method for determining a common coordinate system for registration of a plurality of range images, the method comprising:
identifying, by a processor-based platform configured for processing images, a feature present in a range image of the plurality of range images;
deconstructing, by the processor-based platform, the feature present in the range image into virtual gimbal information for the range image using estimates of two or more orthogonal axes for the feature present in the range image, wherein the virtual gimbal information defines the common coordinate system for the plurality of range images for registration of the range image with a second range image of the plurality of range images from a different pose than the range image;
identifying, by the processor-based platform, the feature present in the second range image of the plurality of range images;
deconstructing, by the processor-based platform, the feature present in the second range image into second virtual gimbal information for the second range image using estimates of two or more orthogonal axes for the feature present in the second range image, wherein the second virtual gimbal information defines a coordinate system for the second range image for translation of the second range image to the common coordinate system; and
registering, by the processor-based platform, one of the range image or the second range image to the other one of the second range image or the range image based upon the common coordinate system.

2. The method of claim 1 , wherein the feature comprises a corner present in the range image.

3. The method of claim 2 , wherein the range image comprises a depth map for a scanned volume.

4. The method of claim 1 , wherein the two or more orthogonal axes comprise three orthogonal axes, and wherein the common coordinate system for the range images comprises a coordinate system having an X axis, a Y axis, and a Z axis.

5. The method of claim 1 , further comprising:
estimating a pose orientation for the range image using the virtual gimbal information for the range image;
estimating a second pose orientation for the second range image using the second virtual gimbal information for the second range image; and
determining relative translation information for registering the range image and the second range image to the common coordinate system.

6. The method of claim 5 , wherein the estimating the pose orientation for the range image generates a pose matrix for the range image and the estimating the second pose orientation for the second range image generates a second pose matrix for the second range image, and wherein the determining the relative translation information generates a translation vector using the pose matrix for the range image and the second pose matrix for the second range image.

7. The method of claim 6 , further comprising:
combining the range image and the second range image.

8. An apparatus for determining a common coordinate system for registration of a plurality of range images, the apparatus comprising:
a memory;
at least one processor coupled to the memory, wherein the at least one processor is configured:
to identify a feature present in a range image of the plurality of range images;
to deconstruct the feature present in the range image into virtual gimbal information for the range image using estimates of two or more orthogonal axes for the feature present in the range image, wherein the virtual gimbal information defines the common coordinate system for the plurality of range images for registration of the range image with a second range image of the plurality of range images from a different pose than the range image;
to identify the feature present in the second range image of the plurality of range images;
to deconstruct the feature present in the second range image into second virtual gimbal information for the second range image using estimates of two or more orthogonal axes for the feature present in the second range image, wherein the second virtual gimbal information defines a coordinate system for the second range image for translation of the second range image to the common coordinate system; and
to register one of the range image or the second range image to the other one of the second range image or the range image based upon the common coordinate system.

9. The apparatus of claim 8 , wherein the feature comprises a corner present in the range image.

10. The apparatus of claim 9 , wherein the range image comprises a depth map for a scanned volume.

11. The apparatus of claim 8 , wherein the two or more orthogonal axes comprise three orthogonal axes, and wherein the common coordinate system for the range image comprises a coordinate system having an X axis, a Y axis, and a Z axis.

12. The apparatus of claim 8 , wherein the at least one processor is further configured:
to estimate a pose orientation for the range image using the virtual gimbal information for the range image;
to estimate a second pose orientation for the second range image using the second virtual gimbal information for the second range image; and
to determine relative translation information for registering the range image and the second range image to the common coordinate system.

13. The apparatus of claim 12 , wherein the at least one processor is further configured:
to generate a pose matrix for the range image for estimating the pose orientation for the range image;
to generate a second pose matrix for the second range image for estimating the second pose orientation for the second range image; and
to generate a translation vector using the pose matrix for the range image and the second pose matrix for the second range image for determining the relative translation information.

14. The apparatus of claim 13 , wherein the at least one processor is further configured:
to combine the range image and the second range image.

15. A non-transitory computer-readable medium having program code recorded thereon for determining a common coordinate system for registration of a plurality of range images, the program code comprising:
program code executable by a computer for causing the computer to:
identify a feature present in a range image of the plurality of range images; and
deconstruct the feature present in the range image into virtual gimbal information for the range image using estimates of two or more orthogonal axes for the feature present in the range image, wherein the virtual gimbal information defines the common coordinate system for the plurality of range images for registration of the range image with a second range image of the plurality of range images from a different pose than the range image;
identify the feature present in the second range image of the plurality of range images;
deconstruct the feature present in the second range image into second virtual gimbal information for the second range image using estimates of two or more orthogonal axes for the feature present in the second range image, wherein the second virtual gimbal information defines a coordinate system for the second range image for translation of the second range image to the common coordinate system; and
register one of the range image or the second range image to the other one of the second range image or the range image based upon the common coordinate system.

16. The non-transitory computer-readable medium of claim 15 , wherein the feature comprises a corner present in the range image.

17. The non-transitory computer-readable medium of claim

Assignees

• Qualcomm Inc

Inventors

• Nash James
• Atanassov Kalin
• Lindner Albrecht Johannes

Classifications

• G06T7/33

  using feature-based methods · CPC title

• G06K9/4604

  Physics · mapped topic

• G06T7/73 Primary

  using feature-based methods · CPC title

• G06T2207/10028

  Range image; Depth image; 3D point clouds · CPC title