Sensor perturbation

Patent metadata

Field	Value
Publication number	US-12007228-B2
Application number	US-202117520496-A
Country	US
Kind code	B2
Filing date	Nov 5, 2021
Priority date	Aug 11, 2017
Publication date	Jun 11, 2024
Grant date	Jun 11, 2024

Abstract

Perception sensors of a vehicle can be used for various operating functions of the vehicle. A computing device may receive sensor data from the perception sensors, and may calibrate the perception sensors using the sensor data, to enable effective operation of the vehicle. To calibrate the sensors, the computing device may project the sensor data into a voxel space, and determine a voxel score comprising an occupancy score and a residual value for each voxel. The computing device may then adjust an estimated position and/or orientation of the sensors, and associated sensor data, from at least one perception sensor to minimize the voxel score. The computing device may calibrate the sensor using the adjustments corresponding to the minimized voxel score. Additionally, the computing device may be configured to calculate an error in a position associated with the vehicle by calibrating data corresponding to a same point captured at different times.

First claim

What is claimed is:

1. A system comprising:
one or more processors; and
one or more computer readable storage media communicatively coupled to the one or more processors and storing instructions that are executable by the one or more processors to perform operations comprising:
receive first sensor data associated with a first position and a first orientation;
associate the first sensor data with a voxel space;
receive second sensor data associated with a second position and a second orientation;
associate the second sensor data with the voxel space;
determine, based at least in part on the first sensor data and the second sensor data, a value associated with a first portion of the voxel space, wherein the value represents a measure of planarity of a subset of the first sensor data and the second sensor data that falls within the first portion and is indicative of how well the subset conforms to a common plane of the first portion;
determine, as a transform and based at least in part on the value, one or more of a rotation or translation associated with the second sensor data, wherein the transform represents a first pose associated with the first sensor data relative to a second pose associated with the second sensor data, wherein determining the transform comprises perturbing one or more of the second position or second orientation by a magnitude and along one or more of a directional dimension or an orientation dimension; and
control, based on the transform, an operation of a sensor or a vehicle.

2. The system of claim 1 , wherein:
perturbing the one or more of the second position or second orientation comprises iteratively:
transforming, as transformed sensor data and by one or more of a set of rotations or a set of translations having varying magnitudes, the second sensor data;
associating the transformed sensor data with the voxel space;
determining, based at least in part on the voxel space, an updated value; and
determining the transform is based at least in part on an optimized updated value.

3. The system of claim 1 , wherein the magnitude associated with the translation or the rotation is based at least in part on a voxel size of individual voxels in the voxel space.

4. The system of claim 1 , wherein the operation includes determining a drift error associated with a position of the vehicle based at least in part on the transform.

5. The system of claim 1 , wherein the operation includes calibrating a sensor associated with the second sensor data.

6. The system of claim 1 , wherein determining the value comprises one or more of:
determining a difference between data associated with a voxel of the first portion of the voxel space and a best fit plane associated with the data, or
performing a statistical analysis of data associated with the voxel.

7. The system of claim 1 , wherein:
the first sensor data is associated with a second portion of the voxel space based on at least one of the first position or the first orientation,
the second sensor data is associated with a third portion of the voxel space based on at least one of the second position or the second orientation, and
the first portion comprises the second portion and the third portion.

8. A method comprising:
receiving first sensor data and second sensor data, the first sensor data associated with a first position and first orientation and the second sensor data associated with a second position and second orientation;
associating the first sensor data and the second sensor data with a voxel space;
determining, based at least in part on the first sensor data and the second sensor data, a value associated with a portion of the voxel space, wherein the value represents a measure of planarity of a subset of the first sensor data and the second sensor data that falls within the portion and is indicative of how well the subset conforms to a common plane of the first portion;
determining, as a transform and based at least in part on the value, one or more of a rotation or a translation to apply to the second sensor data, wherein the transform represents a first pose associated with the first sensor data relative to a second pose associated with the second sensor data, wherein determining the transform comprises perturbing a pose associated with the second sensor data by a magnitude and along one or more of a directional dimension or an orientation dimension; and
controlling, based at least in part on the transform, an operation of a sensor or a vehicle.

9. The method of claim 8 , wherein perturbing includes modifications to two or three dimensions simultaneously.

10. The method of claim 8 , wherein a magnitude associated with the translation or the rotation is based at least in part on a voxel size of individual voxels in the voxel space.

11. The method of claim 8 , wherein determining the transform comprises:
iteratively determining, based at least in part on a set of transforms and the second sensor data, a plurality of updated values; and
determining the transform as one of the set of transforms having an optimum updated value.

12. The method of claim 8 , wherein determining the value comprises one or more of:
determining a difference between data associated with a voxel of the portion of the voxel space and a best fit plane associated with the data, or
performing a statistical analysis of data associated with the voxel.

13. The method of claim 8 , wherein the operation includes one or more of:
determining a drift error associated with a position of the vehicle based at least in part on the transform, or
calibrating the sensor.

14. One or more non-transitory computer-readable storage media communicatively coupled to one or more processors and storing instructions that are executable by the one or more processors to perform operations comprising:
receiving first sensor data and second sensor data;
associating the first sensor data and the second sensor data with a voxel space;
determining, based at least in part on a portion of the voxel space, a value, wherein the value represents a measure of planarity of a subset of the first sensor data and the second sensor data that falls within the portion and is indicative of how well the subset conforms to a common plane of the first portion;
determining, based at least in part on the value, a transform to align the first sensor data and the second sensor data, wherein the transform represents a first pose associated with the first sensor data relative to a second pose associated with the second sensor data, wherein determining the transform comprises perturbing a pose associated with the second sensor data by a magnitude and along one or more of a directional dimension or an orientation dimension; and
controlling, based at least in part on the transform, an operation of a sensor associated with the second sensor data or a vehicle associated with the second sensor data.

15. The one or more non-transitory computer-readable storage media of claim 14 , wherein:
the pose comprises a position and an orientation, and
perturbing the pose includes modifications to two or three dimensions simultaneously.

16. The one or more non-transitory computer-readable storage media of claim 14 , wherein determining the value comprises one or more of:
determining a best fit of a plane with data associated with a voxel of the portion of the voxel space, or
performing a statistical analysis of data associated with the voxel.

17. The one or more non-transitory computer-readable storage media of claim 14 , wherein a magnitude associated

Assignees

• Zoox Inc

Inventors

• Adams Derek
• Baldwin Ian
• Douillard Bertrand Robert
• Levinson Jesse Sol

Classifications

• G01C21/1652

  with ranging devices, e.g. LIDAR or RADAR · CPC title

• G01C21/1656

  with passive imaging devices, e.g. cameras · CPC title

• G01S17/06

  Systems determining position data of a target · CPC title

• G01S17/88

  Lidar systems specially adapted for specific applications · CPC title

• G01S7/4026

  Antenna boresight · CPC title