Method and apparatus with sensor calibration

Patent metadata

Field	Value
Publication number	US-12561838-B2
Application number	US-202217575092-A
Country	US
Kind code	B2
Filing date	Jan 13, 2022
Priority date	Jun 25, 2021
Publication date	Feb 24, 2026
Grant date	Feb 24, 2026

Abstract

A processor-implemented method with sensor calibration includes: estimating a portion of a rotation parameter for a target sensor among a plurality of sensors based on a capture of a reference object; estimating another portion of the rotation parameter for the target sensor based on an intrinsic parameter of the target sensor and a focus of expansion (FOE) determined based on sensing data collected with consecutive frames by the target sensor while the electronic device rectilinearly moves based on one axis; and performing calibration by determining a first extrinsic parameter for the target sensor based on the portion and the other portion of the rotation parameter.

First claim

What is claimed is:

1 . A processor-implemented method with sensor calibration, the method comprising:
estimating a portion of a rotation parameter for a target sensor among a plurality of sensors based on a capture of a reference object; estimating another portion of the rotation parameter for the target sensor based on an intrinsic parameter of the target sensor and a focus of expansion (FOE), the FOE being a point at which an optical flow converges in consecutive frame images, the another portion being determined based on sensing data collected with consecutive frames by the target sensor while an electronic device rectilinearly moves based on one axis; and performing calibration by determining a first extrinsic parameter for the target sensor based on the portion and the another portion of the rotation parameter, wherein the determining of the first extrinsic parameter comprises: determining first candidate extrinsic parameters respectively for at least a portion of a plurality of frames; and determining the first extrinsic parameter by integrating the first candidate extrinsic parameters, wherein the determining of the first extrinsic parameter comprises removing an outlier value from the first candidate extrinsic parameters and integrating the remaining values.

2 . The method of claim 1 , wherein the estimating of the portion of the rotation parameter comprises:
identifying a plurality of corner points from the sensing data as reference points; and determining whether the reference object is in the capture based on the plurality of identified corner points.

3 . The method of claim 1 , wherein the estimating of the portion of the rotation parameter comprises:
identifying a reference marking from the reference object identified from sensing data generated through the target sensor; determining an orientation of the reference object based on the reference marking; and estimating the portion of the rotation parameter based on the determined orientation.

4 . The method of claim 3 , wherein the estimating of the portion of the rotation parameter based on the determined orientation comprises:
determining default coordinate information according to a world coordinate system that is based on the reference marking for a plurality of reference points of the reference object; and estimating the portion of the rotation parameter based on coordinate information determined by projecting the default coordinate information to an image coordinate system based on the determined orientation and coordinate information detected for the plurality of reference points in the sensing data.

5 . The method of claim 1 , wherein the estimating of the portion of the rotation parameter comprises determining either one or both of a rolling component and a pitch component of the rotation parameter.

6 . The method of claim 1 , wherein the estimating of the another portion of the rotation parameter comprises determining a yaw component of the rotation parameter in the first extrinsic parameter.

7 . The method of claim 1 , further comprising:
determining a translation component in the first extrinsic parameter based on dimension data indicating a positional relationship in which the target sensor is disposed in the electronic device.

8 . The method of claim 1 , further comprising:
determining a height component of a translation component in the first extrinsic parameter based on either one or both of dimension data for the target sensor and a translation component between the reference object and the target sensor determined according to the capture of the reference object.

9 . The method of claim 1 , further comprising:
performing coordinate conversion between a coordinate system of the target sensor and a ground coordinate system, using the determined first extrinsic parameter.

10 . The method of claim 1 , further comprising:
performing calibration of a second extrinsic parameter between two different sensors, in response to a common reference object being identified at the same time from sensing data captured by the two different sensors among the plurality of sensors.

11 . The method of claim 10 , wherein the performing of the calibration of the second extrinsic parameter comprises determining, by the two different sensors, the second extrinsic parameter between the target sensor and another sensor using an extrinsic parameter between the reference object and the target sensor and an extrinsic parameter between the reference object and the another sensor.

12 . The method of claim 10 , further comprising:
performing coordinate conversion between a coordinate system of the target sensor and a coordinate system of another sensor, using the determined second extrinsic parameter.

13 . The method of claim 1 , wherein, in response to only a sensor disposed on one side of the electronic device detecting the reference object among the plurality of sensors, the electronic device is configured to move such that a sensor disposed on a side opposite to the one side based on a center of the electronic device detects the reference object.

14 . The method of claim 1 , wherein a range of a field of view of the target sensor comprises a longitudinal axis of a ground coordinate system set for the electronic device.

15 . The method of claim 1 , wherein the plurality of sensors comprise any one or any combination of any two or more of camera sensors, lidar sensors, infrared sensors, and ultrasonic sensors.

16 . The method of claim 1 , wherein the electronic device is mounted on any one or any combination of any two or more of a drone, a vehicle, and a drivable robot.

17 . A non-transitory computer-readable storage medium storing instructions that, when executed by one or more processors, configure the one or more processors to perform the method of claim 1 .

18 . An electronic device, comprising:
one or more processors configured to:
estimate a portion of a rotation parameter for a target sensor among a plurality of sensors based on a capture of a reference object;
estimate another portion of the rotation parameter for the target sensor based on an intrinsic parameter of the target sensor and a focus of expansion (FOE), the FOE being a point at which an optical flow converges in consecutive frame images, the another portion being determined based on sensing data collected with consecutive frames by the target sensor while an electronic device rectilinearly moves based on one axis;
perform calibration by determining a first extrinsic parameter for the target sensor based on the portion and the another portion of the rotation parameter;
determine first candidate extrinsic parameters respectively for at least a portion of a plurality of frames; and
determine the first extrinsic parameter by integrating the first candidate extrinsic parameters; and
wherein the determining of the first extrinsic parameter comprises removing an outlier value from the first candidate extrinsic parameters and integrating the remaining values.

19 . The device of claim 18 , wherein the device is a vehicle further comprising the plurality of sensors disposed such that at least a portion of a field of view of each sensor overlaps a field of view of another sensor.

Assignees

• Samsung Electronics Co Ltd

Inventors

• Sung Kapje
• Park Daeul
• Lee Dongwook
• Lee Jaewoo

Classifications

• G06T7/85 Primary

  Stereo camera calibration · CPC title

• G06T2207/30252

  Vehicle exterior; Vicinity of vehicle · CPC title

• G06T2207/30204

  Marker · CPC title

• G06T7/73

  using feature-based methods · CPC title

• G06T7/97

  Determining parameters from multiple pictures (depth or shape recovery from multiple images G06T7/55; stereo camera calibration G06T7/85) · CPC title