Integrated fiducial marker for simultaneously calibrating sensors of different types
The present teaching relates to different configurations for facilitating calibration of multiple sensors of different types. A plurality of fiducial markers are arranged in space for simultaneously calibrating multiple sensors of different types. Each of the plurality of fiducial markers has a feature point thereon and is provided to enable the multiple sensors to calibrate by detecting the features points and estimating their corresponding 3D coordinates with respect to respective coordinate systems of the multiple sensors.
1. A computer-implemented method comprising:
receiving, by a computing system, a GPS signal associated with a marker;
determining, by the computing system, a location of the marker based on the GPS signal associated with the marker;
calibrating, by the computing system, a visual based sensor based on a visual identification of a feature point on the marker and the location of the marker; and
calibrating, by the computing system, a depth based sensor based on a depth measurement of the feature point on the marker,
wherein the calibrating of the visual based sensor and the calibrating of the depth based sensor are i) performed while a vehicle associated with the computing system is in motion and ii) initiated based on a distance between the vehicle and the marker, wherein the distance between the vehicle and the marker is determined based at least in part on the GPS signal associated with the marker.
2. The computer-implemented method of claim 1 , further comprising:
calibrating, by the computing system, a radar based sensor based on a distance measurement of the feature point on the marker and the location of the marker.
3. The computer-implemented method of claim 1 , further comprising:
identifying, by the computing system, the feature point in a first coordinate system used by the visual based sensor;
identifying, by the computing system, the feature point in a second coordinate system used by the depth based sensor; and
determining, by the computing system, a transformation matrix for converting first points in the first coordinate system to second points in the second coordinate system based on the feature point.
4. The computer-implemented method of claim 1 , wherein the calibrating the visual based sensor comprises:
identifying, by the computing system, an intersection of different color patches on the marker, wherein the intersection corresponds with a location of the feature point on the marker.
5. The computer-implemented method of claim 1 , wherein the calibrating the depth based sensor comprises:
determining, by the computing system, scanned depth values of the marker; and
determining, by the computing system, a local minimum depth value or a peak depth value from the scanned depth values of the marker, wherein the local minimum depth value or the peak depth value corresponds with a location of the feature point on the marker.
6. The computer-implemented method of claim 1 , further comprising:
determining, by the computing system, the marker is within a distance threshold of the visual based sensor and the depth based sensor, wherein the distance threshold is increased based on light conditions, and wherein the distance threshold is decreased based on dark conditions; and
initiating, by the computing system, calibration of the visual based sensor and the depth based sensor.
7. The computer-implemented method of claim 1 , further comprising:
receiving, by the computing system, a command from a user to initiate calibration of the visual based sensor and the depth based sensor; and
initiating, by the computing system, the calibration of the visual based sensor and the depth based sensor based on the command.
8. The computer-implemented method of claim 1 , wherein the visual based sensor and the depth based sensor are calibrated simultaneously.
9. The computer-implemented method of claim 1 , further comprising:
determining, by the computing system, a first orientation of the marker;
determining, by the computing system, a discrepancy between the first orientation of the marker and a second orientation of a vehicle; and
generating, by the computing system, a control signal to position the vehicle based on the discrepancy.
10. The computer-implemented method of claim 1 , wherein the visual based sensor is a stereo camera pair, wherein the depth based sensor is a LiDAR sensor, and wherein the stereo camera pair and the LiDAR sensor are installed on a sensor rack.
11. A system comprising:
at least one processor; and
a memory storing instructions that, when executed by the at least one processor, cause the system to perform operations comprising:
receiving a GPS signal associated with a marker;
determining a location of the marker based on the GPS signal associated with the marker;
calibrating a visual based sensor based on a visual identification of a feature point on the marker and the location of the marker; and
calibrating a depth based sensor based on a depth measurement of the feature point on the marker,
wherein the calibrating of the visual based sensor and the calibrating of the depth based sensor are i) performed while a vehicle associated with the system is in motion and ii) initiated based on a distance between the vehicle and the marker, wherein the distance between the vehicle and the marker is determined based at least in part on the GPS signal associated with the marker.
12. The system of claim 11 , the operations further comprising:
calibrating a radar based sensor based on a distance measurement of the feature point on the marker and the location of the marker.
13. The system of claim 11 , the operations further comprising:
identifying the feature point in a first coordinate system used by the visual based sensor;
identifying the feature point in a second coordinate system used by the depth based sensor; and
determining a transformation matrix for converting first points in the first coordinate system to second points in the second coordinate system based on the feature point.
14. The system of claim 11 , wherein the calibrating the visual based sensor comprises:
identifying an intersection of different color patches on the marker, wherein the intersection corresponds with a location of the feature point on the marker.
15. The system of claim 11 , wherein the calibrating the depth based sensor comprises:
determining scanned depth values of the marker; and
determining a local minimum depth value or a peak depth value from the scanned depth values of the marker, wherein the local minimum depth value or the peak depth value corresponds with a location of the feature point on the marker.
16. A non-transitory computer-readable storage medium including instructions that, when executed by at least one processor of a computing system, cause the computing system to perform:
receiving a GPS signal associated with a marker;
determining a location of the marker based on the GPS signal associated with the marker;
calibrating a visual based sensor based on a visual identification of a feature point on the marker and the location of the marker; and
calibrating a depth based sensor based on a depth measurement of the feature point on the marker,
wherein the calibrating of the visual based sensor and the calibrating of the depth based sensor are i) performed while a vehicle associated with the computing system is in motion and ii) initiated based on a distance between the vehicle and the marker, wherein the distance between the vehicle and the marker is determined based at least in part on the GPS signal associated with the marker.
17. The non-transitory computer-readable storage medium of claim 16 , the operations further comprising:
calibrating a radar based sensor based on a distance measurement of the feature point on the marker and the location of the marker.
18. The computer-implemented method of claim 1 , wherein the vehicle is without access to a specialized environment and the marker is positioned along a highway.
19. The computer-implemented method of claim 1 , wherein the distance between the vehicle and the marker is further determined based on a GPS signal associated with the vehicle.
20. The non-transitory computer-readable storage medium of claim 16 , the operations further comprising:
identifying the feature point in a first coordinate system used by the visual based sensor;
identifying the feature point in a second coordinate system used by the depth based sensor; and
determining a transformation matrix for converting first points in the first coordinate system to second points in the second coordinate system based on the feature point.