Hand-eye calibration methods, systems, and storage media for robots

1 . A hand-eye calibration method for a robot, comprising:

obtaining images of a target body of a calibration target acquired by an imaging device when the robot is in different poses, the calibration target further including an infrared light emitting component and an infrared light modulating component, the infrared light emitting component being provided inside the target body, the infrared light emitting component being configured to emit an infrared light in an infrared light waveband that the imaging device is capable of collecting to the target body, different regions of the target body having different transmittances to the infrared light emitted by the infrared light emitting component, and the infrared light modulating component being configured to control the infrared light emitting component to emit an infrared light of at least one of a preset waveband or a preset intensity;

for any pose,

determining, based on at least one image collected in the pose, a first transformation relationship, in the pose, between a first coordinate system of the calibration target and a second coordinate system of the imaging device;

obtaining a second transformation relationship between a third coordinate system and a fourth coordinate system of the robot; and

determining, based on the first transformation relationships and the second transformation relationships corresponding to the different poses, a third transformation relationship between the second coordinate system and one of the third coordinate system and the fourth coordinate system, the third coordinate system being related to a base of the robot, and the fourth coordinate system being related to an end-of-arm tool (EOAT) of a robotic arm of the robot.

2 . The hand-eye calibration method according to claim 1 , wherein the obtaining images of a target body of a calibration target acquired by an imaging device when the robot is in different poses includes:

adjusting parameter information of at least one of the calibration target or the imaging device; and

obtaining the at least one image collected in the pose by controlling the imaging device to perform imaging on the target body based on the adjusted parameter information.

3 . The hand-eye calibration method according to claim 2 , wherein the adjusting parameter information of at least one of the calibration target or the imaging device includes:

controlling, based on an infrared light waveband that the imaging device is capable of collecting, the infrared light emitting component to emit an infrared light of a corresponding waveband by the infrared light modulating component.

4 . The hand-eye calibration method according to claim 2 , wherein the adjusting parameter information of at least one of the calibration target or the imaging device includes:

controlling, by the infrared light modulating component, the infrared light emitting component to emit an infrared light of a current intensity;

obtaining a current image of the target body acquired by the imaging device under the infrared light of the current intensity;

determining whether a quality of the current image meets a condition; and

in response to a determination that the quality of the current image does not meet the condition, adjusting, by the infrared light modulating component, the current intensity of the infrared light emitted by the infrared light emitting component.

5 . The hand-eye calibration method according to claim 4 , wherein the determining whether a quality of the current image meets a condition includes:

obtaining, based on the current image, feature data of the target body in the current image;

obtaining reference feature data corresponding to the feature data of the target body in the current image; and

determining, based on the feature data and the reference feature data, whether the quality of the current image meets the condition.

6 . The hand-eye calibration method according to claim 5 , wherein the determining, based on the feature data and the reference feature data, whether the quality of the current image meets the condition includes:

obtaining a reprojection error based on the feature data and the reference feature data;

determining whether the reprojection error is greater than a preset threshold; and

in response to a determination that the reprojection error is greater than the preset threshold, determining that the quality of the current image does not meet the condition.

7 . The hand-eye calibration method according to claim 6 , wherein the obtaining a reprojection error based on the feature data and the reference feature data includes:

converting, based on intrinsic parameters and extrinsic parameters of the imaging device, the reference feature data to a current image coordinate system to obtain estimated feature data; and

determining, based on an error between the estimated feature data and the feature data, the reprojection error.

8 . The hand-eye calibration method according to claim 1 , wherein the determining, based on the at least one image collected in the pose, a first transformation relationship between a first coordinate system of the calibration target and a second coordinate system of the imaging device in the pose includes:

the at least one image includes a plurality of images, a count of the plurality of images being greater than a threshold, for each image of the plurality of images,

extracting feature data from each image;

determining, based on the feature data, an extrinsic parameter matrix of the imaging device corresponding to each image; and

determining the first transformation relationship based on a plurality of extrinsic parameter matrices corresponding to the plurality of images.

9 . The hand-eye calibration method according to claim 8 , wherein the determining, based on the feature data, an extrinsic parameter matrix of the imaging device corresponding to each image includes:

performing a filtering operation on the feature data to obtain filtered feature data; and

determining the extrinsic parameter matrix based on the filtered feature data.

10 . The hand-eye calibration method according to claim 8 , wherein the determining, based on the feature data, an extrinsic parameter matrix of the imaging device corresponding to each image includes:

determining, based on the feature data, a plurality of first extrinsic parameter matrices corresponding to the plurality of images; and

performing a filtering operation on the plurality of first extrinsic parameter matrices to determine the plurality of extrinsic parameter matrices.

11 . The hand-eye calibration method according to claim 8 , wherein the determining, based on the feature data, an extrinsic parameter matrix of the imaging device corresponding to each image includes:

performing a filtering operation on the feature data to obtain filtered feature data;

determining, based on the filtered feature data, a plurality of second extrinsic parameter matrices corresponding to the plurality of the images; and

performing a filtering operation on the plurality of second extrinsic parameter matrices to determine the plurality of extrinsic parameter matrices.

12 . The hand-eye calibration method according to claim 1 , wherein the determining, based on the first transformation relationships and the second transformation relationships corresponding to the different poses, a third transformation relationship between the second coordinate system and one of the third coordinate system and the fourth coordinate system includes:

constructing, based on the first transformation relationships and the second transformation relationships corresponding to the different poses, a plurality of sets of input data of a nonlinear optimization model, each set of input data corresponding to the first transformation relationship and the second transformation relationship corresponding to each pose of the different poses; and

determining the third transformation relationship based on the nonlinear optimization model and the plurality of sets of input data.

13 . The hand-eye calibration method according to claim 12 , wherein each set of input data of the plurality of sets of input data includes a first relative transformation and a second relative transformation,

the second relative transformation in each set of input data represents a position transformation of the second coordinate system in the pose corresponding to each set of input data relative to a reference pose,

the first relative transformation in each set of input data represents a position transformation of the fourth coordinate system in the pose corresponding to each set of input data relative to the reference pose,

the first relative transformation is constructed based on the second transformation relationship, and

the second relative transformation is constructed based on the first transformation relationship.

14 . A calibration target, comprising a target body, an infrared light emitting component, and an infrared light modulating component, wherein

the infrared light emitting component is provided inside the target body;

the infrared light emitting component is configured to emit an infrared light to the target body, wherein different regions of the target body have different transmittances to the infrared light emitted by the infrared light emitting component; and

the infrared light modulating component is configured to control the infrared light emitting component to emit an infrared light of at least one of a preset waveband or a preset intensity.

15 . A hand-eye calibration system for a robot, comprising a calibration target, an imaging device, the robot, a processor, and a storage, wherein the storage is configured to store an instruction set, and the processor is configured to execute the instruction set to execute the hand-eye calibration method for the robot including:

obtaining images of a target body of the calibration target acquired by the imaging device when the robot is in different poses, the calibration target further including an infrared light emitting component and an infrared light modulating component, the infrared light emitting component being provided inside the target body, the infrared light emitting component being configured to emit an infrared light in an infrared light waveband that the imaging device is capable of collecting to the target body, different regions of the target body having different transmittances to the infrared light emitted by the infrared light emitting component, and the infrared light modulating component being configured to control the infrared light emitting component to emit an infrared light of at least one of a preset waveband or a preset intensity;

for any pose,

determining, based on at least one image collected in the pose, a first transformation relationship, in the pose, between a first coordinate system of the calibration target and a second coordinate system of the imaging device;

obtaining a second transformation relationship between a third coordinate system and a fourth coordinate system of the robot; and

determining, based on the first transformation relationships and the second transformation relationships corresponding to the different poses, a third transformation relationship between the second coordinate system and one of the third coordinate system and the fourth coordinate system, the third coordinate system being related to a base of the robot, and the fourth coordinate system being related to an end-of-arm tool (EOAT) of a robotic arm of the robot.

16 . The hand-eye calibration system according to claim 15 , wherein the obtaining images of a target body of a calibration target acquired by an imaging device when the robot is in different poses includes:

adjusting parameter information of at least one of the calibration target or the imaging device; and

obtaining the at least one image collected in the pose by controlling the imaging device to perform imaging on the target body based on the adjusted parameter information.

17 . The hand-eye calibration system according to claim 15 , wherein the determining, based on the at least one image collected in the pose, a first transformation relationship between a first coordinate system of the calibration target and a second coordinate system of the imaging device in the pose includes:

the at least one image includes a plurality of images, a count of the plurality of images being greater than a threshold, for each image of the plurality of images,

extracting feature data from each image;

determining, based on the feature data, an extrinsic parameter matrix of the imaging device corresponding to each image; and

determining the first transformation relationship based on a plurality of extrinsic parameter matrices corresponding to the plurality of images.

18 . The hand-eye calibration system according to claim 15 , wherein the determining, based on the first transformation relationships and the second transformation relationships corresponding to the different poses, a third transformation relationship between the second coordinate system and one of the third coordinate system and the fourth coordinate system includes:

constructing, based on the first transformation relationships and the second transformation relationships corresponding to the different poses, a plurality of sets of input data of a nonlinear optimization model, each set of input data corresponding to the first transformation relationship and the second transformation relationship corresponding to each pose of the different poses; and

determining the third transformation relationship based on the nonlinear optimization model and the plurality of sets of input data.

19 . The hand-eye calibration system according to claim 17 , wherein the determining, based on the feature data, an extrinsic parameter matrix of the imaging device corresponding to each image includes:

determining, based on the feature data, a plurality of first extrinsic parameter matrices corresponding to the plurality of images; and

performing a filtering operation on the plurality of first extrinsic parameter matrices to determine the plurality of extrinsic parameter matrices.

20 . The hand-eye calibration system according to claim 17 , wherein the determining, based on the feature data, an extrinsic parameter matrix of the imaging device corresponding to each image includes:

performing a filtering operation on the feature data to obtain filtered feature data;

determining, based on the filtered feature data, a plurality of second extrinsic parameter matrices corresponding to the plurality of the images; and

performing a filtering operation on the plurality of second extrinsic parameter matrices to determine the plurality of extrinsic parameter matrices.