Calibration method, system, device and storage medium for layered motion of robot

Abstract

This invention discloses a method, system, device, and storage medium for calibrating layered motion of a robot. The method includes: acquiring robot teaching point pose data and calibration parameter data; performing translational and rotational sampling according to a preset uniformity rule based on the teaching point pose data and the calibration parameter data to determine valid sampled pose sequence data; controlling the robot's layered motion based on the valid sampled pose sequence data to acquire calibration board image data and actual pose data; determining calibration board pose data based on the calibration board image data; performing validity verification on the calibration board pose data and the actual pose data; and, after all sampled poses have undergone validity verification, completing the robot-camera hand-eye calibration using a pre-constructed hand-eye calibration solution model based on the valid calibration board pose data and the valid actual pose data. This invention achieves beneficial effects in improving calibration efficiency and accuracy.

Description

Technical Field

[0001] This invention relates to the field of industrial robot and machine vision hand-eye calibration technology, and in particular to a calibration method, system, device and storage medium for robot layered motion. Background Technology

[0002] In the field of hand-eye calibration technology, traditional hand-eye calibration sampling operations rely heavily on operators manually teaching and sampling point by point. It requires manual control of robot movement and positioning, as well as capturing and collecting calibration images. Manual sampling involves great uncertainty and subjectivity, and has many inherent defects.

[0003] When operators manually select calibration poses, the lack of standardized and systematic planning, relying entirely on personal experience for arbitrary point placement, not only easily leads to uneven distribution of sampling points, frequent duplicate points, and redundant invalid points, but also causes ill-conditioned calibration matrices and unstable calibration accuracy due to disordered sampling. Furthermore, differences in point selection habits and operating standards among different operators result in extremely poor consistency of samples collected by different operators, leading to inconsistent calibration results with weak universality. Manual operation also presents significant safety risks; the robotic arm is prone to collisions with the surrounding environment or itself, and emergency situations require manual intervention with delayed response. Manual data collection requires repeated fine-tuning of the robot's pose, and collecting sufficient calibration samples is time-consuming, resulting in low overall calibration efficiency. In addition, traditional calibration processes lack online quality assessment mechanisms, making it impossible to control sample validity in real time. The inclusion of invalid samples easily leads to calibration failures, requiring a complete restart after each failure, significantly increasing operating costs and making it difficult to guarantee calibration quality and stability. Existing automated calibration solutions still have technical shortcomings. They fail to achieve hierarchical standardized sampling that decouples position and attitude, cannot fundamentally solve the various defects caused by manual sampling, and are difficult to simultaneously ensure the safety, efficiency, and accuracy of calibration operations. Summary of the Invention

[0004] This invention provides a method, system, device, and storage medium for calibrating layered motion of a robot, so as to realize fully automatic layered motion hand-eye calibration of the robot and camera.

[0005] According to one aspect of the present invention, a method for calibrating layered motion of a robot is provided, the method comprising: Obtain robot teaching point pose data and calibration parameter data, and perform translational sampling and rotational sampling according to a preset uniformity rule based on the teaching point pose data and the calibration parameter data to determine effective sampled pose sequence data; Based on the effective sampled pose sequence data, the robot is controlled to move in layers to obtain calibration board image data and actual pose data. Based on the calibration board image data, the calibration board pose data is determined, and the calibration board pose data and the actual pose data are validated. With all sampled poses validated, the robot and camera hand-eye calibration is completed using a pre-built hand-eye calibration solution model based on the valid calibration board pose data and the valid actual pose data.

[0006] According to another aspect of the present invention, a calibration system for layered motion of a robot is provided, the system comprising: The pose sequence determination module is used to acquire robot teaching point pose data and calibration parameter data, and perform translation sampling and rotation sampling according to a preset uniformity rule based on the teaching point pose data and the calibration parameter data to determine the effective sampled pose sequence data. The actual pose acquisition module is used to control the robot's layered motion based on the effective sampled pose sequence data, and to acquire calibration board image data and actual pose data. The validity verification module is used to determine the calibration board pose data based on the calibration board image data, and to perform validity verification between the calibration board pose data and the actual pose data. The calibration module is used to perform hand-eye calibration between the robot and the camera based on the valid calibration board pose data and the valid actual pose data, using a pre-built hand-eye calibration solution model, after all sampled poses have been validated.

[0007] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: At least one processor; and memory that is communicatively connected to at least one processor; The memory stores a computer program that can be executed by at least one processor, which enables the at least one processor to perform the calibration method for robot layered motion according to any embodiment of the present invention.

[0008] According to another aspect of the present invention, a computer-readable storage medium is provided, which stores computer instructions for causing a processor to execute and implement a calibration method for robot layered motion according to any embodiment of the present invention.

[0009] The technical solution of this invention acquires robot teaching point pose data and calibration parameter data, and relies on a layered motion mechanism of translational and rotational independent sampling based on preset uniform sampling rules. This generates multiple sets of sampling points in an orderly manner, which are then filtered for accessibility and collision to obtain effective sampling pose sequence data. This layered, separated, and orderly automatic sampling method ensures the uniformity and orthogonality of calibration pose sampling, effectively solving the problems of chaotic, uneven, repetitive, and invalid point distribution in manual sampling, and significantly improving the consistency and standardization of samples from different batches and under different working conditions. Based on the effective sampling pose sequence data obtained from standardized layered sampling, the robot is controlled to move automatically and orderly according to layered logic, simultaneously acquiring calibration board image data and the robot's actual pose data. The entire process is automated, eliminating the need for repeated manual teaching and adjustments. This effectively solves the technical problems of traditional manual calibration operations being cumbersome, inefficient, requiring significant human intervention, and prone to collision risks and poor safety. Simultaneously, this invention performs a one-to-one validity check on the calibration board pose data and actual pose data corresponding to the same sampling pose point, accurately eliminating invalid data with abnormal field of view or imaging failure, ensuring that all data involved in the calibration solution are high-quality, valid paired data. This invention relies on a large amount of uniformly distributed, orthogonally dimensional, and reliable hierarchical sampling valid data, and uses a hand-eye calibration solution model for fitting and solving, effectively avoiding the defects of traditional disordered sampling, such as ill-conditioned calibration matrix, unstable solution convergence, and uncontrollable calibration accuracy, significantly improving the stability and calibration accuracy of the hand-eye transformation matrix solution. The overall solution achieves complete automation, hierarchicalization, ordering, and standardization of the robot hand-eye calibration process, greatly improving calibration efficiency and safety while significantly enhancing the uniformity of calibration pose distribution, data reliability, and the stability and universality of calibration results.

[0010] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 A flowchart illustrating a method for calibrating layered motion of a robot, as provided in an embodiment of the present invention; Figure 2aA flowchart illustrating another method for calibrating layered motion of a robot, provided in an embodiment of the present invention; Figure 2b A schematic diagram illustrating the effectiveness judgment of another robot layered motion calibration method provided in an embodiment of the present invention; Figure 2c A flowchart illustrating the calibration implementation of another method for calibrating layered motion of a robot, provided in an embodiment of the present invention. Figure 3 This is a schematic diagram of the structure of a robot layered motion calibration system provided in an embodiment of the present invention; Figure 4 A schematic diagram of the structure of an electronic device for implementing a calibration method for layered motion of a robot according to an embodiment of the present invention. Detailed Implementation

[0013] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0014] The key terms and their definitions used in this application are as follows: Layered robot motion refers to a standardized motion method that breaks down robot calibration and pose movement into two independent dimensions: pure translational motion and pure rotational motion, executed in layers. Translational motion only changes the robot's spatial position while maintaining its posture, and rotational motion only changes the robot's posture while maintaining its spatial position. This application achieves independent layered sampling through layered robot motion, using uniform sampling during translation and rotation. This avoids the defects of traditional manual mixed random sampling, such as uneven sampling points, poor sample consistency, and ill-conditioned calibration matrices, ensuring that hand-eye calibration sampling data is regular and reliable.

[0015] Effective sampled pose sequence data refers to a set of robot calibration poses that are uniformly distributed and safe to execute, generated according to a preset uniformity rule based on robot teaching point pose data and calibration parameter data, and after reachingability detection and collision risk filtering. This sequence contains layered and independent translational and rotational sampling points, providing a reliable motion reference for subsequent automated data acquisition.

[0016] Calibration board pose data refers to the position and orientation data of the calibration board in the camera coordinate system obtained by acquiring images of the calibration board through a camera and performing feature detection and pose calculation. It is one of the core input data for hand-eye calibration.

[0017] Effective actual pose data refers to the real motion pose data fed back by the robot control system after the robot reaches the preset sampling point, which corresponds one-to-one with the pose data of the calibration board. It is used to characterize the actual position and attitude of the robot in the robot coordinate system.

[0018] The hand-eye transformation matrix is ​​a 4×4 homogeneous transformation matrix that can characterize the fixed position and posture transformation relationship between the camera coordinate system and the robot base coordinate system. This matrix can be used to realize the mutual conversion between camera coordinates and robot coordinates, providing a coordinate transformation basis for robot vision-guided operations.

[0019] Preset uniformity rules refer to the point layout rules set in advance to achieve standardized sampling, including the three-axis translation equidistant point sampling rule and the three-axis rotation positive and negative angle uniform point sampling rule. These rules are used to ensure that the distribution of stratified sampling points is uniform, orthogonal, and without redundancy, and to avoid ill-conditioned calibration matrix problems caused by disordered sampling.

[0020] The calibration fitting accuracy data refers to the quantitative data obtained by statistical calculation based on the fitting residuals of all valid paired pose data. It is used to characterize the error level and reliability of the hand-eye transformation matrix solution and is an important basis for judging whether the calibration results are qualified.

[0021] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0022] Figure 1 This is a flowchart illustrating a robot layered motion calibration method provided in an embodiment of the present invention. This embodiment is applicable to industrial robots equipped with vision cameras (eyes outside the hand) for hand-eye coordinate relationship calibration. This method can be executed by a robot layered motion calibration system, which can be implemented in hardware and / or software and can be configured in an electronic device. Figure 1 As shown, the method specifically includes the following steps: S110. Obtain robot teaching point pose data and calibration parameter data, and perform translational sampling and rotational sampling according to a preset uniformity rule based on the teaching point pose data and the calibration parameter data to determine effective sampled pose sequence data.

[0023] The robot teaching point pose data includes initial safe pose data determined by manual teaching, calibration reference pose data, or sampled boundary-limited pose data. Calibration parameter data includes parameters required for the calibration process, such as height range, dimensions, number of layers, number of rows and columns, and rotation angles. Effective sampled pose sequence data includes a set of safe sampled poses after reachability and collision filtering. Preset uniformity rules include equidistant point sampling rules for three-axis translation, uniform point sampling rules for positive and negative angles of three-axis rotation, and zigzag uniform arrangement rules, used to ensure uniform, standardized, and non-redundant distribution of translational and rotational layered sampling points. Translational sampling includes uniform position sampling based on the reference pose in the robot's X, Y, and Z axes, maintaining the robot's posture during the sampling process. Rotational sampling includes uniform posture sampling based on the reference pose in the robot's RX, RY, and RZ axes, maintaining the robot's position during the sampling process.

[0024] Specifically, taking manual teaching as an example, the robot's initial pose data obtained through manual teaching, as well as the calibration parameter data configured by the user, are acquired; based on the initial pose and combined with the calibration parameters, sample poses are generated, and unreachable and collision-prone poses are filtered out, finally obtaining effective sample pose sequence data that can be directly executed; wherein, the preset calibration parameters can be preset according to experience, and this embodiment does not impose specific restrictions on them.

[0025] Optionally, based on the teaching point pose data and the calibration parameter data, translational sampling and rotational sampling are performed according to a preset uniformity rule to determine the effective sampled pose sequence data. This includes: performing translational sampling and rotational sampling according to a preset uniformity rule based on the teaching point pose data and the calibration parameter data to generate multiple calibration sampled pose data; performing reachability detection and collision risk filtering on the multiple calibration sampled pose data to obtain the effective sampled pose sequence data. The calibration sampled pose data includes original candidate sampled pose data generated based on the teaching point and parameters; reachability detection includes detecting whether the robot can move to the sampled pose; collision risk filtering includes eliminating sampled poses that may collide.

[0026] Specifically, based on the teaching point pose data and calibration parameter data, multiple uniformly distributed calibration sampling pose data are generated. Then, accessibility assessment and collision risk detection are performed on each of these candidate poses, eliminating poses that are inaccessible or pose a collision risk, ultimately obtaining a safe and usable valid sampling pose sequence data. The preset collision determination conditions can be pre-set based on experience; this embodiment does not impose specific restrictions on them. For example, the minimum safe distance threshold between robot links and surrounding obstacles, the joint motion limit angle threshold, and the interference avoidance distance threshold between adjacent mechanisms can be set as collision determination conditions.

[0027] For example, the robot's reference teaching point pose data obtained through manual teaching, along with preset calibration parameter data, are acquired. This scheme adopts a robot layered motion sampling mode, dividing the calibration sampling process into two independent stages: pure translational layered sampling and pure rotational layered sampling, performing standardized and uniform point sampling. In the pure translational sampling stage, points are sampled evenly at equal intervals along the X, Y, and Z axes, centered on the teaching reference point, changing only the position without altering the posture. In the pure rotational sampling stage, the reference position remains unchanged, and points are sampled at uniform angles around the positive and negative directions of the RX, RY, and RZ axes, changing only the posture without altering the position. After completing the layered uniform point sampling, accessibility and collision detection are performed on all sampled points to eliminate abnormal points, ultimately obtaining a uniformly distributed, compliant, and reliable valid sampled pose sequence data. The preset translational sampling interval, rotational sampling angle, and sampling quantity can be preset based on experience; this embodiment does not impose specific limitations on them.

[0028] S120. Based on the effective sampled pose sequence data, control the robot's layered motion to obtain calibration board image data and actual pose data.

[0029] The calibration board image data includes images captured by the camera that contain the concentric circle calibration board; the actual pose data includes the real pose data fed back by the robot after it arrives at the sampling point.

[0030] Specifically, the robot is controlled to move sequentially to each sampling position according to the effective sampling pose sequence; each time the robot reaches a point, the camera is controlled to capture the image data of the calibration board, and at the same time, the actual pose data fed back by the robot is obtained to complete a set of data collection.

[0031] Optionally, controlling the robot's layered movement based on the effective sampled pose sequence data to obtain calibration board image data and actual pose data includes: sending the effective sampled pose sequence data to the robot, controlling the robot to move sequentially to each sampled pose in the effective sampled pose sequence data; and when the robot reaches each sampled pose, acquiring calibration board image data captured by the camera and actual pose data fed back by the robot.

[0032] The sampling pose includes each point to be executed in the sequence.

[0033] Specifically, the effective sampled pose sequence data is sent to the robot control system via communication; the control system controls the robot to move sequentially to each sampled pose in the sequence; when the robot confirms that it has reached the current point, it triggers the camera to capture the calibration board image data and simultaneously reads the actual pose data fed back by the robot to complete the single-point data acquisition.

[0034] S130. Determine the calibration board pose data based on the calibration board image data, and perform a validity check on the calibration board pose data and the actual pose data.

[0035] The calibration board pose data includes the pose data of the calibration board in the camera coordinate system calculated from the image; the validity verification includes determining whether the calibration board is within the effective field of view.

[0036] Specifically, the calibration board image data is analyzed and processed using image recognition to calculate the calibration board pose data. The calibration board pose data is then compared with the robot's actual pose data for validity verification. If the calibration board is completely within the camera's effective field of view, unobstructed, and the image is recognizable, the sample corresponding to the current sampling pose is determined to be valid data. If the calibration board is outside the camera's field of view, partially obstructed, or the image is blurry and cannot be stably recognized, the sample corresponding to the current sampling pose is determined to be invalid data.

[0037] Optionally, the calibration board pose data is determined based on the calibration board image data, and the validity of the calibration board pose data and the actual pose data is verified, including: identifying the calibration board image data based on a pre-built concentric circle detection model to obtain the calibration board pose data; Verify whether the calibration board pose data is within the effective field of view of the camera. If the calibration board pose data is within the effective field of view of the camera, mark the calibration board pose data corresponding to the current sampled pose as valid calibration board pose data, and mark the actual pose data paired with the calibration board pose data as valid actual pose data.

[0038] The concentric circle detection model includes a detection model for identifying concentric circle calibration plates in an image, namely an image processing algorithm model constructed based on image edge extraction, contour filtering, and ring fitting rules; the effective field of view of the camera includes the area where the camera can clearly image and stably identify the calibration plate.

[0039] Specifically, a pre-built concentric circle detection model is invoked to acquire the collected calibration board image data as the input data source. The concentric circle detection model sequentially processes the calibration board image data through grayscale conversion, filtering and noise reduction, and edge feature extraction to filter out closed contour lines in the image. Then, according to the constraints of concentricity of inner and outer rings and the ratio of ring radii, concentric circle target contours that meet the characteristics of the calibration board are matched and filtered from all contours. Subsequently, nonlinear optimization is used to solve the parameters of the concentric circle contours, calculating feature parameters such as the center coordinates and ring radius. Based on the feature parameters, the corresponding calibration board pose data is calculated. Next, a validity check is performed to determine whether the overall contour of the calibration board falls completely within the camera's effective field of view. If the calibration board is unobstructed, has a complete contour, and is within the camera's effective field of view, it is considered valid data. This set of calibration board pose data and the corresponding actual pose data are retained and included in the subsequent calibration solution dataset for computation. If the calibration board exceeds the camera's field of view boundary, has local occlusion, or the image is blurred and cannot complete normal contour recognition, it is considered invalid data. The current sampled pose is marked separately, and this set of data is not included in the subsequent hand-eye calibration fitting calculation process; it is only recorded in the log to avoid abnormal data interfering with the calibration solution accuracy and matrix stability. Simultaneously, the algorithm performs another preprocessing step of sample selection. The core mechanism is to gradually eliminate samples with large errors through iterative optimization and error analysis. By calculating the reprojection error of each sample, it is determined whether the sample meets the error threshold requirements. This further avoids affecting the final calibration result. The preset concentricity deviation threshold, ring radius ratio threshold, and camera effective field of view range can be preset based on experience; this embodiment does not impose specific restrictions on them.

[0040] S140. After all sampled poses have been validated, the robot and camera are calibrated using a pre-built hand-eye calibration solution model based on the valid calibration board pose data and the valid actual pose data.

[0041] Among them, the effective calibration board pose data includes the calibration board pose data that has passed the verification; the effective actual pose data includes the robot's actual pose data corresponding to the effective calibration board; and the hand-eye calibration solution model includes a mathematical solution model used to calculate the coordinate transformation relationship between the robot and the camera.

[0042] Specifically, the hand-eye calibration solution model establishes the solution relationship based on the basic equations of hand-eye calibration in outdoor environments. It simultaneously models the transformation relationships between the camera coordinate system, calibration board coordinate system, robot base coordinate system, and robot end effector coordinate system. It adopts a hierarchical solution approach that decouples the position dimension and the attitude dimension, and iteratively fits and optimizes multiple pairs of effective calibration board pose data and effective actual pose data. It continuously corrects the transformation parameters and converges to the optimal solution. Finally, it calculates the hand-eye transformation matrix between the robot and the camera, and outputs calibration fitting accuracy data to evaluate the reliability of the calibration results.

[0043] For example, the hand-eye calibration solution model is expressed by the following formula: ; Where A represents the transformation matrix of the robot end effector relative to the robot base (end effect to base, obtained through robot kinematics), B is the transformation matrix of the calibration plate relative to the camera coordinate system (obtained from the image), X represents the transformation matrix from the camera coordinate system to the robot end effector coordinate system, and Y represents the coordinate system of the robot base relative to the calibration plate. For example, initialize variables and define A, B, X, and Y; process the data into eigen matrices, construct sample data, and store them in sample sets A and B respectively; set the descriptor, parameter type, and cost function type of the optimization problem, and initialize the optimization variables; call the algorithm and optimize X and Y using the Ceres solver, minimizing the error of AX=YB; convert the optimization results into matrix form, providing the solution results for X and Y respectively; after solving X and Y through calibration, combined with the transformation relationship between the robot arm and the camera, the transformation matrix from the camera coordinate system to the robot arm base can be directly calculated. If the eye is outside the hand, the calibration result is Y multiplied by the inverse of B; if the eye is on the hand, the calibration result is A multiplied by X. The hand-eye calibration solution model in this embodiment uses the fundamental matrix equation of hand-eye calibration as a constraint. It adopts a hierarchical iterative optimization method that decouples position and attitude to fit and solve multiple sets of paired pose data. After convergence, it outputs the optimal hand-eye transformation matrix and calibration fitting accuracy data to realize the solution and calibration of the relative position and attitude of the robot and the camera.

[0044] Specifically, after all sampled poses have been collected and verified, all valid calibration board pose data and valid actual pose data are input into the hand-eye calibration solution model; the coordinate transformation relationship is obtained through model calculation, and finally the hand-eye calibration between the robot and the camera is completed.

[0045] The technical solution of this invention acquires robot teaching point pose data and calibration parameter data. Based on the teaching point pose data and calibration parameter data, translational sampling and rotational sampling are performed according to a preset uniformity rule to determine effective sampled pose sequence data. Based on the effective sampled pose sequence data, the robot is controlled to perform layered motion to acquire calibration board image data and actual pose data. Based on the calibration board image data, calibration board pose data is determined, and validity verification is performed on the calibration board pose data and actual pose data. After all sampled poses have been validated, based on the effective calibration board pose data and effective actual pose data, a pre-built hand-eye calibration solution model is used to complete the hand-eye calibration between the robot and the camera. This solves the technical problems of low efficiency, poor safety, and uncontrollable calibration accuracy in traditional manual calibration, achieving the technical effects of improving calibration efficiency, reducing collision risk, improving pose distribution uniformity, and enhancing calibration stability and accuracy.

[0046] Figure 2 is a flowchart of another robot layered motion calibration method provided by an embodiment of the present invention. Based on the above embodiments, this embodiment further refines how to complete the robot and camera hand-eye calibration based on the effective calibration board pose data and the effective actual pose data using a pre-built hand-eye calibration solution model. For specific implementation details, please refer to the technical solution of this embodiment. Technical terms that are the same as or corresponding to those in the above embodiments will not be repeated here.

[0047] As shown in Figure 2, the method specifically includes the following steps: S1401. Determine the effective calibration plate pose data and the effective actual pose data corresponding to the same sampling pose point as a set of effective pose data, and input multiple sets of effective pose data into the hand-eye calibration solution model.

[0048] The hand-eye calibration solution model is a solution model constructed based on the position and attitude decoupling optimization algorithm.

[0049] Among them, the position and pose decoupling optimization algorithm includes an algorithm that separates and independently optimizes the robot's position and pose dimensions; the hand-eye calibration solution model includes a dedicated mathematical calculation model for solving the coordinate transformation relationship between the robot and the camera; and multiple sets of pose data include multiple sets of paired effective calibration plate pose data and effective actual pose data.

[0050] Specifically, multiple sets of pose data after screening and verification are uniformly input into the hand-eye calibration solution model, which is built on the position and pose decoupling optimization algorithm architecture. The input data is sent into the model input end in pairs to provide a complete data source for subsequent hierarchical fitting calculations. Only valid data that has passed verification is included in the calculation, and invalid abnormal data is removed to ensure the regularity and validity of the input data source.

[0051] S1402. By using the hand-eye calibration solution model, perform hierarchical fitting calculations on multiple sets of effective pose data to obtain the hand-eye transformation matrix and calibration fitting accuracy data between the robot and the camera.

[0052] The hierarchical fitting calculation includes a step-by-step iterative fitting and optimization convergence calculation method based on the position layer and the attitude layer; the hand-eye transformation matrix includes matrix data that characterizes the position and attitude transformation relationship between the camera coordinate system and the robot base coordinate system; the calibration fitting accuracy data includes quantitative evaluation data that measures the magnitude of the hand-eye calibration solution error and the quality of the fit.

[0053] Specifically, after importing multiple sets of paired and matched effective calibration board pose data and multiple sets of effective actual pose data into the hand-eye calibration solution model, the model uses the classical hand-eye calibration matrix equation as a basic constraint and employs a hierarchical processing logic that decouples position and attitude to perform fitting calculations. First, the position dimension parameters in multiple sets of data are individually fitted iteratively to solve for the optimal transformation parameters of the position dimension; then, the attitude dimension parameters are independently fitted and converged hierarchically to solve for the optimal transformation parameters of the attitude dimension. By solving for position and attitude step by step and iteratively fitting hierarchically, the matrix ill-conditioning and solution divergence problems that are prone to occur in the overall simultaneous solution are avoided, and the optimal transformation relationship is gradually approximated iteratively. After the model completes the hierarchical fitting calculation convergence, the solution outputs a complete hand-eye transformation matrix that represents the coordinate transformation relationship between the camera and the robot base; at the same time, based on the statistical calculation of the fitting residuals of multiple sets of samples, calibration fitting accuracy data that quantifies the magnitude of the calibration error and the degree of fitting quality is generated. Among them, the preset iteration convergence threshold and fitting residual judgment threshold can be preset according to experience, and this embodiment does not impose specific restrictions on them.

[0054] Preferably, the multiple sets of pose data involved in the calculation consist of multiple sets of valid calibration board pose data and multiple sets of valid actual pose data, all matched in a one-to-one correspondence. Each time the robot moves to a sampling pose point, a unique set of valid calibration board pose data is generated. The two sets of data form a strict pairing association according to the sampling point. Only all verified, one-to-one matched pose data are fed into the model for calculation. After receiving the one-to-one matched dataset, the hand-eye calibration solution model performs hierarchical fitting calculations according to the position and pose decoupling optimization approach. The pose data is split into position and pose dimensions for separate processing, and hierarchical iterative fitting and parameter optimization are completed sequentially. Using multiple sets of paired data to jointly constrain the solution relationship, the transformation relationship between the robot coordinate system and the camera coordinate system is calculated after iterative convergence, generating the corresponding hand-eye transformation matrix. The model performs statistical analysis on the fitting residuals of all paired pose data, quantifies the calibration error level, and generates calibration fitting accuracy data for evaluating the quality of the calibration results.

[0055] Optionally, after obtaining the hand-eye transformation matrix and calibration fitting accuracy data between the robot and the camera, the method further includes: evaluating the calibration quality of the hand-eye transformation matrix based on the calibration fitting accuracy data; and confirming the completion of the hand-eye calibration between the robot and the camera if the calibration quality evaluation results meet preset conditions.

[0056] The calibration quality assessment includes determining whether the calibration results are reliable based on the fitting accuracy; the preset conditions include a set of rules that must be met to determine whether the calibration is qualified.

[0057] Specifically, after obtaining the hand-eye transformation matrix and calibration fitting accuracy data, the calibration fitting accuracy data is used to evaluate the quality of the hand-eye transformation matrix; it is then determined whether the evaluation results meet the preset qualification conditions; if all conditions are met, the calibration is confirmed to be valid and the calibration process is completed; if not, the parameters can be adjusted and recalibrated.

[0058] Optional preset conditions include: the calibration fitting accuracy data is not greater than the preset accuracy threshold; the number of sets of effective calibration board pose data and effective actual pose data is not less than the preset minimum sample size; and the hand-eye transformation matrix is ​​a non-ill-conditioned legal matrix.

[0059] Among them, the preset accuracy threshold includes the maximum allowable calibration error; the preset minimum number of samples includes the minimum number of valid data sets required to ensure calibration stability; and the non-illness valid matrix includes a numerically stable and usable transformation matrix.

[0060] Specifically, the calibration quality assessment must simultaneously meet three preset conditions: First, the calibration fitting accuracy data is less than or equal to the preset accuracy threshold; second, the number of sets of effective calibration board pose data and effective actual pose data is not less than the preset minimum sample size; third, the solved hand-eye transformation matrix is ​​a non-ill-conditioned legal matrix; the calibration is deemed qualified when all three conditions are met; among them, the preset accuracy threshold and the preset minimum sample size can be preset based on experience, and this embodiment does not impose specific restrictions on them.

[0061] S1403. Based on the hand-eye transformation matrix and calibration fitting accuracy data, complete the hand-eye calibration between the robot and the camera.

[0062] Specifically, after obtaining the hand-eye transformation matrix and the calibration fitting accuracy data, the two are combined as the judgment and benchmark for calibration completion. The hand-eye transformation matrix establishes a fixed position and posture transformation mapping relationship between the camera coordinate system and the robot base coordinate system, enabling mutual conversion from visual image coordinates to the robot's actual motion coordinates. The calibration fitting accuracy data quantitatively reflects the fitting error and reliability of this matrix solution. The system uses the hand-eye transformation matrix as the core benchmark for coordinate transformation, and combines it with the calibration fitting accuracy data to comprehensively verify and judge the quality of the calibration results. Under the premise that the accuracy data is within a reasonable and qualified range, the current hand-eye transformation matrix is ​​recognized as valid, and the matrix parameters are officially fixed and saved, completing the overall hand-eye calibration between the robot and the camera, providing a reliable coordinate transformation basis for subsequent robot visual grasping, visual guided motion, and other operations. The preset accuracy qualification threshold can be pre-set based on experience, and this embodiment does not impose specific restrictions on it.

[0063] Figure 2b This is a schematic diagram illustrating the effectiveness judgment of another robot layered motion calibration method provided in an embodiment of the present invention; as shown. Figure 2b As shown in the figure, the camera's 2D field of view, the effective area of ​​the calibration point, and two calibration point pose states are illustrated to explain the judgment logic for validating the sampling pose in this invention. The yellow area represents the camera's complete 2D field of view, and the light gray area represents the preset effective imaging area of ​​the calibration point. This area is set to 40%-60% of the camera's 2D field of view to avoid feature distortion and decreased calculation accuracy caused by the calibration plate being at the edge of the field of view. The left side shows the effective calibration point state: the concentric circle calibration plate is completely within the preset effective imaging area, the image is complete and without edge truncation, and feature detection and pose calculation can be performed normally, thus it is determined to be a valid data point. The right side shows the invalid calibration point state: part of the concentric circle calibration plate is outside the effective imaging area, there is edge truncation or incomplete imaging, the feature calculation accuracy cannot be guaranteed, it is determined to be an invalid data point and will not participate in subsequent hand-eye calibration solutions.

[0064] Figure 2c A flowchart illustrating the calibration implementation of another robot layered motion calibration method provided in this embodiment of the invention; as shown. Figure 2c As shown, the method includes: S1: Initial point for manual teaching.

[0065] Specifically, the operator holds a teach pendant and records the first collision-free joint angle position P1. The robot then moves to P1 using MoveJ.

[0066] S2: Establish a communication connection.

[0067] Specifically, the robot control terminal establishes communication with the calibration software via Socket and continuously sends the current joint angle, flange pose, TCP parameters, and robot status.

[0068] S3: Image acquisition and visual inspection.

[0069] Specifically, the camera takes a picture of the calibration board and calls an ellipse detection or concentric circle detection algorithm to determine whether the calibration board is located in the center of the field of view; If the detection fails, the system will issue a warning and proceed with a retry or filter the points. If the detection is successful, record the current pose of the calibration plate and the pose of the robotic arm.

[0070] S4: Automatic planning and deployment of multi-pose.

[0071] Specifically, the system automatically generates N robot poses based on the hand-eye calibration scheme (Scheme B), sends them to the robotic arm for execution in sequence, and waits for feedback.

[0072] Optionally, the inverse kinematics solution can be obtained by calling the robotic arm kinematics library to pre-filter unreachable poses.

[0073] S5: Repeated sampling process.

[0074] Specifically, repeat steps S4 to S5 for each set of postures until N sets of calibration plate posture and robotic arm posture data are collected.

[0075] S6: Calibration, solution, and result output.

[0076] Specifically, samples with large errors are removed, and a certain number of pure rotation and pure translation samples are ensured, for example, more than 3 samples each. The N sets of posture data are then input into the automatic hand-eye calibration solver, and the calibration results are output and displayed.

[0077] S7: Process Diagnosis and Abnormal Handling.

[0078] S7.1 Analysis of reprojection error results.

[0079] Specifically, the reprojection error of the calibration solution is calculated, the error distribution of each set of sampled data is analyzed, and samples with out-of-tolerance errors are screened and labeled.

[0080] S7.2 Valid Sample Guidelines.

[0081] Specifically, based on the reprojection error analysis results, the system automatically removes invalid or low-quality samples and guides operators to collect additional valid attitude data to improve calibration accuracy.

[0082] S7.3 Camera exposure adaptive adjustment.

[0083] Specifically, during the sampling process, the image quality of the calibration board is monitored in real time. If the edges of the concentric circles are blurred or the contrast is insufficient, the system automatically adjusts the camera exposure parameters and re-acquires the image until the detection requirements are met.

[0084] S7.4 Visual Analysis.

[0085] Specifically, the current robotic arm model data is read, and the point cloud of the robotic arm body collected by the camera is obtained. The two are then superimposed and displayed in the same coordinate system for consistency visualization analysis.

[0086] S8: End process.

[0087] Specifically, the software sends out a termination signal to complete the calibration process.

[0088] The technical solution of this invention solves the technical problems of traditional solution methods, such as matrix ill-conditioning, inability to adapt to separately sampled position and pose data, and unstable accuracy due to overfitting, by using a position and pose decoupling optimization algorithm to construct a solution model and hierarchically fitting multiple sets of pose data to solve the hand-eye transformation matrix. It achieves the beneficial effects of adapting to hierarchical sampling data, improving solution stability, reducing overfitting risk, and ensuring reliable calibration accuracy.

[0089] Figure 3 This is a schematic diagram of a robot layered motion calibration system provided in an embodiment of the present invention. Figure 3 As shown, the system includes: a pose sequence determination module 310, an actual pose acquisition module 320, a validity verification module 330, and a calibration module 340.

[0090] The module 310 is used to acquire robot teaching point pose data and calibration parameter data, and to perform translational sampling and rotational sampling according to a preset uniformity rule based on the teaching point pose data and the calibration parameter data to determine the effective sampled pose sequence data; the actual pose acquisition module 320 is used to control the robot's layered motion based on the effective sampled pose sequence data to acquire calibration board image data and actual pose data; the validity verification module 330 is used to determine calibration board pose data based on calibration board image data and to perform validity verification on the calibration board pose data and the actual pose data; the calibration module 340 is used to complete the hand-eye calibration of the robot and the camera based on the effective calibration board pose data and the effective actual pose data, using a pre-built hand-eye calibration solution model, after all sampled poses have been verified for validity.

[0091] The technical solution of this invention acquires robot teaching point pose data and calibration parameter data. Based on the teaching point pose data and calibration parameter data, translational sampling and rotational sampling are performed according to a preset uniformity rule to determine effective sampled pose sequence data. Based on the effective sampled pose sequence data, the robot is controlled to perform layered motion to acquire calibration board image data and actual pose data. Based on the calibration board image data, calibration board pose data is determined, and validity verification is performed on the calibration board pose data and actual pose data. After all sampled poses have been validated, based on the effective calibration board pose data and effective actual pose data, a pre-built hand-eye calibration solution model is used to complete the hand-eye calibration between the robot and the camera. This solves the technical problems of low efficiency, poor safety, and uncontrollable calibration accuracy in traditional manual calibration, achieving the technical effects of improving calibration efficiency, reducing collision risk, improving pose distribution uniformity, and enhancing calibration stability and accuracy.

[0092] Optionally, the pose sequence determination module 310 includes: The sampling pose generation unit is used to generate multiple calibration sampling pose data by performing translational sampling and rotational sampling according to a preset uniformity rule based on the teaching point pose data and calibration parameter data. The sequence generation unit is used to perform reachability detection and collision risk filtering on multiple calibrated sampled pose data to obtain effective sampled pose sequence data.

[0093] Optionally, the actual pose acquisition module 320 includes: The motion control unit is used to send the effective sampled pose sequence data to the robot and control the robot to move sequentially to each sampled pose in the effective sampled pose sequence data. The actual pose acquisition unit is used to acquire the calibration board image data captured by the camera and the actual pose data fed back by the robot when the robot reaches each sampling pose.

[0094] Optionally, the validity verification module 330 includes: The image recognition unit is used to recognize the calibration board image data based on the pre-built concentric circle detection model to obtain the calibration board pose data. The verification unit is used to verify whether the calibration board pose data is within the effective field of view of the camera. If the calibration board pose data is within the effective field of view of the camera, the calibration board pose data corresponding to the current sampled pose is marked as valid calibration board pose data, and the actual pose data paired with the calibration board pose data is marked as valid actual pose data.

[0095] Optionally, the hand-eye calibration solution model is a solution model constructed based on a position and attitude decoupling optimization algorithm; correspondingly, the calibration module 340 includes: The model input unit is used to determine the effective calibration plate pose data and the effective actual pose data corresponding to the same sampling pose point as a set of effective pose data, and input multiple sets of effective pose data into the hand-eye calibration solution model; The hierarchical fitting unit is used to perform hierarchical fitting calculations on multiple sets of effective pose data through the hand-eye calibration solution model, so as to obtain the hand-eye transformation matrix between the robot and the camera and the calibration fitting accuracy data. The calibration unit is used to perform hand-eye calibration between the robot and the camera based on the hand-eye transformation matrix and calibration fitting accuracy data.

[0096] Optionally, the calibration module 340 also includes: The calibration quality assessment unit is used to assess the calibration quality of the hand-eye transformation matrix based on the calibration fitting accuracy data after obtaining the hand-eye transformation matrix between the robot and the camera and the calibration fitting accuracy data. The hand-eye calibration unit is used to confirm the completion of hand-eye calibration between the robot and the camera when the calibration quality assessment results meet the preset conditions.

[0097] Optional, preset conditions include: The calibration fitting accuracy data should not exceed the preset accuracy threshold. The number of sets of valid calibration plate pose data and valid actual pose data is not less than the preset minimum sample size; The hand-eye transformation matrix is ​​a non-illness-prone legal matrix.

[0098] The robot layered motion calibration system provided in this embodiment of the invention can execute the robot layered motion calibration method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0099] Figure 4 This is a schematic diagram of an electronic device for implementing the robot layered motion calibration method according to embodiments of the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0100] like Figure 4As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0101] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0102] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the calibration of layered motion of a robot.

[0103] In some embodiments, the calibration of the method robot's layered motion can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or mounted on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the calibration of the method robot's layered motion described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform the calibration of the method robot's layered motion by any other suitable means (e.g., by means of firmware).

[0104] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include: implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0105] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0106] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0107] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0108] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0109] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0110] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0111] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for calibrating layered motion of a robot, characterized in that, include: Obtain robot teaching point pose data and calibration parameter data, and perform translational sampling and rotational sampling according to a preset uniformity rule based on the teaching point pose data and the calibration parameter data to determine effective sampled pose sequence data; Based on the effective sampled pose sequence data, the robot is controlled to move in layers to obtain calibration board image data and actual pose data. Based on the calibration board image data, the calibration board pose data is determined, and the calibration board pose data and the actual pose data are validated. With all sampled poses validated, the robot and camera hand-eye calibration is completed using a pre-built hand-eye calibration solution model based on the valid calibration board pose data and the valid actual pose data.

2. The method according to claim 1, characterized in that, The step of determining effective sampled pose sequence data by performing translational and rotational sampling based on the teaching point pose data and the calibration parameter data according to a preset uniformity rule includes: Based on the teaching point pose data and the calibration parameter data, translation sampling and rotation sampling are performed according to a preset uniformity rule to generate multiple calibration sampling pose data. Accessibility detection and collision risk filtering are performed on multiple calibrated sampled pose data to obtain the effective sampled pose sequence data.

3. The method according to claim 1, characterized in that, The process of controlling the robot's layered motion based on the effective sampled pose sequence data, and acquiring calibration board image data and actual pose data, includes: The effective sampled pose sequence data is sent to the robot, and the robot is controlled to move sequentially to each sampled pose in the effective sampled pose sequence data; When the robot reaches each sampling pose, it acquires the calibration board image data captured by the camera and the actual pose data fed back by the robot.

4. The method according to claim 1, characterized in that, The step of determining the calibration board pose data based on the calibration board image data and performing a validity check on the calibration board pose data and the actual pose data includes: The calibration board image data is identified based on a pre-built concentric circle detection model to obtain the calibration board pose data; Verify whether the calibration board pose data is within the effective field of view of the camera. If the calibration board pose data is within the effective field of view of the camera, mark the calibration board pose data corresponding to the current sampled pose as valid calibration board pose data, and mark the actual pose data paired with the calibration board pose data as valid actual pose data.

5. The method according to claim 1, characterized in that, The hand-eye calibration solution model is a solution model constructed based on a position and pose decoupling optimization algorithm; the hand-eye calibration of the robot and camera is completed using the pre-constructed hand-eye calibration solution model based on the effective calibration board pose data and the effective actual pose data, including: The effective calibration plate pose data and the effective actual pose data corresponding to the same sampling pose point are determined as a set of effective pose data, and multiple sets of the effective pose data are input into the hand-eye calibration solution model. The hand-eye calibration solution model is used to perform hierarchical fitting calculations on multiple sets of effective pose data to obtain the hand-eye transformation matrix and calibration fitting accuracy data between the robot and the camera. The hand-eye calibration between the robot and the camera is completed based on the hand-eye transformation matrix and the calibration fitting accuracy data.

6. The method according to claim 5, characterized in that, After obtaining the hand-eye transformation matrix and calibration fitting accuracy data between the robot and the camera, the process further includes: The calibration quality of the hand-eye transformation matrix is ​​evaluated based on the calibration fitting accuracy data. If the calibration quality assessment results meet the preset conditions, the hand-eye calibration of the robot and the camera is confirmed to be complete.

7. The method according to claim 6, characterized in that, The preset conditions include: The calibration fitting accuracy data is not greater than a preset accuracy threshold. The number of sets of the effective calibration plate pose data and the effective actual pose data is not less than the preset minimum sample size; The hand-eye transformation matrix is ​​a non-illness-prone legal matrix.

8. A calibration system for layered motion of a robot, characterized in that, include: The pose sequence determination module is used to acquire robot teaching point pose data and calibration parameter data, and perform translation sampling and rotation sampling according to a preset uniformity rule based on the teaching point pose data and the calibration parameter data to determine the effective sampled pose sequence data. The actual pose acquisition module is used to control the robot's layered motion based on the effective sampled pose sequence data, and to acquire calibration board image data and actual pose data. The validity verification module is used to determine the calibration board pose data based on the calibration board image data, and to perform validity verification between the calibration board pose data and the actual pose data. The calibration module is used to perform hand-eye calibration between the robot and the camera based on the valid calibration board pose data and the valid actual pose data, using a pre-built hand-eye calibration solution model, after all sampled poses have been validated.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and a memory communicatively connected to the at least one processor; The memory stores a computer program that can be executed by the at least one processor, which is then executed by the at least one processor to enable the at least one processor to perform the calibration method for robot layered motion according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed by a processor, implement the calibration method for the layered motion of the robot according to any one of claims 1-7.

Images