A method and apparatus for compensating microscope image-stage coordinates
By using a microscope image-stage coordinate compensation method, single-axis and dual-axis errors are separated, and an error model is constructed. This solves the problem of insufficient explanation of stage motion errors and improves the positioning accuracy and operational quality of the microscope imaging equipment.
Patent Information
- Application Number
- CN202611293389.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-25
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2046-08-25
AI Technical Summary
In existing microscopic imaging systems, the motion error of the stage is difficult to explain effectively, especially during continuous dual-axis motion. Current technology cannot clearly distinguish between single-axis error and dual-axis coupling error, resulting in insufficient positioning accuracy and incomplete error model, which makes it impossible to achieve high-precision target positioning and area revisiting.
By employing a microscope image-stage coordinate compensation method, and utilizing the differences in marker point positions and image increments, single-axis positioning errors, dual-axis coupling errors, and motion sequence-related errors are separated. An error model is constructed to achieve precise compensation of stage command coordinates. This method is applicable to pathological slide scanners, cytology scanners, fluorescence microscopes, material microscopic inspection equipment, and other microscopic imaging systems.
It significantly improves the positioning accuracy of the stage and the operational quality of the imaging equipment, ensures the accuracy and integrity of the error model, achieves precise decoupling of the stage motion error, and enhances the stability and positioning accuracy of the microscopic imaging equipment in high-precision scenarios.
Smart Images

Figure CN122836983B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to geometric calibration and precision motion compensation technology for microscopic imaging equipment, and particularly to a method and apparatus for compensating microscope image-stage coordinates. Background Technology
[0002] In modern microscopic imaging systems, in order to achieve high-precision target localization, area revisiting, and local rescanning, a static mapping between the microscope image coordinates and the physical coordinates of the stage is usually established using a calibration plate, and the stage movement is controlled based on this mapping relationship.
[0003] However, in practical applications, the motion accuracy of the stage is constrained by a variety of complex factors, and its motion trajectory is not ideal. Specifically, the actual motion error of the stage mainly comes from the nonlinearity of the single-axis scale, the hysteresis caused by changes in the direction of motion, the inter-axis load coupling during dual-axis motion, the insufficient stability of the system after the previous axis motion, and the specific historical influence of the motion sequence.
[0004] While existing conventional single-axis calibration techniques can describe the inherent errors of each axis during independent motion, they lack effective explanatory power for the composite errors generated by continuous biaxial motion. Although directly comparing the endpoint differences between paths X followed by Y and Y followed by X can reveal some phenomena caused by the different motion order, this method has inherent limitations and cannot clearly answer the following key questions: In the total endpoint error of continuous biaxial motion, which parts can be explained by the superposition of the errors of the two single axes, and which parts are additional errors generated by the continuous combination of the two axes?
[0005] Furthermore, directly subtracting the original endpoint residuals of the two combined paths will cause common coupling errors, which are independent of the axis sequence but exist in both paths, to cancel each other out, thus losing important error information. On the other hand, simply adding the single-axis residuals obtained at different intermediate positions will ignore the local scale, rotation and projection response of the image coordinates to the physical displacement, resulting in inaccurate baseline construction. Summary of the Invention
[0006] The purpose of this invention is to provide a microscope image-stage coordinate compensation method and device, which is based on the separation of single-axis superposition reference and dual-axis combination excess residual based on a common endpoint. By utilizing the position of marker points in the microscope image, the single-axis positioning error, dual-axis common coupling error and motion sequence related error of the two-dimensional or multi-axis motorized stage are separated, and the stage command coordinates corresponding to the image target are corrected accordingly. The method and system are applicable to pathological slide scanners, cytology scanners, fluorescence microscopes, material microscopic inspection equipment, semiconductor microscopic inspection equipment and other microscopic imaging systems with motorized stages.
[0007] To achieve the above objectives, this technical solution provides a microscope image-stage coordinate compensation method, including the following steps: S1: Obtain the basic static coordinate mapping of the microscopic image and the stage; S2: For the same four-point structure on the working area of the stage, execute the first continuous combination motion of the first axis followed by the second axis and the second continuous combination motion of the second axis followed by the first axis respectively, and obtain the standardized single-axis motion that matches each motion segment of the two continuous combination motions respectively. S3: Obtain the order-preserving and flipping components under different calibration conditions. For each calibration condition, obtain the difference between the actual image increment of the marker point in each standardized single-axis motion and the predicted image increment based on the basic static mapping to obtain the corresponding single-axis residual. Based on the local coordinate response of the basic static mapping between the obtained position of the single-axis residual and the common endpoint, convert each single-axis residual into a common endpoint equivalent residual. Based on the common endpoint equivalent residual matched with the motion segments of each continuous combined motion, construct the corresponding single-axis superposition reference. Obtain the measured endpoint residuals of the two continuous combined motions respectively, and subtract the measured endpoint residuals from the corresponding single-axis superposition references to obtain the combined excess residual. Determine the order-preserving and order-flipping components based on the preservation and flipping relationships of the combined excess residuals corresponding to the two continuous combined motions under axis order exchange. S4: Construct an error model based on calibration data obtained under different calibration conditions, where the calibration data includes single-axis residuals, order-preserving components, and order-reversing components; S5: Based on the target position and the planned motion path, query the single-axis residual, order-keeping component, and order-flipping component from the self-error model. Determine the direction of the order-flipping component according to the planned motion path. Use the single-axis residual, order-keeping component, and order-flipping component to determine the stage command compensation amount. Determine the stage command coordinates based on the stage command compensation amount and the target position.
[0008] Compared with existing technologies, this technical solution has the following characteristics and beneficial effects: This scheme achieves precise decoupling of complex motion errors of the stage by using a single-axis superposition reference with a common endpoint and a dual-axis combined excess residual separation technique. It can clearly divide the total error into single-axis positioning error, a sequence-preserving component representing the dual-axis common coupling, and a sequence-reversing component representing the influence of motion history. This effectively answers the question of which continuous dual-axis motion errors are explained by single-axis superposition and which are additionally generated, providing a foundation for accurate identification of error sources. Compared with existing technologies, this scheme significantly improves the accuracy and completeness of error modeling. By converting single-axis residuals at different locations into equivalent residuals at a common endpoint, it overcomes the defect of directly superimposing and ignoring local coordinate responses. Simultaneously, by utilizing the decomposition principle under axis sequence exchange, it avoids the problem of axis sequence-independent common coupling errors being canceled out by direct subtraction, ensuring that the error model retains all key physical characteristics and avoids the loss of key error information. Ultimately, it achieves adaptive dynamic compensation based on the planned motion path, significantly improving the positioning accuracy of the stage. The system can dynamically query and combine single-axis residuals, order-preserving components, and order-flipping components with the correct signs based on the actual sequence and direction of motion axes, thereby generating optimal stage command coordinates. Combined with an independent validation set and a component update mechanism, this scheme can effectively correct positioning deviations in tasks such as target centering and region revisiting, significantly improving the operational quality and stability of microscopic imaging equipment in high-precision scenarios. Attached Figure Description
[0009] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of a microscope image-stage coordinate compensation method according to an embodiment of this application; Figure 2 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of this application. Detailed Implementation
[0010] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with one or more embodiments of this specification. Rather, they are merely examples of apparatuses and methods consistent with some aspects of one or more embodiments of this specification as detailed in the appended claims.
[0011] It should be noted that the steps of the corresponding methods are not necessarily performed in the order shown and described in this specification in other embodiments. In some other embodiments, the methods may include more or fewer steps than described in this specification. Furthermore, a single step described in this specification may be broken down into multiple steps in other embodiments; and multiple steps described in this specification may be combined into a single step in other embodiments.
[0012] Example 1 like Figure 1 As shown, this solution provides a microscope image-stage coordinate compensation method, including the following steps: S1: Obtain the basic static coordinate mapping of the microscopic image and the stage; S2: For the same four-point structure on the working area of the stage, execute the first continuous combination motion of the first axis followed by the second axis and the second continuous combination motion of the second axis followed by the first axis respectively, and obtain the standardized single-axis motion that matches each motion segment of the two continuous combination motions respectively. S3: Obtain the order-preserving and flipping components under different calibration conditions. For each calibration condition, obtain the difference between the actual image increment of the marker point in each standardized single-axis motion and the predicted image increment based on the basic static mapping to obtain the corresponding single-axis residual. Based on the local coordinate response of the basic static mapping between the obtained position of the single-axis residual and the common endpoint, convert each single-axis residual into a common endpoint equivalent residual. Based on the common endpoint equivalent residual matched with the motion segments of each continuous combined motion, construct the corresponding single-axis superposition reference. Obtain the measured endpoint residuals of the two continuous combined motions respectively, and subtract the measured endpoint residuals from the corresponding single-axis superposition references to obtain the combined excess residual. Determine the order-preserving and order-flipping components based on the preservation and flipping relationships of the combined excess residuals corresponding to the two continuous combined motions under axis order exchange. S4: Construct an error model based on calibration data obtained under different calibration conditions, where the calibration data includes single-axis residuals, order-preserving components, and order-reversing components; S5: Based on the target position and the planned motion path, query the single-axis residual, order-keeping component, and order-flipping component from the self-error model. Determine the direction of the order-flipping component according to the planned motion path. Use the single-axis residual, order-keeping component, and order-flipping component to determine the stage command compensation amount. Determine the stage command coordinates based on the stage command compensation amount and the target position.
[0013] Regarding step S1: The microscope image-stage coordinate compensation method proposed in this scheme is applicable to microscopic imaging devices with X-axis and Y-axis stages, wherein the microscopic imaging device can be a pathological slide scanner, a cytology scanner, a fluorescence microscope, a material microscopic inspection device, or other microscopic imaging devices with a motorized stage.
[0014] In some embodiments, the basic static mapping is used to give the nominal response of the physical displacement of the stage in the image coordinates of the microscope, which can be expressed as F(m). The basic static mapping can be obtained by affine mapping, homography mapping, local mapping or other mapping methods that can describe the correspondence between image coordinates and physical coordinates of the stage.
[0015] In order to establish the calibration relationship between the microscope image and the stage, this scheme further obtains the physical marker positions and the microscopic image marker positions. The physical marker positions are the positions of the physical marker points on the stage, and the microscopic image marker positions are the positions of the physical marker points on the microscopic image. Correspondingly, multiple physical marker points are set in the working area of the stage of the microscopic imaging device, and the microscopic images corresponding to the physical marker points are obtained through the microscopic imaging device.
[0016] In some embodiments, physical markers are used to establish the correspondence between the physical coordinates of the stage and the coordinates of the microscopic image. The physical markers may be checkerboard corner points, dots, cross marks, coded marks, speckle-related points, fluorescent microspheres, wafer marks, or natural tissue features in the microscopic image.
[0017] For the i-th physical marker point, its physical marker position on the platform is represented as: m i =(x i ,y i ) Where x i ,y i Let x and y be the x and y coordinates of the i-th physical marker point; Correspondingly, the location of the microscopic image marker is represented as u i .
[0018] In some embodiments, the markers may be checkerboard corners, dots, crosses, coded markers, speckle-related dots, fluorescent microspheres, wafer markers, or natural tissue features.
[0019] In some embodiments, the basic static mapping records the mapping relationship between the physical marker position and the microscopic image marker position of the same physical marker point. That is, the basic static mapping is established based on the physical coordinates of the stage and the corresponding microscopic image coordinates of multiple physical marker points. The basic static mapping can be used to describe the microscopic image coordinate relationship corresponding to the change of the physical coordinates of the stage under ideal static conditions.
[0020] Furthermore, step S1 of this solution also acquires the stage motion record, wherein the stage motion record includes at least the motion start point, motion end point, motion axis, motion direction, motion distance, and encoder position. In some embodiments, the stage motion record further includes motion speed, acceleration, and intermediate waiting time. Regarding step S2: This scheme sets up a four-point structure in the working area of the stage, consisting of a common starting point, the middle position of the first axis, the middle position of the second axis, and a common ending point, and executes motion to obtain a first continuous combination motion, a second continuous combination motion, and a standardized single-axis motion.
[0021] In some embodiments, the first axis midpoint position refers to the midpoint obtained after only performing the first axis movement from the common starting point, and the second axis midpoint position refers to the midpoint obtained after only performing the second axis movement from the common starting point, wherein the first axis is different from the second axis.
[0022] In some embodiments, the first continuous combined motion and the second continuous combined motion have the same common starting point, common ending point, and the same command displacement. The only difference is the execution order of the two continuous combined motions. That is, the first continuous combined motion is a continuous combined motion in which the first axis motion is executed first and then the second axis motion is executed, and the second continuous combined motion is a continuous combined motion in which the second axis motion is executed first and then the first axis motion is executed.
[0023] In some embodiments, standardized single-axis motion is a motion segment that executes only one motion axis under a specified preload direction and stabilization waiting conditions.
[0024] In some embodiments, the standardized single-axis motion includes a first standardized single-axis motion that moves only along the first axis from a common starting point to the middle position of the first axis, a second standardized single-axis motion that moves only along the second axis from a common starting point to the middle position of the second axis, a third standardized single-axis motion that moves only along the second axis from the middle position of the first axis to a common ending point, and a fourth standardized single-axis motion that moves only along the X-axis from the middle position of the second axis to the common ending point. The first continuous combined motion corresponds to the first standardized single-axis motion and the third standardized single-axis motion, and the second continuous combined motion corresponds to the second standardized single-axis motion and the fourth standardized single-axis motion.
[0025] Specifically, this scheme performs at least the following six types of motions on the four-point structure formed by the common starting point m_0, the midpoint of the X-axis m_X, the midpoint of the Y-axis m_Y, and the common ending point m_T: 1. First standardized single-axis motion: moving only along the X-axis from m_0 to m_X; 2. Second standardized single-axis motion: moving only along the Y-axis from m_0 to m_Y; 3. Third standardized single-axis motion: motion from m_X along only the Y-axis to m_T; 4. Fourth standardized single-axis motion: motion from m_Y along only the X-axis to m_T; 5. First continuous combined motion: Continuously execute X-axis movement from m_0 to Y-axis movement to m_T; 6. Second continuous combined motion: After continuously executing the Y-axis motion from m_0, the X-axis motion is executed to m_T.
[0026] The motion process for the first continuous combination motion is as follows: m_0-m_X-m_T Therefore, the first motion segment of the first continuous combined motion is matched with the first standardized single-axis motion, and the second motion segment is matched with the third standardized single-axis motion; The motion process corresponding to the second continuous combination motion is as follows: m_0-m_Y-m_T Therefore, the first motion segment of the second continuous combined motion is matched with the second standardized single-axis motion, and the second motion segment is matched with the fourth standardized single-axis motion.
[0027] The purpose of this scheme in standardizing single-axis motion is to obtain a single-axis reference that matches each motion segment of two consecutive combined motions.
[0028] In some embodiments, before performing standardized single-axis motion, it is preferable to move the stage into the corresponding common starting point in a preset uniform direction so that each standardized single-axis motion has a basically consistent mechanical preload state; after the stage reaches the common starting point, the position loop error and encoder feedback speed can be detected, and the absolute value of the position loop error is not greater than a preset position error threshold. The absolute value of the encoder feedback speed is not greater than the preset speed threshold. And the duration of the above state reaches the preset stabilization time. Once the stage has reached a stable state, the corresponding standardized single-axis motion is executed.
[0029] It is worth mentioning that the initial position of the stage can be achieved by preloading in a fixed direction, returning after passing the target, extending the stabilization wait, or locking the closed loop servo. The key is that the initial mechanical state of each single axis segment can be repeated.
[0030] Furthermore, the position error threshold, velocity threshold, and settling time can be determined based on the positioning accuracy of the image markers, the repeatability of the stage, the encoder resolution, and the control cycle.
[0031] For example, in one implementation, the preset position error threshold can be set to ±1 micrometer, the preset velocity threshold can be set to 0.01 millimeter / second, and the above state is required to be maintained continuously (i.e., for a preset settling time) for 100 milliseconds; the specific values can be adjusted according to the accuracy level of the equipment. Data of standardized single-axis motion that does not meet the sufficient stability conditions are not used to construct a single-axis superposition reference.
[0032] Unlike standardized single-axis motion, the two continuous combined motions are preferably executed sequentially according to the actual operating speed, acceleration, and intermediate waiting time of the microscope imaging equipment. That is, when executing the first continuous combined motion, the second axis motion is executed directly according to the actual operating state after the first axis motion is completed; when executing the second continuous combined motion, the first axis motion is executed according to the actual operating state after the second axis motion is completed. It is not required that the two axes reach the fully stable state required by standardized single-axis motion, thus ensuring that the motion history effects such as hysteresis, residual vibration, and insufficient stability caused by the previous motion axis on the subsequent motion axis are retained in the actual measurement results of the continuous combined motion.
[0033] Regarding step S3: Different calibration conditions refer to different calibration positions, movement distances, movement directions, movement speeds, intermediate waiting times, and equipment states. Step S3 of this scheme is used to obtain calibration data under different calibration conditions so that step S4 can construct the single-axis reference error model, the order-keeping component model, and the order-flipping component model.
[0034] In some embodiments, for the same combination of position and displacement, the first and second consecutive combinations of motion are acquired using low-speed, medium-speed, and high-speed states of the formal scan, respectively, and with different intermediate waiting times. Single-axis references can share the position structure, but state condition terms should be added when the speed or waiting time significantly alters the single-axis response. Non-zero models are only built for components that are significantly higher than the image localization noise in independent, repetitive data to avoid unnecessary model complexity.
[0035] Specifically, regarding uniaxial residuals: This scheme obtains the corresponding single-axis residual based on the difference between the actual image increment of the corresponding marker point in each standardized single-axis motion and the predicted image increment based on the basic static mapping prediction.
[0036] Specifically, in each standardized single-axis motion segment, the difference between the marker point from the first position to the second position is used as the actual image increment; the predicted image increment of the marker point from the first position to the second position is obtained according to the basic static mapping; and the difference between the actual image increment and the predicted image increment is determined as the single-axis residual corresponding to the standardized single-axis motion segment.
[0037] For example, for the first position To the second position The standardized single-axis segment is defined with the corresponding single-axis residual as: in This represents the actual image increment from the first position to the second position. The value of the second position in the underlying static mapping. The value at the first position in the basic static mapping. For uniaxial residuals.
[0038] In some embodiments, the single-axis residuals corresponding to the four standardized single-axis motions are calculated respectively to obtain the single-axis residuals corresponding to the first standardized single-axis motion, the second standardized single-axis motion, the third standardized single-axis motion, and the fourth standardized single-axis motion.
[0039] In some embodiments, each standardized single-axis motion can be repeated multiple times, and the median, robust mean, or weighted result based on image positioning reliability is used as the final single-axis residual based on the multiple measurement results. Through the above processing, the impact of single image positioning noise and random error of repeated stage positioning on the single-axis residual can be reduced.
[0040] Regarding equivalent residuals at a common endpoint: Since different single-axis residuals are obtained at different motion positions, the local image coordinate responses corresponding to different single-axis residuals are not exactly the same. If the single-axis residuals obtained at different positions are directly added together, the difference in coordinate response at different spatial positions may be directly introduced into the single-axis superposition reference. Therefore, it is necessary to convert each single-axis residual into a common endpoint equivalent residual under the common endpoint based on the local coordinate response relationship between the single-axis residual acquisition position and the common endpoint of the basic static mapping.
[0041] Specifically, in some embodiments, the equivalent residual of the common endpoint is obtained by at least one of local first-order response, local affine mapping, bilinear interpolation, B-spline transformation or piecewise short-stroke accumulation, and the validity of the equivalent residual of the common endpoint is determined based on the back substitution error.
[0042] Each uniaxial residual is converted into a common endpoint equivalent residual.
[0043] In some embodiments, when a local first-order response is used, the equivalent residual at the common endpoint can be expressed as:
[0044] in, The local response matrix of the basic static mapping to physical coordinates, It is a common destination. It is the first position. For the common endpoint equivalent residuals after conversion, This is to obtain the local coordinate transformation relationship from the residual to the common endpoint. It represents the generalized inverse of the local response matrix.
[0045] Furthermore, when using the local response matrix for coordinate transformation, if the local response matrix does not meet the conditions for direct inversion, its generalized inverse can be used for calculation.
[0046] Transform the uniaxial residuals obtained from the original location to the common endpoint:
[0047] To evaluate the effectiveness of the equivalent transformation of the common endpoint, this scheme establishes a back-substitution transformation that returns the original residuals to their original location after obtaining the equivalent residuals from the common endpoint:
[0048] It represents the generalized inverse of the local response matrix.
[0049] And obtain the residual after back substitution:
[0050] Further calculation of back-substitution error:
[0051] in, The smaller the value, the more consistent the single-axis residual is with the original single-axis residual after being transformed to the common endpoint and then substituted back to the original position, indicating that the equivalent transformation of the common endpoint is more reliable.
[0052] By comparing back-substitution errors The system uses a preset back-substitution error threshold to determine whether the current common endpoint equivalent transformation is valid. At the same time, when using local first-order responses for common endpoint equivalent transformation, the condition number of the local response matrix is further calculated.
[0053] Only when the independent verification error of the basic static mapping meets the requirements, the condition number of the local response matrix is not higher than the preset threshold, and the back substitution error is not higher than the preset back substitution error threshold, will the equivalent residual of the common endpoint after transformation be used to construct a single-axis superposition benchmark. This can effectively control the equivalent transformation of the common endpoint from three aspects: the accuracy of the basic mapping, the numerical stability of the local transformation, and the reversibility of the transformation result.
[0054] Furthermore, if any of the above conditions are not met, the current local first-order transformation is not used, and at least one of the following processing methods is adopted: 1. Shorten the spatial step size of single-axis calibration motion to reduce the error caused by local linear approximation during long-distance conversion; 2. Add local calibration points between the residual acquisition location and the common endpoint to improve the description accuracy of the local coordinate response; 3. Replace the local first-order response with a local affine lookup table, bilinear interpolation, or B-spline transformation; 4. Break down long-distance transformations into multiple short-distance transformations and accumulate them segment by segment; 5. Re-establish or update the underlying static mappings; 6. Mark the result of this transformation as invalid and do not use it for single-axis superposition reference calculation.
[0055] By performing the above-mentioned basic mapping accuracy verification and conversion back-substitution verification, the impact of nonlinear distortion or calibration error in the basic static mapping on the transmission of the error to the single-axis superimposed reference and the combined excess residual is reduced.
[0056] In addition, in some embodiments, before performing the transformation of the common endpoint equivalent residual in step S4, the accuracy of the basic static mapping is verified using independent calibration points that were not involved in the establishment of the basic static mapping.
[0057] Specifically, the predicted microscopic image position of the independent calibration points is obtained through the basic static mapping, and the mapping residual of the current independent calibration point is obtained by combining the predicted microscopic image position. Based on the mapping residual of each independent calibration point, the average mapping residual, the preset quantile residual, or the maximum mapping residual of the independent calibration points is calculated. When the average mapping residual, the preset quantile residual, or the maximum mapping residual of the independent calibration points is not higher than the preset mapping error threshold, it is determined that the basic static mapping meets the accuracy requirements, and subsequent single-axis residual transformation is allowed to be performed using the basic static mapping.
[0058] When the mapping residual exceeds the threshold, the static calibration is re-executed, or the original basic static mapping is replaced by local affine mapping, piecewise mapping, or local measured lookup table.
[0059] In some embodiments, the mapping error threshold can be determined based on image marker positioning noise, stage repetitive positioning error, and the system's allowable target arrival error.
[0060] In some embodiments, the first The mapping residual of each independent calibration point can be expressed as:
[0061] in, For the first The location of each independent calibration point The predicted microscopic image position is obtained by fundamental static mapping for independent calibration points.
[0062] Regarding single-axis stacking references: After completing the equivalent transformation of the common endpoint and confirming the validity of the transformation result, this scheme selects the corresponding equivalent residual of the common endpoint based on the actual motion paths of the two continuous combined motions, and constructs single-axis superposition references respectively.
[0063] Specifically, the equivalent residuals of the common endpoints of the first and third standardized single-axis motions are superimposed as the single-axis superposition reference for the first continuous combined motion; the equivalent residuals of the common endpoints of the second and fourth standardized single-axis motions are superimposed as the single-axis superposition reference for the second continuous combined motion.
[0064] Specifically, for the first continuous combined motion, its single-axis superimposed reference is expressed as: = T m0 →T r m0 → m X + T m X→T r m X→ m T in T serves as the single-axis superposition reference for the first continuous combined motion. m0 →T r m0 → m X is the equivalent residual of the common endpoint of the first standardized single-axis motion, and T is... m X→T r m X→ m T is the equivalent residual of the common endpoint of the third-standardized single-axis motion.
[0065] For the second continuous combined motion, its single-axis superposition reference is expressed as: = T m0 →T r m0 → m Y + T m Y→T r m Y→ m T in T serves as the single-axis superposition reference for the second continuous combined motion. m0 →T r m0 → m Y is the equivalent residual of the common endpoint of the second-normalized single-axis motion, and T is... m Y→T r m Y→ m T is the equivalent residual of the common endpoint of the fourth-standardized single-axis motion.
[0066] Each continuous combination of motion uses a single-axis reference corresponding to its two motion segments, and does not use single-axis measurements that do not match the middle position of the path. Single-axis segments are preferably repeatedly acquired, and the median, robust mean, or weighted result based on image positioning reliability is used.
[0067] It should be noted that the single-axis superposition benchmark is not simply the direct addition of any two single-axis residuals. Instead, after undergoing equivalent transformation to a common endpoint and completing validity verification, the matching single-axis residuals are selected and superimposed according to the specific motion path. Therefore, the single-axis superposition benchmark can serve as a benchmark for subsequent identification of additional errors in dual-axis continuous combined motion.
[0068] Regarding the combined excess residual: The formula for obtaining the combined excess residual of the first continuous combination motion is as follows:
[0069] in The measured endpoint residual of the first continuous combination motion. As the single-axis superposition reference for the first continuous combination motion, This is the combined excess residual of the first continuous combination of motions; The formula for obtaining the combined excess residual of the second continuous combined motion is as follows:
[0070] in This represents the measured endpoint residual of the second continuous combined motion. As a single-axis superposition reference for the second continuous combined motion, This is the combined excess residual of the second continuous combination motion.
[0071] The combined excess residuals obtained by this scheme represent additional errors that cannot be explained by the superposition of standardized single-axis errors and only occur in continuous combined motions of two axes.
[0072] Regarding order-reversed components and order-preserving components: Since the first and second continuous combined motions have the same starting point, the same ending point, the same command displacements for each axis, and the same or matching velocity, acceleration, and waiting time conditions, the first and second continuous combined motions can be regarded as a pair of combined motions that only exchange the execution order of axes. Therefore, according to the symmetric-antisymmetric decomposition principle of the axis sequence exchange operator, any pair of combined excess residuals can be uniquely represented as the order-preserving component maintained after axis sequence exchange and the order-reversing component whose sign changes after axis sequence exchange.
[0073] In some embodiments, the average of the sum of the combined excess residuals of the second consecutive combined motion and the combined excess residuals of the first consecutive combined motion is taken as the order-preserving component, and the average of the difference between the combined excess residuals of the second consecutive combined motion and the combined excess residuals of the first consecutive combined motion is taken as the order-reversing component.
[0074] The calculation method for the order-preserving components is as follows:
[0075] The calculation method for the order-reversed components is as follows:
[0076] in This is the combined excess residual for the second consecutive combination of motions. This is the combined excess residual of the first continuous combination motion.
[0077] The decomposition of the order-preserving component and the order-reversing component in this scheme is a rigorous algebraic reparameterization of the two types of axis-order observations. It does not require the overall mechanical response of the stage to be linear with respect to position, velocity, load or temperature, nor does it require that various mechanical errors be linearly superimposed on each other in terms of physical mechanism.
[0078] It needs to be explained that, from the perspective of the axis order exchange operator, let the exchange operator be... Converting the path observation of the first continuous combined motion into the path observation of the second continuous combined motion, then That is, restoring the original axis order by swapping the axes twice consecutively, the corresponding symmetric and antisymmetric projections are as follows:
[0079] in, To extract the parts that remain unchanged after the axis sequence is swapped. To extract the portion whose sign or direction of action changes after axis sequence exchange, any combination error can be expressed as the sum of these two projection components.
[0080] In some embodiments, the order retention component is used to characterize the biaxial combined coupling error retained after the shaft order is changed, that is, to characterize the combined coupling that exists in both shaft orders, including common load changes, structural deformation, inter-shaft stiffness coupling, and other biaxial nonlinear effects that are not sensitive to shaft order changes; the order reversal component is used to characterize the motion history error that changes with the shaft order, that is, to characterize the motion history effect that changes the direction or sign of action with the change of execution order, including preload direction, hysteresis, lag, residual vibration after the previous shaft motion, and instability.
[0081] For example, if a biaxial nonlinear error contains a coupling term related to the product of the two-axis displacements:
[0082] Because the execution order of the X-axis and Y-axis was swapped, The error itself remains unchanged, and it mainly enters the system. If an error is caused by directional preloading or hysteresis of the preceding axis of motion on the following axis of motion, and exhibits opposite effects in the XY and YX paths, then this error mainly enters... If a complex nonlinear error contains both axis-sequence-independent and axis-sequence-dependent components, then its corresponding components will be entered into... and .
[0083] Therefore, this invention does not assume that nonlinear coupling does not exist, but decomposes it according to the response properties of nonlinear coupling under axis sequence exchange. The order-preserving component and the order-reversing component can still be expressed as nonlinear functions of target position, movement distance, direction, velocity, acceleration, intermediate waiting time, temperature or stage load, respectively. The corresponding path condition compensation relationship can be established through piecewise lookup tables, direction and axis sequence condition splines, local polynomials or state space models.
[0084] When the measurement reliability of the two paths of the first continuous combined motion and the second continuous combined motion is different, a reliability-weighted symmetric-antisymmetric decomposition can also be used. The half-sum and half-difference are the preferred implementations when the reliability of the two paths is the same or approximately the same.
[0085] Regarding step S4: Error models are constructed based on calibration data obtained under different calibration conditions. The calibration data includes single-axis residuals, order-preserving components, and order-reversing components. Therefore, the corresponding error models include single-axis reference error model, order-preserving component model, and order-reversing component model.
[0086] In some embodiments, the error model can establish a corresponding relationship based on the target position, movement distance, movement direction, movement speed, intermediate waiting time, and equipment status parameters such as stage temperature and stage load.
[0087] In some embodiments, the single-axis reference error model is used to describe the mechanical error generated when a single-axis motion is performed alone; the order-preserving component model is used to describe the combined coupling error that is related to the combined motion of two axes and maintains the same action trend under different axis sequences; and the order-reversing component model is used to describe the motion history error that changes the action direction or sign as the execution order of the first axis and the second axis changes.
[0088] In some embodiments, the error model may employ a piecewise lookup table, directional condition spline, local polynomial, state-space model, or other models capable of establishing a mapping relationship between path conditions and corresponding error components.
[0089] It should be noted that the specific mathematical form of the error model does not constitute the core limitation of this invention. This invention mainly utilizes the differences between different error components under single-axis motion, dual-axis combined motion, and axis sequence exchange conditions to perform error modeling and compensation.
[0090] Regarding step S4: In the step of "querying the single-axis residual, sequence-keeping component, and sequence-reversal component in the self-error model based on the target position and planned motion path, determining the direction of the sequence-reversal component according to the planned motion path, and determining the stage command compensation amount using the single-axis residual, sequence-keeping component, and sequence-reversal component", when the planned motion path is a combined motion, the sign function is determined according to the planned motion axis sequence, and the stage command compensation amount is obtained by summing the single-axis reference error, sequence-keeping component, and sequence-reversal component of the X-axis and Y-axis, where the sign function is used to change the sign of the sequence-reversal component; when the planned motion path is a single-axis motion, the stage command compensation amount is obtained by summing the single-axis reference error of the X-axis and Y-axis.
[0091] When the planned motion path is a combined motion, the stage command compensation is obtained as follows: b_stage(m,a) = b_X + b_Y + b_keep + s(a)b_flip Where b_stage(m,a) is the stage command compensation amount, b_X is the X-axis single-axis reference error, b_Y is the Y-axis single-axis reference error, b_keep is the order retention component, b_flip is the order flip component, and s(a) is the sign function.
[0092] In some embodiments, when the axis order of the planned motion path is X-axis first then Y-axis, s(a) = +1; when the axis order of the planned motion path is Y-axis first then X-axis, s(a) = -1.
[0093] In addition, when the planned motion path is only a single-axis motion, no order maintenance component and order reversal component are added. The stage command compensation amount is determined only by using the single-axis reference error corresponding to the current actual motion axis.
[0094] When only X-axis motion is performed: b_stage(m,a) = b_X, When only Y-axis motion is performed: b_stage(m,a) = b_Y, Where b_stage(m,a) is the stage command compensation amount, b_X is the X-axis single-axis reference error, and b_Y is the Y-axis single-axis reference error.
[0095] In the step of "obtaining the stage command coordinates based on the stage command compensation amount and the target position", the stage command compensation amount is converted into physical coordinate compensation through the local inverse response of the basic static mapping, the target position is converted into the nominal target position through the local inverse response of the basic static mapping, and the stage command coordinates are obtained by combining the nominal target position and the physical coordinate compensation correction.
[0096] Specifically, the stage command coordinates are represented as follows: m_cmd = m_target -Δm_stage(m_target,a) Where m_target is the nominal target position, and Δm_stage(m_target,a) is the physical coordinate compensation.
[0097] That is, the target position in this scheme is first obtained by obtaining the nominal physical coordinates through existing static mapping, and then the single-axis, order-preserving and order-reversing components are selected for command correction according to the planned motion path.
[0098] In addition, to avoid the calibration data used for model fitting and the model validation data from interfering with each other, this scheme divides the acquired marker points and planned motions into a fitting set and an independent validation set. The fitting set is used to establish the single-axis baseline error model, the order-preserving component model, and the order-reversing component model, while the marker points and planned motions in the independent validation set do not participate in the fitting of the above error models.
[0099] After the candidate error model is established, the real arrival task is performed at the calibration position in the independent validation set that is not involved in the model fitting. The real arrival task includes at least the single-axis real arrival task, the XY axis sequence real arrival task, and the YX axis sequence real arrival task.
[0100] Based on the actual arrival point measurement results, calculate the remaining amount of single-axis compensation, the remaining amount of order maintenance, the remaining amount of order reversal, and the arrival error between the target physical position and the actual physical position. When the remaining amount of single-axis compensation, the remaining amount of order maintenance, the remaining amount of order reversal, and the error between the target physical position and the actual physical position are all not higher than the corresponding preset error threshold, the candidate error model is deemed to have passed the verification and is allowed to be used for subsequent stage command compensation.
[0101] In some embodiments, the single-axis compensation residual is used to evaluate the residual error of the single-axis reference error model after actual compensation, the order-keeping residual is used to evaluate the combined coupling error compensation effect of the order-keeping component model, the order-flipping residual is used to evaluate the motion history error compensation effect of the order-flipping component model, and the arrival error between the target physical position and the actual arrived physical position is used to evaluate the final positioning effect after model compensation.
[0102] In addition, when at least one remaining quantity exceeds the corresponding preset error threshold during the independent verification process, the current candidate error model is not directly activated. Instead, the corresponding model update object is determined based on the remaining quantity that exceeds the threshold.
[0103] When the remaining amount of single-axis compensation exceeds the corresponding threshold, the single-axis baseline error sub-model corresponding to the remaining amount of single-axis compensation is updated, without changing the validated order-preserving component model and order-flipping component model.
[0104] When the order-preserving residual exceeds the corresponding threshold, supplement the calibration data related to the biaxial interaction, and update the order-preserving component model using the supplemented co-coupling data.
[0105] When the remaining amount of sequence reversal exceeds the corresponding threshold, priority is given to supplementing calibration data related to the motion history. The calibration data related to the motion history includes at least intermediate waiting time, mechanical backlash, reversal state, and other state data that affect the sequence relationship of motion. The sequence reversal component model is then updated using the supplemented data.
[0106] By updating each component as described above, the model update focuses on the error sources that exceed the error limit in actual verification, avoiding the need to refit the entire error model due to a single error component exceeding the limit.
[0107] Furthermore, for scanning devices with a Z-axis or a rotation axis, corresponding common endpoint structures can be set for X / Y, X / Z, Y / Z or translation and rotation axes respectively, and single-axis segmented reference motion and continuous combination motion of two opposite axis sequences can be performed pair by pair. The single-axis error, order maintenance coupling component and order reversal coupling component of each axis pair can be obtained according to the aforementioned method.
[0108] After error compensation for each axis pair is completed, independent verification is still performed in a three-axis combined motion task that is not involved in model building. If the actual endpoint residual of the three-axis combined motion, after deducting the compensation amounts of each single axis and the pairwise combined compensation amounts of X / Y, X / Z, and Y / Z, is not higher than the preset residual threshold, then it is considered that the error separation of each pair can meet the positioning accuracy requirements of the current equipment, and no further three-axis joint compensation term is established. The preset residual threshold can be determined based on the positioning noise of image marker points, the repeated positioning error of the stage, and the prediction error of the two-axis compensation model in the independent verification data. If, after deducting the single-axis error and the pairwise combined error, the remaining error is still stably higher than the preset residual threshold in multiple repeated measurements, then the remaining error is determined as the three-axis combined excess residual, expressed as: in This indicates one of the following three-axis execution sequences: XYZ, XZY, YXZ, YZX, ZXY, or ZYX. Indicates execution order The measured endpoint residual obtained after completing the triaxial continuous combined motion. This represents the predicted superposition result formed by the equivalent transformation of three single-axis residuals that match each motion segment of the three-axis path, after being converted to a common endpoint. This represents the predicted superposition result formed by the pairwise coupled components of X / Y, X / Z, and Y / Z according to the current three-axis execution order. This indicates the excess residual of the three-axis combination that cannot be explained by single-axis error or coupling error between pairs of axes, and only occurs in continuous three-axis combined motion.
[0109] Subsequently, based on the relationship between the combined excess residual of the three axes and the target position, the execution order of the three axes, the direction of motion, the speed of motion, the acceleration, and the stabilization waiting time, a three-axis joint compensation model is established. The three-axis joint compensation model can adopt a piecewise lookup table, direction and axis sequence condition splines, local polynomials, or a state-space model; the specific model form does not constitute the core of this invention.
[0110] In actual motion control, the stage command compensation quantities include, in sequence: single-axis compensation quantities corresponding to each motion axis, pairwise combination compensation quantities corresponding to each axis pair in the current motion path, and three-axis joint compensation quantities corresponding to the current three-axis execution order when the three-axis combination excess residual is stably higher than the preset residual threshold. The three-axis joint compensation term is only enabled when the pairwise error separation and pairwise combination compensation still cannot meet the preset positioning accuracy. When the three-axis residual error is not significantly higher than the image positioning noise, stage repetitive positioning error and two-axis model prediction error, the three-axis joint compensation term is not established to avoid introducing high-order parameters lacking stable physical basis and unnecessary model complexity.
[0111] Example 2 This second embodiment provides a specific application example: A pathology slide scanner includes an area array camera, a 20x microscope objective, an X / Y motorized stage, and an encoder. It uses calibration slides with coded crosshairs, with the markers covering the center, four corners, and edges of the work area.
[0112] First, a basic static mapping is obtained through routine calibration, enabling the stage to move the target marker near the image center. Four positions, m0, m_X, m_Y, and m_T, are set in each selected region. The single-segment displacement is preferably 0.2 mm to 5 mm. Before the standardized single-axis segment motion begins, the stage enters the starting point from a uniform direction and waits for sufficient stabilization. Continuous combined motions are then executed continuously using the nominal speed, acceleration, and intermediate waiting time of the formal scan. Each type of motion is repeated 3 to 10 times.
[0113] Calculate the four single-axis residuals based on the location of the marker point image, and transform them to the common endpoint through local response or segmented lookup table; form the single-axis superposition reference for XY and YX respectively; calculate the combined excess residual and the two types of axis sequence components, and establish a local compensation table; then perform target centering on independent marker points, and record the average error, 95% error, maximum edge error, and the remaining amount after compensation of the two types of axis sequence components.
[0114] Example 3 Referring to Embodiment 1, this solution provides a microscope image-stage coordinate compensation device, comprising the following: The mapping module is used to obtain the basic static coordinate mapping between the microscopic image and the stage; The motion control module is used to execute a first continuous combination motion of X-axis followed by Y-axis and a second continuous combination motion of Y-axis followed by X-axis for the same four-point structure on the working area of the stage, and to obtain the standardized single-axis motion that matches each motion segment of the two continuous combination motions respectively. The calibration module is used to obtain the order-preserving and flipping components under different calibration conditions. For each calibration condition, the difference between the actual image increment of the marker point in each standardized single-axis motion and the predicted image increment based on the basic static mapping is obtained to obtain the corresponding single-axis residual. Based on the local coordinate response of the basic static mapping between the obtained position of the single-axis residual and the common endpoint, each single-axis residual is converted into a common endpoint equivalent residual. Based on the common endpoint equivalent residual matched with the motion segments of each continuous combined motion, corresponding single-axis superposition references are constructed respectively. The measured endpoint residuals of the two continuous combined motions are obtained respectively, and the difference between the measured endpoint residuals and the corresponding single-axis superposition references is calculated to obtain the combined excess residual. Based on the preservation and flipping relationships of the combined excess residuals corresponding to the two continuous combined motions under axis order exchange, the order-preserving and order-flipping components are determined respectively. The data acquisition module is used to construct an error model based on calibration data acquired under different calibration conditions. The calibration data includes single-axis residuals, order-preserving components, and order-reversing components. The compensation module is used to query the single-axis residual, order-keeping component, and order-flipping component in the self-error model based on the target position and planned motion path, determine the direction of the order-flipping component according to the planned motion path, determine the stage command compensation amount using the single-axis residual, order-keeping component, and order-flipping component, and determine the stage command coordinates based on the stage command compensation amount and the target position.
[0115] The content of this embodiment three is the same as that of embodiment one, and will not be described in detail here.
[0116] Example 4 This embodiment also provides an electronic device, see reference. Figure 2It includes a memory 402 and a processor 401, the memory 402 storing a computer program and the processor 401 being configured to run the computer program to perform the steps in any of the embodiments of the microscope image-stage coordinate compensation method described above.
[0117] Specifically, the processor 401 may include a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.
[0118] The memory 402 may include a large-capacity memory 402 for data or instructions. The memory 402 can be used to store or cache various data files that need to be processed and / or communicated, as well as possible computer program instructions executed by the processor 401.
[0119] The processor 401 reads and executes computer program instructions stored in the memory 402 to implement any of the microscope image-stage coordinate compensation methods in the above embodiments.
[0120] Optionally, the electronic device may further include a transmission device 403 and an input / output device 404, wherein the transmission device 403 is connected to the processor 401 and the input / output device 404 is connected to the processor 401.
[0121] The transmission device 403 can be used to receive or send data via a network. Specific examples of the network described above may include wired or wireless networks provided by the communication provider of the electronic device. In one example, the transmission device includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 403 may be a Radio Frequency (RF) module used for wireless communication with the Internet.
[0122] The input / output device 404 is used to input or output information. In this embodiment, the input information may be the target position and planned motion, and the output information may be the stage command coordinates, etc.
[0123] Optionally, in this embodiment, the processor 401 can be configured to perform the following steps via a computer program: S1: Obtain the basic static coordinate mapping of the microscopic image and the stage; S2: For the same four-point structure on the working area of the stage, execute the first continuous combination motion of X-axis first and Y-axis and the second continuous combination motion of Y-axis first and X-axis respectively, and obtain the standardized single-axis motion that matches each motion segment of the two continuous combination motions respectively. S3: Obtain the order-preserving and flipping components under different calibration conditions. For each calibration condition, obtain the difference between the actual image increment of the marker point in each standardized single-axis motion and the predicted image increment based on the basic static mapping to obtain the corresponding single-axis residual. Based on the local coordinate response of the basic static mapping between the obtained position of the single-axis residual and the common endpoint, convert each single-axis residual into a common endpoint equivalent residual. Based on the common endpoint equivalent residual matched with the motion segments of each continuous combined motion, construct the corresponding single-axis superposition reference. Obtain the measured endpoint residuals of the two continuous combined motions respectively, and subtract the measured endpoint residuals from the corresponding single-axis superposition references to obtain the combined excess residual. Determine the order-preserving and order-flipping components based on the preservation and flipping relationships of the combined excess residuals corresponding to the two continuous combined motions under axis order exchange. S4: Construct an error model based on calibration data obtained under different calibration conditions, where the calibration data includes single-axis residuals, order-preserving components, and order-reversing components; S5: Based on the target position and the planned motion path, query the single-axis residual, order-keeping component, and order-flipping component from the self-error model. Determine the direction of the order-flipping component according to the planned motion path. Use the single-axis residual, order-keeping component, and order-flipping component to determine the stage command compensation amount. Determine the stage command coordinates based on the stage command compensation amount and the target position.
[0124] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementations, and will not be repeated here.
[0125] Generally, various embodiments can be implemented in hardware or dedicated circuitry, software, logic, or any combination thereof. Some aspects of the invention can be implemented in hardware, while others can be implemented by firmware or software executed by a controller, microprocessor, or other computing device, but the invention is not limited thereto. Although various aspects of the invention may be shown and described as block diagrams, flowcharts, or using some other graphical representation, it should be understood that, by way of non-limiting example, these blocks, apparatuses, systems, techniques, or methods described herein can be implemented in hardware, software, firmware, dedicated circuitry or logic, general-purpose hardware or controllers or other computing devices, or some combination thereof.
[0126] Embodiments of the present invention can be implemented by computer software, which may be executable by a data processor of a mobile device, such as a processor entity, or by hardware, or by a combination of software and hardware. Computer software or programs (also referred to as program products), including software routines, applets, and / or macros, can be stored in any device-readable data storage medium, and they include program instructions for performing specific tasks. A computer program product may include one or more computer-executable components configured to perform embodiments when the program is run. One or more computer-executable components may be at least one piece of software code or a portion thereof. Additionally, it should be noted that any block in the logical flow of the figures may represent a program step, or interconnected logical circuitry, blocks and functions, or a combination of program steps and logical circuitry, blocks and functions. The software may be stored on physical media such as memory chips or blocks of storage implemented within a processor, magnetic media such as hard disks or floppy disks, and optical media such as, for example, DVDs and their data variants, CDs, etc. The physical medium is a non-transient medium.
[0127] Those skilled in the art should understand that the technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments have been described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
Claims
1. A method for compensating microscope image-stage coordinates, characterized in that, Includes the following steps: S1: Obtain the basic static coordinate mapping of the microscopic image and the stage; S2: For the same four-point structure on the working area of the stage, execute a first continuous combination motion of first axis followed by second axis and a second continuous combination motion of second axis followed by first axis, and obtain the standardized single-axis motion that matches each motion segment of the two continuous combination motions respectively; wherein the four-point structure has a common starting point, the middle position of the first axis, the middle position of the second axis, and a common ending point, the first continuous combination motion and the second continuous combination motion have the same common starting point, the same ending point, and the same command displacement, the first continuous combination motion is a continuous combination motion in which the first axis motion is executed first and then the second axis motion, and the second continuous combination motion is a continuous combination motion in which the second axis motion is executed first and then the first axis motion; S3: Obtain the order-preserving and flipping components under different calibration conditions. For each calibration condition, obtain the difference between the actual image increment of the marker point in each standardized single-axis motion and the predicted image increment based on the basic static mapping to obtain the corresponding single-axis residual. Based on the local coordinate response of the basic static mapping between the obtained position of the single-axis residual and the common endpoint, convert each single-axis residual into a common endpoint equivalent residual. Based on the common endpoint equivalent residual matched with the motion segments of each continuous combined motion, construct the corresponding single-axis superposition reference. Obtain the measured endpoint residuals of the two continuous combined motions respectively, and subtract the measured endpoint residuals from the corresponding single-axis superposition references to obtain the combined excess residual. Determine the order-preserving and order-flipping components based on the preservation and flipping relationships of the combined excess residuals corresponding to the two continuous combined motions under axis order exchange. S4: Construct an error model based on calibration data obtained under different calibration conditions, where the calibration data includes single-axis residuals, order-preserving components, and order-reversing components; S5: Based on the target position and the planned motion path, query the single-axis residual, order-keeping component, and order-reversal component from the self-error model. Determine the direction of the order-reversal component according to the planned motion path. Use the single-axis residual, order-keeping component, and order-reversal component to determine the stage command compensation amount. Determine the stage command coordinates based on the stage command compensation amount and the target position. Among them, the average value of the sum of the combined excess residual of the second continuous combined motion and the combined excess residual of the first continuous combined motion is taken as the order-keeping component, and the average value of the difference between the combined excess residual of the second continuous combined motion and the combined excess residual of the first continuous combined motion is taken as the order-reversal component.
2. The microscope image-stage coordinate compensation method according to claim 1, characterized in that, Obtain the physical marker position and the microscopic image marker position, where the physical marker position is the position of the physical marker point on the stage, and the microscopic image marker position is the position of the physical marker point on the microscopic image. The basic static mapping records the mapping relationship between the physical marker position and the microscopic image marker position of the same physical marker point.
3. The microscope image-stage coordinate compensation method according to claim 1, characterized in that, In each standardized single-axis motion segment, the difference between the marker point from the first position to the second position is used as the actual image increment. The predicted image increment of the marker point from the first position to the second position is obtained according to the basic static mapping. The difference between the actual image increment and the predicted image increment is determined as the single-axis residual corresponding to the standardized single-axis motion segment.
4. The microscope image-stage coordinate compensation method according to claim 1, characterized in that, The equivalent residual of the common endpoint is obtained through at least one of the following methods: local first-order response, local affine mapping, bilinear interpolation, B-spline transformation, or piecewise short-stroke accumulation. The validity of the equivalent residual of the common endpoint is determined based on the back substitution error. The transformed equivalent residual of the common endpoint is used to construct a single-axis superposition reference only when the independent verification error of the basic static mapping meets the requirements, the condition number of the local response matrix is not higher than the preset threshold, and the back substitution error is not higher than the preset back substitution error threshold.
5. The microscope image-stage coordinate compensation method according to claim 1, characterized in that, The system obtains the predicted microscopic image position of the independent calibration points through the basic static mapping, and obtains the mapping residual of the current independent calibration point by combining the predicted microscopic image position. Based on the mapping residual of each independent calibration point, the system calculates the average mapping residual, the preset quantile residual, or the maximum mapping residual of the independent calibration points. When the average mapping residual, the preset quantile residual, or the maximum mapping residual of the independent calibration points is not higher than the preset mapping error threshold, the system determines that the basic static mapping meets the accuracy requirements and allows the system to perform subsequent single-axis residual transformations using the basic static mapping.
6. The microscope image-stage coordinate compensation method according to claim 1, characterized in that, The error models include the single-axis reference error model, the order retention component model, and the order reversal component model. The single-axis reference error model is used to describe the mechanical error generated when performing a single-axis motion alone. The order-preserving component model is used to describe the combined coupling error related to dual-axis combined motion and which maintains the same action trend under different axis sequences; the order-reversing component model is used to describe the motion history error that changes the action direction or sign as the execution order of the first and second axes changes.
7. A microscope image-stage coordinate compensation device, characterized in that, Including the following: The mapping module is used to obtain the basic static coordinate mapping between the microscopic image and the stage; The motion control module is used to execute a first continuous combination motion (X-axis first, then Y-axis) and a second continuous combination motion (Y-axis first, then X-axis) on the same four-point structure on the working area of the stage, and to acquire standardized single-axis motions that match each motion segment of the two continuous combination motions. The four-point structure has a common starting point, a middle position of the first axis, a middle position of the second axis, and a common ending point. The first and second continuous combination motions have the same common starting point, a common ending point, and the same command displacement. The first continuous combination motion is a continuous combination motion in which the first axis motion is executed first and then the second axis motion, and the second continuous combination motion is a continuous combination motion in which the second axis motion is executed first and then the first axis motion. The calibration module is used to acquire order-preserving and flipping components under different calibration conditions. For each calibration condition, it obtains the difference between the actual image increment of the marker point in each standardized single-axis motion and the predicted image increment based on the base static mapping, thus obtaining the corresponding single-axis residual. Based on the local coordinate response of the base static mapping between the acquired single-axis residual and the common endpoint, each single-axis residual is converted into a common endpoint equivalent residual. Based on the common endpoint equivalent residual matched with the motion segments of each continuous combined motion, corresponding single-axis stacking references are constructed. Two types of continuous... The measured endpoint residual of the continuous combined motion is obtained, and the difference between the measured endpoint residual and the corresponding single-axis superposition datum is calculated to obtain the combined excess residual. The order preservation component and the order reversal component are determined according to the preservation relationship and reversal relationship of the combined excess residuals corresponding to the two continuous combined motions under axis order exchange. The average value of the sum of the combined excess residuals of the second continuous combined motion and the combined excess residuals of the first continuous combined motion is taken as the order preservation component, and the average value of the difference between the combined excess residuals of the second continuous combined motion and the combined excess residuals of the first continuous combined motion is taken as the order reversal component. The data acquisition module is used to construct an error model based on calibration data acquired under different calibration conditions. The calibration data includes single-axis residuals, order-preserving components, and order-reversing components. The compensation module is used to query the single-axis residual, order-keeping component, and order-flipping component in the self-error model based on the target position and planned motion path, determine the direction of the order-flipping component according to the planned motion path, determine the stage command compensation amount using the single-axis residual, order-keeping component, and order-flipping component, and determine the stage command coordinates based on the stage command compensation amount and the target position.
8. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to run the computer program to perform the microscope image-stage coordinate compensation method according to any one of claims 1 to 6.
Citation Information
Patent Citations
Three-dimensional measuring method
CN102818532A
Motion platform
WO2024081644A1
