Integrated circuit visual positioning calibration method based on pulse displacement reference and medium

Abstract

The invention relates to the technical field of integrated circuit packaging, solves the problem that a calibration process in the prior art lacks a stable displacement reference consistent with actual movement of equipment, and provides an integrated circuit visual positioning calibration method based on a pulse displacement reference and a medium. The method comprises the following steps: controlling a motion platform to move along a first direction and a second direction from a preset starting point according to a preset pulse number, and obtaining a calibration data set; acquiring pulse deviation data and pixel change data according to the calibration data set; according to the pulse deviation data and the pixel change data, obtaining a rotation deviation angle between the motion coordinate system and the imaging coordinate system; and when a target chip comprising an integrated circuit is subjected to die bonding, compensating a conversion relation between an imaging coordinate system and a motion coordinate system according to the rotation deviation angle. According to the invention, pulse displacement is introduced as a unified calibration reference, and the reliability of a visual positioning result in an integrated circuit die bonding process is improved.

Description

Technical Field

[0001] This invention relates to the field of integrated circuit packaging, and in particular to a visual positioning calibration method and medium for integrated circuits based on pulse displacement reference. Background Technology

[0002] In integrated circuit packaging equipment, especially die bonding equipment, the accuracy of visual positioning directly determines the accuracy of chip placement. To achieve high-precision visual positioning, it is usually necessary to calibrate the correspondence between the imaging coordinate system and the motion platform coordinate system, including the visual equivalent and the directional or angular relationship between the two coordinate systems.

[0003] In existing technologies, some solutions attempt to calibrate the motion direction of a displacement stage using visual methods. For example, CN113847874B provides a vision-based method for calibrating the displacement direction of a displacement stage. This method involves setting a calibration plate on the displacement stage, acquiring calibration images at multiple positions during the displacement process, and combining this with a camera imaging model to solve for the displacement direction vector of the displacement stage in the camera coordinate system using the least squares method, thereby achieving the calibration of the displacement direction.

[0004] However, the above technical solution still has a key technical problem: its displacement reference depends on the set movement step size or continuous displacement model of the displacement stage. During the calibration process, the encoder pulse inside the motion control system is not used as a direct and closed-loop verifiable displacement reference, which causes the correspondence between the displacement direction and the visual pixel change to still be affected by the accumulation of mechanical errors, displacement execution deviation and model assumption errors.

[0005] Therefore, there is an urgent need for a visual positioning calibration method that can establish a mapping relationship between pixel space and motion space based on a stable displacement reference inside the device, so as to improve the engineering consistency and usability of the calibration results. Summary of the Invention

[0006] In view of this, embodiments of the present invention provide an integrated circuit visual positioning calibration method and medium based on pulse displacement reference, in order to solve the problem in the prior art that the calibration process lacks a stable displacement reference consistent with the actual movement of the device, resulting in insufficient engineering consistency and usability of the mapping relationship between pixel space and motion space.

[0007] In a first aspect, embodiments of the present invention provide an integrated circuit visual positioning and calibration method based on a pulse displacement reference, used in a die bonding device. The die bonding device includes a motion control module, an imaging module, and a calibration module. The motion control module includes a motion platform and an encoding controller. The calibration module includes calibration features placed on the motion platform. The encoding controller controls the motion platform to move in at least two preset directions based on a pulse-driven method. The method includes: The motion platform is controlled to move from a preset starting point along a first direction and a second direction according to a preset number of pulses, and a calibration data set is obtained. The calibration data set includes a calibration image set and the actual number of pulses. The first direction and the second direction are perpendicular. Based on the calibration data set, obtain pulse deviation data and pixel change data; Based on the pulse deviation data and pixel change data, the rotation offset angle between the motion coordinate system of the motion platform and the imaging coordinate system of the imaging module is obtained; When performing die bonding on a target chip including an integrated circuit, coordinate rotation compensation processing is performed on the pixel displacement of the target chip according to the rotation offset angle to generate pulse displacement commands in the motion coordinate system to control the movement of the motion platform.

[0008] Preferably, the step of controlling the motion platform to move from a preset starting point along a first direction and a second direction according to a preset number of pulses to obtain a calibration data set includes: The motion platform is controlled to move the calibration feature to the center region of the field of view of the imaging module; The motion platform is controlled to move from a preset starting point along a first direction according to the preset pulse count and preset moving speed; The imaging module is controlled to acquire the first starting point image of the calibration features before motion and the first ending point image after motion, while the first actual pulse count of the encoder controller is acquired after motion. After the motion platform returns to the preset starting point, the motion platform is controlled to move from the preset starting point along the second direction according to the preset number of pulses and the preset moving speed; The imaging module is controlled to acquire the second starting point image of the calibration features before motion and the second ending point image after motion, while acquiring the second actual pulse count of the encoder controller after motion; Wherein, the preset pulse number is greater than the preset pulse number threshold, and the preset moving speed is less than the preset low speed threshold.

[0009] Preferably, the preset starting point includes a preset number of designated starting point positions distributed at different locations on the motion platform. After controlling the imaging module to acquire the second starting point image of the calibration features before motion and the second ending point image after motion, and simultaneously acquiring the second actual pulse count of the encoding controller after motion, the method further includes: The process involves controlling the motion platform to move to another designated starting position and returning to the step of controlling the motion platform to move from the preset starting point along the first direction according to the preset number of pulses and the preset moving speed, until all designated starting positions have been traversed, and a calibration data set corresponding to each designated starting position is obtained.

[0010] Preferably, obtaining pulse deviation data and pixel change data based on the calibration data set includes: Based on the differences between the first actual pulse count and the second actual pulse count and the preset pulse count, the preset number of pulse deviation data is obtained, wherein the pulse deviation data includes pulse deviation data corresponding to the first direction and the second direction respectively; Based on the calibration image set corresponding to each specified starting position, the preset number of pixel change data is obtained, wherein the pixel change data includes pixel change data corresponding to the first direction and the second direction respectively.

[0011] Preferably, obtaining the rotational offset angle between the motion coordinate system of the motion platform and the imaging coordinate system of the imaging module based on the pulse deviation data and pixel change data includes: Based on the pulse deviation data, obtain the actual displacement reference data corresponding to the controlled movement in the first direction and the second direction; Based on the pixel change data, obtain the pixel displacement components of the controlled movement in the first and second directions in the imaging coordinate system. Based on the directional correspondence between the actual displacement reference data and the pixel displacement components, an overdetermined mapping model is constructed to characterize the projection relationship between the first direction and the second direction in the imaging coordinate system. The overdetermined mapping model is solved using the least squares fitting method to obtain the local rotation offset angles corresponding to each specified starting position; Based on the local rotation offset angle at each specified starting point position and the preset consistency judgment rule, the rotation offset angle between the motion coordinate system of the motion platform and the imaging coordinate system of the imaging module is obtained.

[0012] Preferably, the rotational offset angle between the motion coordinate system of the motion platform and the imaging coordinate system of the imaging module is obtained based on the local rotational offset angle at each specified starting point position and a preset consistency judgment rule, including: Obtain the local rotation offset angle corresponding to each specified starting position to obtain a set of local angles; For each local rotation offset angle in the set of local angles, the least squares solution residual of the overdetermined mapping model is used to calculate the residual index of the local rotation offset angle. Based on the residual index, the confidence weight of each local rotation offset angle is obtained, wherein the confidence weight is negatively correlated with the residual index; Based on the confidence weight, the local angle set is weighted and fused to obtain candidate rotation offset angles; Calculate the deviation of each local rotation offset angle in the local angle set relative to the candidate rotation offset angle to obtain the deviation set; The set of deviations is compared with a preset consistency threshold to obtain a consistency judgment result; If the consistency judgment result meets the preset consistency conditions, the candidate rotation offset angle is taken as the final rotation offset angle.

[0013] Preferably, the step of performing coordinate rotation compensation processing on the pixel displacement of the target chip, including the integrated circuit, according to the rotation offset angle during die bonding, and generating a pulse displacement command in the motion coordinate system to control the movement of the motion platform includes: The imaging module obtains the target pixel position of the target chip in the imaging coordinate system and determines the pixel displacement of the target chip. Based on the rotation offset angle, the pixel displacement is subjected to coordinate rotation compensation processing to obtain the compensated displacement. Based on the compensated displacement and pulse equivalent, obtain the pulse displacement command in the motion coordinate system; The motion platform is controlled to move according to the pulse displacement command in order to complete the die bonding of the target chip including the integrated circuit.

[0014] Preferably, obtaining the pulse displacement command in the motion coordinate system based on the compensated displacement and pulse equivalent includes: Obtain the difference between the current ambient temperature and the preset temperature value to get the temperature deviation data; Based on the temperature deviation data and the preset temperature compensation coefficient, the pulse equivalent compensation value is obtained; The pulse equivalent is compensated according to the pulse equivalent compensation value to obtain the compensated pulse equivalent; The pulse displacement command is obtained based on the compensated displacement and the compensated pulse equivalent.

[0015] Preferably, before compensating for the conversion relationship between the imaging coordinate system and the motion coordinate system based on the rotation offset angle when performing die bonding on the target chip including the integrated circuit, the method further includes: The motion platform is controlled to perform test movements in the first and second directions according to the preset test pulse displacement; The imaging module acquires pixel displacement data of the target features before and after the test movement. Based on the pixel displacement data, the rotation offset angle, and the pulse equivalent, the corresponding predicted pulse displacement is calculated in reverse. The predicted pulse displacement is compared with the test pulse displacement to obtain the calibration error; If the calibration error is less than a preset error threshold, the calibration result is deemed valid.

[0016] Secondly, embodiments of the present invention provide an integrated circuit visual positioning and calibration system based on a pulse displacement reference. The system includes a die-bonding device and a processor. The die-bonding device includes a motion control module, an imaging module, and a calibration module. The motion control module includes a motion platform and an encoding controller. The encoding controller controls the motion platform to move in at least two preset directions based on a pulse-driven method and outputs the actual number of pulses. The calibration module includes calibration features disposed on the motion platform. The imaging module is used to acquire calibration images of the calibration features and the target pixel position of the target chip in the imaging coordinate system. The processor is communicatively connected to the die-bonding device and is configured to execute the method described in the first aspect.

[0017] In summary, the beneficial effects of the present invention are as follows: The integrated circuit visual positioning calibration method and medium based on pulse displacement reference provided in this invention uses the actual number of pulses output by the pulse drive and encoding controller inside the die bonding equipment as the displacement reference. During controlled movement along the first and second directions, a calibration image set is simultaneously acquired, ensuring a correspondence between pixel change data and actual displacement data within the same control link. Therefore, calibration no longer relies on external gauges or manual alignment judgment; the calibration process can be automatically executed and data acquired by the equipment, reducing uncertainties caused by human intervention and improving the repeatability and consistency of calibration results.

[0018] Secondly, the deviation between the preset pulse number and the actual pulse number is explicitly introduced into the calibration dataset to further generate pulse deviation data, which, together with pixel change data, is used to solve for the rotational offset angle between the motion coordinate system and the imaging coordinate system. This design enables the calibration model to reflect the actual response at the motion execution level, rather than being derived solely from ideal displacement or simple image displacement. This reduces systematic mapping deviations caused by motion platform execution errors, transmission backlash, or control drift, thereby improving the reliability of the rotational offset angle solution.

[0019] Finally, in the production process, this invention uses the calculated rotation offset angle to compensate for the conversion relationship between the imaging coordinate system and the motion coordinate system, enabling the visual positioning output to correct coordinate system inconsistencies before being converted into motion commands. This compensation mechanism can suppress cross-coupling errors caused by camera mounting misalignment or coordinate system misalignment, reduce offset, rotation, or cumulative errors during die bonding alignment, improve the die bonding positioning accuracy and yield of target chips, including integrated circuits, and allow the calibration results to directly serve subsequent engineering applications in die bonding control. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments of the present invention will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, and these are all within the protection scope of the present invention.

[0021] Figure 1 This is a flowchart illustrating an integrated circuit visual positioning calibration method based on a pulse displacement reference according to an embodiment of the present invention.

[0022] Figure 2 This is a flowchart illustrating step S1 of an embodiment of the present invention.

[0023] Figure 3 This is a flowchart illustrating step S2 of an embodiment of the present invention. Detailed Implementation

[0024] The features and exemplary embodiments of various aspects of the present invention will now be described in detail. To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only configured to explain the present invention and are not configured to limit the present invention. For those skilled in the art, the present invention can be practiced without some of these specific details. The following description of the embodiments is merely intended to provide a better understanding of the present invention by illustrating examples of the invention.

[0025] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes said element.

[0026] Example 1 This invention provides an integrated circuit visual positioning and calibration method based on a pulse displacement reference, using a die-bonding device. The die-bonding device includes a motion control module, an imaging module, and a calibration module. The motion control module includes a motion platform and an encoding controller. The calibration module includes calibration features placed on the motion platform. The encoding controller controls the motion platform to move in at least two preset directions based on a pulse-driven method. The die bonding equipment used in this embodiment of the invention is an integrated integrated circuit die bonding equipment. Its hardware structure is designed around the coordinated design of visual positioning and precision motion control. It mainly includes a motion control module, an imaging module, and a calibration module. Each module works together to provide basic support for subsequent visual positioning calibration based on pulse displacement reference.

[0027] The motion control module is used to achieve precise displacement control during the calibration and die bonding processes. This module includes at least a motion platform and an encoder controller. The motion platform, as a load-bearing and motion-execution component, supports the calibration features and the integrated circuit chip, and performs linear or planar motion along a preset direction under the control of the encoder controller. The encoder controller is connected to the drive unit and position feedback unit of the motion platform, controlling the displacement of the motion platform through pulse driving and outputting pulse data corresponding to the actual motion process in real time, thereby providing a reliable data source for obtaining the displacement reference.

[0028] The imaging module is used to acquire images of the calibration features and the target chip's position before and after motion. It may include an industrial camera, lens, and a matching light source. The imaging module is fixedly mounted above or to the side of the die bonding equipment, and its imaging coordinate system has a fixed but unknown spatial relationship with the motion coordinate system of the motion platform. The calibration images acquired by the imaging module provide a basis for subsequent extraction of pixel position changes of the calibration features.

[0029] The calibration module provides a calibration reference that can be stably identified by the imaging module. It includes calibration features and is disposed on the motion platform. The calibration features can be geometric identifiers of a preset shape and size, which maintain a fixed positional relationship with the motion platform during its follow-up motion. This ensures that pixel changes in the calibration features within the imaging module's field of view accurately reflect the actual displacement of the motion platform.

[0030] Through the coordinated operation of the aforementioned hardware structures, this embodiment of the invention can directly utilize the motion control and imaging capabilities of the die bonding device itself to complete the subsequent visual positioning and calibration process without the need for additional external measuring devices, providing a hardware foundation for achieving high-precision, compensable integrated circuit die bonding visual positioning.

[0031] Please see Figure 1-3 The method includes: S1. Control the motion platform to move from a preset starting point along the first direction and the second direction according to the preset number of pulses, and obtain a calibration data set. The calibration data set includes a calibration image set and the actual number of pulses. The first direction and the second direction are perpendicular. The motion platform is controlled to move in two mutually perpendicular directions by a preset pulse number-driven encoder controller. A calibration image set is acquired before and after the movement, and the corresponding actual pulse count is read to form a calibration data set. The key to this step is to utilize the device's internal pulse drive to achieve repeatable displacement input and record the response changes of calibration features in the imaging coordinate system using the calibration image set. This allows for the simultaneous acquisition of displacement input from the motion side and displacement output from the visual side within the same set of calibration data, providing a foundation for establishing the correspondence between pixel space and motion space.

[0032] S2. Based on the calibration data set, obtain pulse deviation data and pixel change data; Based on the aforementioned calibration dataset, two types of core data are extracted: pulse deviation data and pixel change data. Pulse deviation data, obtained from the difference between the preset pulse number and the actual pulse number, reflects the deviation of the motion platform during controlled movement. Pixel change data, derived from the pixel position changes of calibration features in the calibration image set, reflects the projection of the same controlled movement onto the imaging coordinate system. By acquiring both types of data simultaneously, the motion execution state can be linked to the visual measurement results in subsequent solutions, avoiding systematic errors introduced by relying solely on ideal displacement or single image displacement derivations.

[0033] S3. Based on the pulse deviation data and pixel change data, obtain the rotation offset angle between the motion coordinate system of the motion platform and the imaging coordinate system of the imaging module; Pulse deviation data and pixel variation data are used as joint inputs to calculate the rotational offset angle between the motion coordinate system and the imaging coordinate system. Specifically, two mutually perpendicular controlled movement directions are typically represented as projection directions with a certain angle in the imaging coordinate system. By modeling and fitting the directional projection relationship of multiple sets of calibration data, the rotational offset angle between the two coordinate systems can be obtained. This step can automatically identify cross-coupling caused by camera mounting angle or coordinate system inconsistency, so that subsequent conversions no longer assume that the imaging coordinate axis and the motion coordinate axis are naturally aligned, thus eliminating positioning errors caused by coordinate system misalignment at the source.

[0034] S4. When performing die bonding on the target chip including the integrated circuit, the conversion relationship between the imaging coordinate system and the motion coordinate system is compensated according to the rotation offset angle.

[0035] The obtained rotational offset angle is applied to the die-bonding process to compensate for the conversion between the imaging coordinate system and the motion coordinate system when aligning target chips, including integrated circuits, during die-bonding. In other words, when the imaging module provides the pixel displacement or pixel coordinate difference of the target chip, the displacement is first compensated for by coordinate transformation based on the rotational offset angle, and then converted into the pulse displacement command that the motion platform should execute, combined with parameters such as visual equivalent. This ensures that the actual alignment direction of the motion platform is consistent with the visual measurement direction. This compensation significantly reduces the systematic error introduced by the camera mounting angle, improves die-bonding alignment accuracy and consistency, and ultimately enhances the yield and stability of the integrated circuit packaging process.

[0036] Preferably, the step of controlling the motion platform to move from a preset starting point along a first direction and a second direction according to a preset number of pulses to obtain a calibration data set includes: S11. Control the motion platform to move the calibration feature to the center region of the field of view of the imaging module; The control motion platform moves the calibration features to the center of the imaging module's field of view. By placing the calibration features at the center of the field of view, the impact of lens distortion, edge imaging errors, and uneven illumination on pixel positioning accuracy can be effectively reduced, making the subsequently acquired calibration images more representative and stable, thereby improving the reliability of pixel change data.

[0037] S12. Control the motion platform to move from a preset starting point along a first direction according to the preset pulse count and preset moving speed; The motion platform is controlled to move from a preset starting point along a first direction based on a preset number of pulses and a preset moving speed. This step involves applying a known pulse input in the first direction to induce a controllable displacement of the motion platform. The preset number of pulses serves as a theoretical displacement input, providing a reference benchmark for subsequent analysis of the relationship between the motion execution state and the visual response. Simultaneously, using a preset moving speed helps to standardize motion conditions and avoid introducing additional dynamic errors due to speed differences.

[0038] S13. Control the imaging module to acquire the first starting point image of the calibration features before motion and the first ending point image after motion, and simultaneously acquire the first actual pulse count of the encoder controller after motion; The control imaging module acquires the first starting point image of the calibration features before motion and the first ending point image after motion, and reads the first actual pulse count output by the encoder controller after motion is completed. By acquiring images before and after motion, the pixel position changes of the calibration features in the imaging coordinate system can be accurately obtained; while the synchronously acquired first actual pulse count is used to reflect the actual execution of the motion platform in this controlled movement, thus making the image-side data correspond one-to-one with the motion-side data.

[0039] In one embodiment, during the controlled movement of the motion platform, the motion control module employs an S-curve acceleration / deceleration control method to control the start and stop processes of the motion platform. This ensures a smooth transition in speed between acceleration, constant speed, and deceleration phases, thereby avoiding mechanical vibration, impact, or positioning overshoot caused by sudden starts and stops. This acceleration / deceleration control method effectively improves the motion stability and repeatability of the motion platform during calibration.

[0040] Simultaneously, after the motion platform reaches the starting and ending points of the displacement, a preset stabilization delay is added before image acquisition. This allows the motion platform to remain stationary for a period of time after completing the displacement, waiting for the mechanical vibration to fully decay before the imaging module acquires the image. This stabilization delay further reduces the impact of residual vibration on the calibration image quality and pixel positioning accuracy, thereby improving the reliability and consistency of the acquired calibration data.

[0041] S14. After the motion platform returns to the preset starting point, control the motion platform to move from the preset starting point along the second direction according to the preset number of pulses and the preset moving speed; After the motion platform returns to the preset starting point, it is again controlled to move along the second direction according to the preset number of pulses and the preset moving speed. By performing the same controlled movement in the second direction, which is perpendicular to the first direction, two sets of independent and orthogonal displacement data can be generated, providing the necessary conditions for subsequent analysis of the directional relationship between the motion coordinate system and the imaging coordinate system.

[0042] S15. Control the imaging module to acquire the second starting point image of the calibration features before motion and the second ending point image after motion, and simultaneously acquire the second actual pulse count of the encoder controller after motion; The control imaging module acquires the second starting point image of the calibration features before motion and the second ending point image after motion, and acquires the second actual pulse count of the encoding controller after motion is completed. This step corresponds to step S13 and is used to acquire pixel change data and actual pulse execution data under controlled movement in the second direction, thereby constructing a complete two-dimensional calibration data foundation.

[0043] Wherein, the preset pulse number is greater than the preset pulse number threshold, and the preset moving speed is less than the preset low speed threshold.

[0044] Furthermore, by limiting the preset number of pulses to be greater than a preset pulse number threshold, the motion displacement amplitude can be ensured to be large enough to reduce the relative impact of image noise and subpixel positioning error on the results; by limiting the preset moving speed to be less than a preset low speed threshold, the impact of acceleration and deceleration shocks and mechanical vibrations on motion stability and image acquisition quality can be reduced, thereby further improving the accuracy and repeatability of calibration data.

[0045] Preferably, the preset starting point includes a preset number of designated starting point positions distributed at different locations on the motion platform. After controlling the imaging module to acquire the second starting point image of the calibration features before motion and the second ending point image after motion, and simultaneously acquiring the second actual pulse count of the encoding controller after motion, the method further includes: S16. Control the motion platform to move to another designated starting position, and return to the step of controlling the motion platform to move from the preset starting point along the first direction according to the preset pulse number and preset moving speed, until the designated starting positions are traversed and a calibration data set corresponding to each designated starting position is obtained.

[0046] In step S16, after completing one controlled movement along the first and second directions and acquiring the starting point image, ending point image, and second actual pulse count corresponding to the second direction, the motion platform is controlled to move the calibration feature to another designated starting point position. The designated starting point position can be distributed in different areas within the field of view of the imaging module to cover different positional states that the motion platform may encounter during actual operation.

[0047] Subsequently, at the new designated starting point, the controlled movement steps in the first and second directions, based on a preset number of pulses and a preset moving speed, are executed again, and the corresponding calibration images and actual pulse counts are acquired simultaneously. By repeating the above calibration process at multiple different starting points, multiple sets of independent but structurally consistent calibration data can be obtained.

[0048] This step effectively reduces the impact of random errors caused by a single initial position, making the calibration results independent of any specific location conditions. Furthermore, multi-position calibration data can be used for fitting and consistency analysis in subsequent parameter solving processes, thereby improving the reliability of the rotation offset angle calculation between the motion coordinate system and the imaging coordinate system, and enhancing the overall calibration accuracy.

[0049] Preferably, obtaining pulse deviation data and pixel change data based on the calibration data set includes: S21. Based on the differences between the first actual pulse count and the second actual pulse count and the preset pulse count, obtain the preset number of pulse deviation data, wherein the pulse deviation data includes pulse deviation data corresponding to the first direction and the second direction respectively; In step S21, pulse deviation data is obtained by calculating the difference between the first actual pulse count corresponding to the first direction and the second actual pulse count corresponding to the second direction and the preset pulse count, respectively. The pulse deviation data reflects the difference between the actual displacement and the theoretical input displacement of the motion platform during controlled movement, and distinguishes the motion states in the first and second directions. By introducing pulse deviation data, the problem of ignoring motion execution level errors by using only the preset displacement as the ideal input can be avoided, enabling the subsequent calculation of calibration parameters to truly reflect the actual response characteristics of the motion control system.

[0050] S22. Based on the calibration image set corresponding to each specified starting point position, obtain the preset number of pixel change data, wherein the pixel change data includes pixel change data corresponding to the first direction and the second direction respectively.

[0051] In step S22, for each specified starting point position corresponding to the calibration image set, the pixel position changes of the calibration features in the imaging coordinate system are extracted to obtain a preset number of pixel change data. These pixel change data correspond to controlled movements in the first and second directions, respectively. To improve the stability and accuracy of the pixel change data, in practice, a preset stabilization delay can be added at the displacement starting point and the displacement ending point, allowing the motion platform to wait for mechanical vibration to decay before image acquisition, thus avoiding the influence of transient vibrations on the image positioning results.

[0052] During pixel location extraction, the calibration features can be located in both the pre- and post-motion images, and sub-pixel edge extraction algorithms can be used to accurately determine the feature edges or centers. For example, grayscale centroid methods, edge curve fitting, or contour fitting methods can be employed to precisely calculate the coordinates (u1, v1) and (u2, v2) of the feature points in the two images, thereby obtaining the pixel displacement components. To further suppress the effects of random noise and illumination fluctuations, multiple frames of images can be continuously acquired at the same specified starting point, and the feature coordinates corresponding to the multiple frames can be averaged to improve the consistency of the pixel displacement data.

[0053] Furthermore, for sub-pixel localization optimization, feature localization methods based on Gaussian fitting or surface fitting can be employed to fit and solve the edge response or central region grayscale distribution of the calibrated features, improving the feature point localization accuracy to the order of 0.1 pixels or higher, thereby further enhancing the accuracy and repeatability of pixel change data. Through the above processing, stable and reliable pixel change data can be obtained, providing a high-quality data foundation for subsequent calculation of the rotational offset angle between the motion coordinate system and the imaging coordinate system based on multiple sets of data.

[0054] Preferably, obtaining the rotational offset angle between the motion coordinate system of the motion platform and the imaging coordinate system of the imaging module based on the pulse deviation data and pixel change data includes: S31. Based on the pulse deviation data, obtain the actual displacement reference data corresponding to the controlled movement in the first direction and the second direction; The actual displacement reference data corresponding to the controlled movement in the first and second directions is determined based on the pulse deviation data. This actual displacement reference data is obtained by combining the preset number of pulses with the actual pulse execution, and is used to characterize the true displacement input of the motion platform in the two orthogonal directions. By using the displacement data corrected for pulse deviation as the reference, systematic errors introduced by using only ideal or commanded displacement as a reference can be avoided, allowing subsequent coordinate relationship analysis to be based on the actual execution state of the motion control system.

[0055] S32. Based on the pixel change data, obtain the pixel displacement components of the controlled movement in the first direction and the second direction in the imaging coordinate system. Pixel displacement components in the imaging coordinate system are extracted based on pixel change data, representing controlled movement in the first and second directions. These pixel displacement components reflect the actual projection changes of the calibration features in the imaging coordinate system when the motion platform moves in the corresponding direction. By distinguishing the pixel displacement components in the first and second directions, two independent directional responses can be obtained, providing fundamental data for analyzing the projection relationship of the motion directions in the imaging coordinate system.

[0056] S33. Based on the directional correspondence between the actual displacement reference data and the pixel displacement components, construct an overdetermined mapping model to characterize the projection relationship between the first direction and the second direction in the imaging coordinate system. Based on the directional correspondence between the aforementioned actual displacement reference data and pixel displacement components, an overdetermined mapping model is constructed to characterize the projection relationship between the first and second directions in the imaging coordinate system. This overdetermined mapping model introduces multiple sets of calibration data from different specified starting positions, holistically modeling the displacement input in orthogonal directions of the motion coordinate system and the projection output in the imaging coordinate system. This avoids excessive influence of single data points on the results and improves the model's ability to suppress noise and local errors.

[0057] S34. Solve the overdetermined mapping model using the least squares fitting method to obtain the local rotation offset angle corresponding to each specified starting position; In this step, the overdetermined mapping model is not constructed based on a single calibration position, but rather constructed and solved separately for each specified starting position. Specifically, at each specified starting position, the motion platform performs multiple sets of controlled movements along the first and second directions, corresponding to a set of independent actual displacement reference data and pixel displacement components. Using the multiple sets of calibration data corresponding to the specified starting position as constraints, the overdetermined mapping model is solved using a least-squares fitting method to obtain the rotational offset angle reflecting the relative relationship between the motion coordinate system and the imaging coordinate system at the specified starting position. This rotational offset angle is the local rotational offset angle corresponding to the specified starting position.

[0058] During actual operation, the spatial position of the motion platform in the die bonding equipment may be affected by assembly errors, minor deformations of the guide rails, or platform attitude deviations. Therefore, the relative rotational relationship between the motion coordinate system and the imaging coordinate system may not be entirely consistent at different specified starting points. By solving for the local rotational offset angle at each specified starting point, the problem of local errors masking the true extent of errors caused by simply merging all calibration data can be avoided. This ensures that each local rotational offset angle accurately reflects the relative coordinate system relationship at the corresponding spatial position.

[0059] Based on this, the obtained local rotation offset angles at each specified starting point not only characterize the rotational characteristics under different spatial positions, but also provide basic data support for subsequent filtering, weighting, or fusion of each local rotation offset angle based on consistency judgment rules. This method of solving at different positions and then processing them uniformly can balance overall stability and local differences under the constraints of multiple sets of calibration data, thus providing more reliable parameter basis for rotation compensation in subsequent visual positioning and motion control calculations, further improving the positioning accuracy and consistency during integrated circuit die bonding.

[0060] Preferably, the overdetermined mapping model is as follows: b=A*p In the formula, vector b is used to characterize the pixel displacement components of the controlled movement in the first and second directions in the imaging coordinate system, matrix A is used to characterize the actual displacement reference data corresponding to the controlled movement in the first and second directions, and vector p is used to characterize the set of transformation parameters between the motion coordinate system and the imaging coordinate system. The set of transformation parameters includes the rotation offset angle and the pulse equivalent. The pulse equivalent is used to characterize the pulse displacement corresponding to the unit pixel displacement of the motion platform along the first and second directions.

[0061] In this embodiment, the pulse equivalent is a pre-determined known parameter. The pulse equivalent is used to characterize the actual displacement corresponding to a unit pulse of the motion platform. It can be obtained through factory calibration, linear calibration during installation and commissioning, or pre-calculated based on encoder resolution and mechanical transmission parameters. Before performing this step, the pulse equivalent has been stored in the control system as a basic motion parameter and remains unchanged during the calibration process. It is not used as a variable to be solved in the parameter estimation of the overdetermined mapping model.

[0062] In the above embodiments, the overdetermined mapping model b = A * p is introduced to uniformly model the relationship between the motion coordinate system and the imaging coordinate system. Its technical meaning and function are explained below.

[0063] Here, vector b is used to characterize the pixel displacement component of the calibration feature in the imaging coordinate system when the motion platform moves in a controlled manner along the first and second directions. This pixel displacement component is extracted from the aforementioned pixel change data, reflects the projection result of the actual motion in the imaging coordinate system, and is a direct output of the visual measurement side.

[0064] Matrix A represents the actual displacement reference data corresponding to the controlled movement in the first and second directions. This actual displacement reference data is derived from the actual number of pulses output by the encoder controller, and is determined by combining a preset number of pulses and pulse deviation. It reflects the true displacement input of the motion platform in the two orthogonal directions. By incorporating the actual displacement reference data into the model, it is ensured that the mapping relationship is established based on the actual state of motion execution, rather than ideal or assumed displacement conditions.

[0065] Vector p is used to characterize the set of transformation parameters between the motion coordinate system and the imaging coordinate system. This set of transformation parameters includes at least a rotation offset angle and a pulse equivalent. The rotation offset angle describes the relative rotation relationship between the motion coordinate system and the imaging coordinate system; the pulse equivalent characterizes the pulse displacement relationship corresponding to the unit pixel displacement when the motion platform moves along the first direction and the second direction, thus reflecting the scale correspondence between the pixel space and the pulse space.

[0066] By incorporating multiple sets of actual displacement reference data and their corresponding pixel displacement components into a unified mathematical framework and solving the transformation parameter set as a whole, the rotational offset angle and related scale parameters between the motion coordinate system and the imaging coordinate system can be accurately obtained while suppressing the influence of random noise and local errors. This model provides a clear mathematical foundation for subsequent least-squares fitting to solve the transformation parameters and enables the calibration results to directly serve the visual positioning and motion control conversion in the subsequent integrated circuit die bonding process.

[0067] S35. Based on the local rotation offset angle of each specified starting point position and the preset consistency judgment rule, obtain the rotation offset angle between the motion coordinate system of the motion platform and the imaging coordinate system of the imaging module.

[0068] The local rotation offset angle in this step refers to the rotation offset angle obtained from multiple sets of calibration data at each specified starting position. Because the motion platform may have assembly deviations, minor changes in guide rail attitude, or differences in local forces at different travel positions, the relative rotation relationship between the motion coordinate system and the imaging coordinate system may exhibit slight inconsistencies at different specified starting positions. Therefore, after obtaining multiple local rotation offset angles, it is necessary to further determine the final rotation offset angle used for die bonding compensation through a preset consistency judgment rule. This consistency judgment rule can be understood as a set of criteria used to measure whether each local rotation offset angle is stably distributed around the same true rotation relationship, and based on this, decide whether to directly fuse them to obtain a unified angle, or to exclude abnormal angles before determining the result.

[0069] In implementation, the local rotation offset angles at each specified starting point are first compared and analyzed to generate measurable differences. This can be achieved by calculating the angle difference between any two angles or the deviation of each angle from the candidate center value, thus obtaining an index reflecting the degree of dispersion. This dispersion index is then compared with a preset consistency threshold to obtain a consistency judgment result. If the result is consistent, the final rotation offset angle is determined according to preset rules. For example, averaging or weighted fusion of each local rotation offset angle yields a rotation offset angle that better represents the overall trend. If the result is inconsistent, the local rotation offset angles are processed according to preset rules. For example, angles with deviations exceeding the threshold are removed, or their weights are reduced and re-fused until the consistency condition is met, at which point the final rotation offset angle is output. This comparison and filtering fusion method avoids the adverse effects of angle anomalies at individual specified starting points caused by image extraction errors or instantaneous motion perturbations on the final calibration result.

[0070] This step converges scattered local rotational offset angles into rotational offset angles suitable for engineering applications under consistency rules, making coordinate rotation compensation for target chip pixel displacement more stable and reliable during subsequent die bonding. On one hand, the fused rotational offset angles more accurately reflect the relative rotation relationship between the imaging coordinate system and the motion coordinate system within the actual working range, reducing angle fluctuations introduced by random noise. On the other hand, by identifying and suppressing abnormal local angles, the consistency of calibration results across different locations and batches during die bonding can be improved, thereby reducing the accumulation of visual positioning conversion errors and improving the positioning accuracy and yield of integrated circuits during die bonding.

[0071] Preferably, the rotational offset angle between the motion coordinate system of the motion platform and the imaging coordinate system of the imaging module is obtained based on the local rotational offset angle at each specified starting point position and a preset consistency judgment rule, including: S351. Obtain the local rotation offset angle corresponding to each specified starting point position to obtain a set of local angles; This step summarizes the local rotation offset angles corresponding to each specified starting point position, forming a set of local angles. This provides an input basis for subsequent judgments on whether these angles are distributed around the same true rotation relationship, which is beneficial for obtaining stable and usable calibration results even when there are slight differences in different spatial positions.

[0072] S352. For each local rotation offset angle in the set of local angles, calculate the least squares residual of the overdetermined mapping model and calculate the residual index of the local rotation offset angle. Least squares residuals refer to the error between the predicted and actual observed values ​​of each set of calibration data when solving an overdetermined mapping model at a specified starting position. The residual index can be understood as a quantitative representation of this error, such as a numerical value reflecting the overall fitting error level. By calculating the residual index for each local rotation offset angle, the interpretability of that angle for its corresponding calibration data can be objectively evaluated, avoiding judgment based solely on the angle's magnitude, thus providing a measurable basis for subsequent confidence weight allocation.

[0073] S353. Based on the residual index, obtain the confidence weight of each local rotation offset angle, wherein the confidence weight is negatively correlated with the residual index; The confidence weight is used to characterize the reliability of each local rotation offset angle. The smaller the residual index, the better the local angle fits its corresponding multiple sets of calibration data, and the higher its reliability. Therefore, a larger confidence weight is assigned to it. The setting of negative correlation enables the system to automatically weaken the impact of local angles with large fitting errors on the final result, reducing the risk of angle offset caused by image extraction jitter, local motion perturbations, or single data anomalies.

[0074] S354. Based on the confidence weight, perform weighted fusion processing on the local angle set to obtain candidate rotation offset angles; Weighted fusion processing refers to allocating the contribution of each angle within a set of local angles according to its corresponding confidence weight, thereby obtaining a candidate rotation offset angle that represents the overall trend. This candidate angle is not simply averaged, but rather tends to be dominated by high-confidence angles, thus obtaining a more stable estimate under constraints of multiple locations and multiple sets of data. This approach retains the advantages of multi-starting-point information while suppressing the influence of low-quality data on the results, making the candidate rotation offset angle closer to the actual relative rotation relationship of the device's coordinate system.

[0075] S355. Calculate the deviation of each local rotation offset angle in the local angle set relative to the candidate rotation offset angle to obtain the deviation set; The deviation is used to characterize the degree of difference between each local rotation offset angle and the candidate rotation offset angle, and can be understood as the quantification of the angle difference. By calculating the deviation set, the discrete distribution of each local angle relative to the candidate angle can be made explicit, providing direct input for subsequent consistency threshold judgment. The purpose of this step is to identify whether there are angles in the local angle set that deviate significantly from the overall trend, and to avoid abnormal estimations at individual locations from corrupting the final calibration results.

[0076] S356. Compare the set of deviations with a preset consistency threshold to obtain a consistency judgment result; A preset consistency threshold is used to define the acceptable range of angular deviations, while the consistency judgment result indicates whether the current local angle set meets the overall consistency requirement. When comparing the set of deviations with the consistency threshold, it can be determined according to preset rules whether there is a deviation exceeding the threshold, or whether the number exceeding the threshold exceeds the allowable proportion, thus outputting a consistent or inconsistent judgment result. Through this threshold-based judgment, the process of determining the rotational offset angle can be elevated from a simple evaluation to a verifiable quality control process, improving the reliability of the calibration results in engineering applications.

[0077] S357. If the consistency judgment result meets the preset consistency condition, the candidate rotation offset angle is taken as the final rotation offset angle.

[0078] If the consistency judgment result meets the preset consistency conditions, the candidate rotation offset angle is output as the final rotation offset angle. This means that the candidate angle can represent the overall trend of most high-confidence data under weighted fusion, and also meets the stability requirements through the consistency threshold check. The final rotation offset angle can be directly used to perform coordinate rotation compensation on the target chip pixel displacement during the die bonding stage, and further converted into pulse displacement commands in the motion coordinate system, thereby reducing the direction error in the pixel-to-motion command conversion and improving the die bonding positioning accuracy and consistency.

[0079] In a preferred embodiment, when the consistency judgment result of step S356 does not meet the preset consistency condition, the process of determining the rotation offset angle is not directly terminated. Instead, an iterative processing mechanism based on confidence weight adjustment is introduced to further optimize the local rotation offset angle set in order to improve the stability and reliability of the final rotation offset angle.

[0080] In this embodiment, the method further includes the following after step S357: S358: If the consistency judgment result does not meet the preset consistency condition, reduce the angle confidence weight corresponding to the local rotation offset angle whose deviation is greater than the preset consistency threshold, and return to execute steps S354 to S356 until the consistency judgment result meets the preset consistency condition or the number of iterations reaches the preset upper limit.

[0081] Specifically, if the consistency judgment result does not meet the preset consistency condition, local rotation offset angles with deviations greater than the preset consistency threshold are first identified in the deviation set. For the identified local rotation offset angles, it is considered that their corresponding calibration results deviate significantly from the current overall rotation trend, possibly due to local image extraction errors, instantaneous motion disturbances, or abnormal calibration data. For these local rotation offset angles, their corresponding confidence weights are reduced, or adjusted to values ​​below the preset lower limit, thereby reducing the impact of these local rotation offset angles on subsequent fusion processes.

[0082] After adjusting the confidence weights, steps S354 to S356 are re-executed based on the updated confidence weight set. This involves re-weighting and fusing the local angle set to obtain new candidate rotation offset angles. The deviation of each local rotation offset angle relative to the candidate rotation offset angle is then recalculated, and a consistency judgment is performed again. Through this method, the candidate rotation offset angles gradually converge towards the direction dominated by high-confidence local angles.

[0083] The above iterative process can be repeated until the consistency judgment result meets the preset consistency condition, or the number of iterations reaches the preset upper limit. When the consistency judgment result finally meets the preset consistency condition, the currently obtained candidate rotation offset angle is output as the final rotation offset angle; when the number of iterations reaches the upper limit and the consistency condition is still not met, the current candidate rotation offset angle can be output as an approximate rotation offset angle, or the process of re-acquiring the calibration data set can be triggered to ensure that the system still has controllable processing results under abnormal conditions.

[0084] Through the above iterative processing scheme, the interference of abnormal local rotation offset angles on the overall result can be gradually suppressed in the case of local abnormal calibration results. The final rotation offset angle is mainly determined by the calibration data with small fitting residuals and high stability, thereby improving the consistency and reliability of the rotation offset angle under different spatial positions and different calibration batches, and further improving the accuracy of visual positioning and motion control conversion in the integrated circuit die bonding process.

[0085] Preferably, the step of compensating for the conversion relationship between the imaging coordinate system and the motion coordinate system based on the rotation offset angle during die bonding of the target chip including the integrated circuit includes: S41. Obtain the target pixel position of the target chip in the imaging coordinate system according to the imaging module, and determine the pixel displacement of the target chip; The imaging module acquires the target pixel position of the target chip in the imaging coordinate system, and determines the pixel displacement of the target chip relative to the reference position. This pixel displacement characterizes the offset of the target chip in the image plane, is a direct result of visual positioning output, and provides input data for subsequent motion conversion.

[0086] S42. Based on the rotation offset angle, perform coordinate rotation compensation processing on the pixel displacement to obtain the compensated displacement. Based on the previously obtained rotation offset angle, coordinate rotation compensation is performed on the pixel displacement to obtain the compensated displacement. This step corrects the pixel displacement in the imaging coordinate system to a displacement expression consistent with the direction of the motion coordinate system, thereby eliminating the directional coupling error introduced by the rotational deviation between the imaging and motion coordinate systems, and ensuring that the visual measurement results match the actual motion direction.

[0087] S43. Based on the compensated displacement and pulse equivalent, obtain the pulse displacement command in the motion coordinate system; Based on the compensated displacement and pulse equivalent, pulse displacement commands in the motion coordinate system are obtained. The pulse equivalent is used to characterize the pulse displacement relationship corresponding to a unit pixel displacement. By converting the compensated displacement into the number of pulses, control commands that the motion control module can execute can be directly generated, enabling the visual positioning results to be seamlessly converted into the actual drive input of the motion platform.

[0088] S44. Control the motion platform to move according to the pulse displacement command to complete the die bonding of the target chip including the integrated circuit.

[0089] The motion platform is controlled by the aforementioned pulse displacement command to perform corresponding displacement actions to complete the die bonding operation on the target chip, including the integrated circuit. By introducing rotational offset angle compensation during the die bonding process, the alignment action of the motion platform can accurately correspond to the visual positioning result, thereby improving the positioning accuracy and consistency during the chip die bonding process and reducing the mounting deviation and defect rate caused by coordinate system inconsistency. In another embodiment, when the global consistency judgment result does not meet the preset consistency condition, the solution result of the rotational offset angle is not directly rejected. Instead, the determination process of the rotational offset angle is switched from a global single parameter to partitioned parameterized modeling to adapt to the possible differences in local rotational relationships that may exist in different spatial regions of the motion platform.

[0090] S3781: If the global consistency judgment result does not meet the preset consistency condition, the working area of ​​the motion platform is divided according to the spatial distribution relationship of each specified starting point position on the motion platform to obtain at least two position partitions.

[0091] Specifically, the spatial position identifiers or coordinate information of each designated starting point position in the motion coordinate system of the motion platform are first obtained, and the working area is divided according to a preset partitioning rule. The partitioning rule can be based on the segmentation of the motion platform's travel direction, or it can be based on the grid division method on the working plane of the motion platform to form multiple adjacent regions, so that each position partition covers a continuous travel segment or a continuous working surface. By introducing spatial partitioning, the determination of subsequent rotation offset angles no longer depends on the global consistency assumption, but allows the establishment of more realistic local rotation relationships in different spatial regions, thereby providing a basis for spatial inconsistency compensation in large-stroke die bonding equipment.

[0092] S3782: Map each specified starting point position to the corresponding position partition to form a set of partition starting points corresponding to each position partition.

[0093] Specifically, for each specified starting point location, its corresponding location partition is determined based on its spatial location identifier and the coverage area of ​​each location partition. The identifier of the specified starting point location is then written into the corresponding partition starting point set. If a specified starting point location exists near the partition boundary, it can be assigned to a location partition with a closer distance or a higher coverage ratio according to a preset boundary assignment rule, or it can be simultaneously marked as a boundary starting point for subsequent processing. Through the above mapping process, the originally mixed local rotation offset angle samples are classified according to spatial regions, ensuring that the data used by each partition has clear spatial homogeneity during partition modeling, and reducing the interference of cross-regional differences on the solution of partition rotation offset angles.

[0094] S3783: For each location partition, extract the local rotation offset angle from the partition start point set corresponding to that location partition to form a partition local angle set.

[0095] Specifically, based on the starting point identifiers recorded in the partition starting point set, angle samples corresponding one-to-one with the partition starting point set are selected from the obtained local rotation offset angle set and combined to form the partition local angle set. To improve the reliability of the partition angle solution, auxiliary quality information of each angle sample can be retained simultaneously in this step, such as its corresponding fitting residual index or pixel displacement validity indicator, so as to perform quality screening or weighting processing within the partition later. By forming the partition local angle set, the angle samples within each location partition can reflect the common rotation trend of the spatial region, providing a data basis for determining the partition rotation offset angle.

[0096] S3784: Calculate the partition angle dispersion index based on the local angle set of each partition, and determine the partition rotation offset angle of the corresponding position partition if the partition angle dispersion index meets the partition consistency condition.

[0097] Specifically, for each local angle set of a partition, the partition angle dispersion index is first calculated and compared with a preset partition dispersion threshold to determine whether the partition has usable consistency. If the partition angle dispersion index meets the partition consistency condition, the partition rotation offset angle is determined based on the local angle set of that partition. The partition rotation offset angle can be the statistical center value of the set or a representative angle determined with the minimum deviation as the target within the partition. If the partition angle dispersion index of a certain partition does not meet the partition consistency condition, further subdivision of the partition can be triggered, or the partition can be marked as a low-confidence partition for a more conservative strategy in the subsequent compensation stage. By applying consistency constraints again within the partition, the direct introduction of local abnormal angles into the partition compensation parameters can be avoided, thus ensuring that the partition rotation offset angle is both spatially specific and maintains sufficient stability and usability.

[0098] S3785: Establish a correlation between the rotation offset angle of each position partition and the position partition, and generate a set of position-related rotation compensation parameters.

[0099] Specifically, each position partition is assigned a partition identifier, and this partition identifier is mapped to the corresponding partition rotation offset angle and stored to obtain a set of position-related rotation compensation parameters. During die bonding, the position partition to which the target chip belongs can be determined based on its spatial position or the position of the motion platform calculated from the target pixel position in the imaging coordinate system. The corresponding partition rotation offset angle is then selected for coordinate rotation compensation of the pixel displacement. By establishing a partition parameter mapping relationship, rotation compensation is upgraded from a single global parameter to a parameter system that adaptively selects parameters based on spatial position. This maintains the compensation accuracy in each region even with global inconsistencies, thereby improving the consistency and reliability of die bonding positioning under long-stroke, multi-position conditions.

[0100] Based on this embodiment, steps S42~S44 are adjusted as follows: The corresponding position partition is determined based on the target pixel position of the target chip, and the partition rotation offset angle matching the position partition is obtained. Based on the partition rotation offset angle, the pixel displacement is subjected to coordinate rotation compensation processing to obtain the compensated displacement. Based on the compensated displacement and pulse equivalent, a pulse displacement command in the motion coordinate system is generated; Specifically, based on the position of the target pixel in the imaging coordinate system, the working area of ​​the motion platform corresponding to the target chip is determined, and then its position partition is determined. The partition rotation offset angle corresponding to this position partition is selected from the set of position-related rotation compensation parameters. If the global consistency judgment result meets the preset consistency condition, then a unified rotation offset angle is selected as the partition rotation offset angle.

[0101] Subsequently, the pixel displacement is rotated and compensated based on the selected partition rotation offset angle, so that the compensated displacement is consistent with the motion coordinate system in the direction, reducing the directional coupling error caused by the coordinate system rotation deviation and improving the positioning consistency under different spatial positions.

[0102] Finally, by converting the compensated displacement into the corresponding number of pulses, a pulse displacement command that the motion control module can directly execute is generated to drive the motion platform to complete the positioning and die bonding of the target chip.

[0103] Preferably, obtaining the pulse displacement command in the motion coordinate system based on the compensated displacement and pulse equivalent includes: S431. Obtain the difference between the current ambient temperature and the preset temperature value to get the temperature deviation data; This step obtains the current ambient temperature of the die bonding equipment and compares it with a preset temperature value to obtain temperature deviation data. This temperature deviation data reflects the temperature change of the equipment's operating environment relative to the calibration or reference state, providing a basis for subsequent pulse equivalent temperature compensation. By introducing the ambient temperature factor, the adverse effects of thermal expansion and contraction of the mechanical structure or changes in transmission characteristics caused by temperature variations on the accuracy of displacement conversion can be avoided.

[0104] S432. Obtain the pulse equivalent compensation value based on the temperature deviation data and the preset temperature compensation coefficient; Based on the temperature deviation data and the preset temperature compensation coefficient, the pulse equivalent compensation value is calculated. The temperature compensation coefficient is used to characterize the sensitivity of the pulse equivalent to temperature changes. This step converts the temperature change into a correction value for the pulse equivalent, thereby achieving quantitative modeling of the environmental impact.

[0105] S433. Compensate the pulse equivalent according to the pulse equivalent compensation value to obtain the compensated pulse equivalent; The original pulse equivalent is compensated and corrected based on the above pulse equivalent compensation value to obtain the compensated pulse equivalent. The compensated pulse equivalent can more realistically reflect the actual displacement characteristics of the motion platform under the current environmental conditions, making the generation of subsequent pulse displacement commands closer to the actual response of the motion platform.

[0106] S434. Based on the compensated displacement and the compensated pulse equivalent, obtain the pulse displacement command.

[0107] Based on the compensated displacement and the compensated pulse equivalent, the pulse displacement command for controlling the motion platform is calculated and obtained. By introducing the temperature-compensated pulse equivalent in this step, the pulse displacement command converted from the visual positioning result can maintain high consistency and accuracy under different environmental conditions, thereby further improving the positioning accuracy and stability during the integrated circuit die bonding process.

[0108] Preferably, before compensating for the conversion relationship between the imaging coordinate system and the motion coordinate system based on the rotation offset angle when performing die bonding on the target chip including the integrated circuit, the method further includes: S01. Control the motion platform to perform test movement in the first and second directions according to the preset test pulse displacement; The motion platform is controlled to perform test movements in the first and second directions according to preset test pulse displacements. These test pulse displacements can be known and repeatable displacement inputs, used to simulate typical motion states of the motion platform during actual production, thus providing a reference for subsequent verification.

[0109] S02. Obtain pixel displacement data of the target features before and after the test movement through the imaging module; The imaging module acquires pixel displacement data of the target features in the imaging coordinate system before and after the test movement. This pixel displacement data reflects the actual response of the target features in the image plane when the motion platform performs the test pulse displacement, and serves as the visual basis for verifying the effectiveness of the calibration parameters.

[0110] S03. Based on the pixel displacement data, the rotation offset angle, and the pulse equivalent, the corresponding predicted pulse displacement is calculated in reverse. Based on the pixel displacement data, and combined with the previously obtained rotation offset angle and pulse equivalent, the predicted pulse displacement is calculated in reverse. This reverse calculation process is used to verify whether the motion command deduced from the visual measurement results can be consistent with the known test input under given calibration parameters.

[0111] S04. Compare the predicted pulse displacement with the test pulse displacement to obtain the calibration error; The predicted pulse displacement is compared with the actual executed test pulse displacement to obtain the corresponding calibration error. This calibration error is used to quantify the degree of deviation of the calibration parameters in the actual motion-imaging conversion process and is a direct indicator of the accuracy of the calibration results.

[0112] S05. If the calibration error is less than a preset error threshold, the calibration result is determined to be valid.

[0113] When the calibration error is less than a preset error threshold, the current calibration result is deemed valid, and the process is allowed to proceed to the subsequent production process, including die bonding and coordinate compensation of the target chip containing the integrated circuit. This verification step allows for a validity self-check of the calibration results before they are officially put into production, thereby avoiding systematic positioning errors caused by calibration anomalies and improving the stability and reliability of the integrated circuit die bonding process.

[0114] Example 2 Furthermore, in conjunction with the integrated circuit visual positioning and calibration method based on pulse displacement reference in the above embodiments, this invention can be implemented using a computer-readable storage medium. This computer-readable storage medium stores computer program instructions; when executed by a processor, these computer program instructions implement any of the integrated circuit visual positioning and calibration methods based on pulse displacement reference in the above embodiments.

[0115] It should be clarified that the present invention is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of the present invention.

[0116] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0117] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0118] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0119] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0120] It should also be noted that the exemplary embodiments mentioned in this invention describe methods or systems based on a series of steps or apparatus. However, this invention is not limited to the order of the steps described above; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0121] The above description is merely a specific embodiment of the present invention. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the protection scope of the present invention.

Claims

1. A visual positioning calibration method for integrated circuits based on pulse displacement reference, characterized in that, A die bonding device is used, the die bonding device comprising: a motion control module, an imaging module, and a calibration module, wherein the motion control module includes a motion platform and an encoding controller, the calibration module includes calibration features placed on the motion platform, and the encoding controller controls the motion platform to move in at least two preset directions based on a pulse-driven method, the method comprising: The motion platform is controlled to move from a preset starting point along a first direction and a second direction according to a preset number of pulses, and a calibration data set is obtained. The calibration data set includes a calibration image set and the actual number of pulses. The first direction and the second direction are perpendicular. Based on the calibration data set, obtain pulse deviation data and pixel change data; Based on the pulse deviation data and pixel change data, the rotation offset angle between the motion coordinate system of the motion platform and the imaging coordinate system of the imaging module is obtained; When performing die bonding on a target chip including an integrated circuit, coordinate rotation compensation processing is performed on the pixel displacement of the target chip according to the rotation offset angle to generate pulse displacement commands in the motion coordinate system to control the movement of the motion platform.

2. The integrated circuit visual positioning calibration method based on pulse displacement reference according to claim 1, characterized in that, The step of controlling the motion platform to move from a preset starting point along a first direction and a second direction according to a preset number of pulses, and acquiring a calibration data set, includes: The motion platform is controlled to move the calibration feature to the center region of the field of view of the imaging module; The motion platform is controlled to move from a preset starting point along a first direction according to the preset pulse count and preset moving speed; The imaging module is controlled to acquire the first starting point image of the calibration features before motion and the first ending point image after motion, while the first actual pulse count of the encoder controller is acquired after motion. After the motion platform returns to the preset starting point, the motion platform is controlled to move from the preset starting point along the second direction according to the preset number of pulses and the preset moving speed; The imaging module is controlled to acquire the second starting point image of the calibration features before motion and the second ending point image after motion, while acquiring the second actual pulse count of the encoder controller after motion; Wherein, the preset pulse number is greater than the preset pulse number threshold, and the preset moving speed is less than the preset low speed threshold.

3. The integrated circuit visual positioning calibration method based on pulse displacement reference according to claim 2, characterized in that, The preset starting point includes a preset number of designated starting point positions distributed at different locations on the motion platform. After controlling the imaging module to acquire the second starting point image of the calibration features before motion and the second ending point image after motion, and simultaneously acquiring the second actual pulse count of the encoding controller after motion, the method further includes: The process involves controlling the motion platform to move to another designated starting position and returning to the step of controlling the motion platform to move from the preset starting point along the first direction according to the preset number of pulses and the preset moving speed, until all designated starting positions have been traversed, and a calibration data set corresponding to each designated starting position is obtained.

4. The integrated circuit visual positioning calibration method based on pulse displacement reference according to claim 3, characterized in that, The step of obtaining pulse deviation data and pixel change data based on the calibration data set includes: Based on the differences between the first actual pulse count and the second actual pulse count and the preset pulse count, the preset number of pulse deviation data is obtained, wherein the pulse deviation data includes pulse deviation data corresponding to the first direction and the second direction respectively; Based on the calibration image set corresponding to each specified starting position, the preset number of pixel change data is obtained, wherein the pixel change data includes pixel change data corresponding to the first direction and the second direction respectively.

5. The integrated circuit visual positioning calibration method based on pulse displacement reference according to any one of claims 1-4, characterized in that, The step of obtaining the rotational offset angle between the motion coordinate system of the motion platform and the imaging coordinate system of the imaging module based on the pulse deviation data and pixel change data includes: Based on the pulse deviation data, obtain the actual displacement reference data corresponding to the controlled movement in the first direction and the second direction; Based on the pixel change data, obtain the pixel displacement components of the controlled movement in the first and second directions in the imaging coordinate system. Based on the directional correspondence between the actual displacement reference data and the pixel displacement components, an overdetermined mapping model is constructed to characterize the projection relationship between the first direction and the second direction in the imaging coordinate system. The overdetermined mapping model is solved using the least squares fitting method to obtain the local rotation offset angles corresponding to each specified starting position; Based on the local rotation offset angle at each specified starting point position and the preset consistency judgment rule, the rotation offset angle between the motion coordinate system of the motion platform and the imaging coordinate system of the imaging module is obtained.

6. The integrated circuit visual positioning calibration method based on pulse displacement reference according to claim 5, characterized in that, Based on the local rotation offset angles at each specified starting point position and a preset consistency judgment rule, the rotation offset angle between the motion coordinate system of the motion platform and the imaging coordinate system of the imaging module is obtained, including: Obtain the local rotation offset angle corresponding to each specified starting position to obtain a set of local angles; For each local rotation offset angle in the set of local angles, the least squares solution residual of the overdetermined mapping model is used to calculate the residual index of the local rotation offset angle. Based on the residual index, the confidence weight of each local rotation offset angle is obtained, wherein the confidence weight is negatively correlated with the residual index; Based on the confidence weight, the local angle set is weighted and fused to obtain candidate rotation offset angles; Calculate the deviation of each local rotation offset angle in the local angle set relative to the candidate rotation offset angle to obtain the deviation set; The set of deviations is compared with a preset consistency threshold to obtain a consistency judgment result; If the consistency judgment result meets the preset consistency conditions, the candidate rotation offset angle is taken as the final rotation offset angle.

7. The integrated circuit visual positioning calibration method based on pulse displacement reference according to any one of claims 1-4, characterized in that, When performing die bonding on a target chip including an integrated circuit, the process of performing coordinate rotation compensation processing on the pixel displacement of the target chip according to the rotation offset angle, and generating pulse displacement commands in the motion coordinate system to control the movement of the motion platform, includes: The imaging module obtains the target pixel position of the target chip in the imaging coordinate system and determines the pixel displacement of the target chip. Based on the rotation offset angle, the pixel displacement is subjected to coordinate rotation compensation processing to obtain the compensated displacement. Based on the compensated displacement and pulse equivalent, obtain the pulse displacement command in the motion coordinate system; The motion platform is controlled to move according to the pulse displacement command in order to complete the die bonding of the target chip including the integrated circuit.

8. The integrated circuit visual positioning calibration method based on pulse displacement reference according to claim 7, characterized in that, The step of obtaining the pulse displacement command in the motion coordinate system based on the compensated displacement and pulse equivalent includes: Obtain the difference between the current ambient temperature and the preset temperature value to get the temperature deviation data; Based on the temperature deviation data and the preset temperature compensation coefficient, the pulse equivalent compensation value is obtained; The pulse equivalent is compensated according to the pulse equivalent compensation value to obtain the compensated pulse equivalent; The pulse displacement command is obtained based on the compensated displacement and the compensated pulse equivalent.

9. The integrated circuit visual positioning calibration method based on pulse displacement reference according to claim 7, characterized in that, Before compensating for the conversion relationship between the imaging coordinate system and the motion coordinate system based on the rotation offset angle when performing die bonding on the target chip including the integrated circuit, the method further includes: The motion platform is controlled to perform test movements in the first and second directions according to the preset test pulse displacement; The imaging module acquires pixel displacement data of the target features before and after the test movement. Based on the pixel displacement data, the rotation offset angle, and the pulse equivalent, the corresponding predicted pulse displacement is calculated in reverse. The predicted pulse displacement is compared with the test pulse displacement to obtain the calibration error; If the calibration error is less than a preset error threshold, the calibration result is deemed valid.

10. An integrated circuit visual positioning calibration system based on pulse displacement reference, characterized in that, The system includes a die-bonding device and a processor. The die-bonding device includes a motion control module, an imaging module, and a calibration module. The motion control module includes a motion platform and an encoding controller. The encoding controller controls the motion platform to move in at least two preset directions based on a pulse-driven method and outputs the actual number of pulses. The calibration module includes calibration features set on the motion platform. The imaging module is used to acquire calibration images of the calibration features and the target pixel position of the target chip in the imaging coordinate system. The processor is communicatively connected to the die-bonding device and is configured to execute the integrated circuit visual positioning calibration method based on pulse displacement reference as described in any one of claims 1-9.

Images