A multi-station visual guidance automatic feeding and discharging method and control system

Abstract

The application discloses a kind of multi-station visual guidance's automatic feeding and discharging method and control system, belong to the field of semiconductor manufacturing and testing, through top station preliminary screening positioning, front position accurate positioning and deviation calculation, reverse position independent review, the multi-stage visual guidance process of test seat position target confirmation, in combination with high-precision motion control, realize the whole process real-time compensation and correction of chip from grabbing to placing, ±0.1mm of repeat positioning accuracy is realized, the damage of chip and test Socket is avoided, and rapid changeover and whole process data traceability can be realized by parameter configuration, significantly improve test yield, equipment reliability and comprehensive efficiency.

Description

Technical Field

[0001] This invention relates to the field of semiconductor manufacturing and testing, and in particular to a multi-station vision-guided automatic loading and unloading method and control system. Background Technology

[0002] In semiconductor back-end manufacturing, chips must undergo a series of rigorous electrical and reliability tests after packaging to screen out potentially defective products and ensure the quality and long-term reliability of finished products. Among these, the High-Accelerated Stress Test (HAST) is a crucial step in evaluating the reliability of chips under harsh environments of high temperature, high humidity, and high pressure. Currently, the pre-test material preparation mainly faces the following challenges:

[0003] (1) Inherent defects of manual and semi-automatic modes: Traditional methods rely heavily on operators to manually place the chip into the corresponding test socket on the test board (such as the HAST board). This process is extremely inefficient and cannot meet the requirements of batch testing capacity. At the same time, human judgment and manual operation are prone to placement deviations, angle shifts or improper force, which not only lead to poor test contact and inaccurate results, but also easily scratch the chip pins or damage the expensive test socket, causing direct economic losses.

[0004] (2) Limitations of existing automation solutions: Currently proposed automated loading and unloading equipment typically uses relatively simple vision positioning systems, often performing only a single, coarse positioning from a single angle. These solutions lack the ability to assess the front and back faces of chips and their multi-degree-of-freedom spatial orientation. It also has the ability to systematically and collaboratively test the integrity of the socket (such as whether there are foreign objects or whether the pins are deformed). Since it is impossible to perform closed-loop visual feedback and compensation at multiple key stations such as picking, transfer, and placement, its actual repeatability positioning accuracy is difficult to consistently reach ±0.1mm. This accuracy threshold is crucial to ensuring reliable contact between multi-pin chips such as Quad Flat No-leads Packages (QFN) and the socket.

[0005] (3) Insufficient flexibility and diagnostic capabilities: Existing equipment is complex to change and debug when faced with chips of different sizes and packages and test boards of various specifications, and has poor adaptability. At the same time, there is a general lack of complete process diagnosis and data traceability functions. When misoperation or test failure occurs, it is difficult to quickly locate whether the problem is with the chip, the placement, or the socket itself, which affects the overall efficiency of the equipment and the production yield.

[0006] Therefore, there is an urgent need in this field for a fully automated loading and unloading solution with high integration, multi-stage visual positioning and correction capabilities, flexible operation, and intelligent diagnostic functions, in order to break through the current bottlenecks in production capacity and quality in chip testing, especially in high-reliability testing. Summary of the Invention

[0007] The purpose of this invention is to overcome the problems of the prior art and provide a multi-station vision-guided automatic loading and unloading method and control system.

[0008] The objective of this invention is achieved through the following technical solution: a multi-station vision-guided automatic loading and unloading method, which includes the following steps:

[0009] The chip is transferred to the top station, a top image of the chip is acquired, the front and back sides are determined and screened based on the top image, and the front-facing chips are initially positioned.

[0010] The control actuator picks up the chip based on the preliminary positioning results and moves the chip to the front workstation. It acquires a front image of the chip, calculates the first current position of the chip based on the front image, and calculates the first position deviation between the first current position and the standard position; and / or, calculates the first current angle of the chip based on the front image, and calculates the first angle deviation between the first current angle and the standard angle.

[0011] The chip is transferred to the reverse side station, the reverse side image of the chip is acquired, the second current position and / or the second current angle of the chip are independently calculated based on the reverse side image, and the first current position and the second current position are cross-validated, and / or the first current angle and the second current angle are cross-validated to generate a verification result;

[0012] Acquire test socket images, detect residual chips on the test socket based on the test socket images, and calculate the current position of the test socket;

[0013] Based on the first deviation result, the verification result, and the current position of the test socket, the control actuator compensates and corrects the placement position and / or angle of the chip, and places the chip into the test socket; the first deviation result includes the first position deviation and / or the first angle deviation.

[0014] In one example, the step of determining and filtering the front and back sides based on the top image, and initially locating the filtered front-facing chips, includes:

[0015] The top image is processed to identify the first chip region, extract the outline of the first chip, and calculate the center coordinates of the first pixel of the chip.

[0016] Extract character or texture features from the chip surface and match them with a preset front character texture template to determine and filter the front and back of the chip;

[0017] Based on the camera calibration model, the center coordinates of the first pixel of the front-facing chip are converted into preliminary world coordinates in the mechanical coordinate system, thus completing the initial positioning.

[0018] In one example, calculating the first current position of the chip based on the frontal image includes:

[0019] The front image is processed to identify the second chip region, extract the outline of the second chip, and calculate the center coordinates of the second pixel of the chip as the first current position;

[0020] And / or, calculate the first current angle of the chip based on the frontal image, including:

[0021] Identify a first preset feature mark on the frontal image, and calculate the angle of the first preset feature mark in the frontal image as the first current angle;

[0022] And / or, independently calculate the second current position of the chip based on the reverse image, including:

[0023] The reverse image is processed to identify the third chip region, extract the outline of the third chip, and calculate the center coordinates of the third pixel of the chip as the second current position;

[0024] And / or, independently calculate the chip's second current angle based on the reverse image, including:

[0025] Identify a second preset feature mark on the reverse image, and calculate the angle of the second preset feature mark in the image as the second current angle.

[0026] In one example, the step of cross-validating the first current position with the second current position, and / or cross-validating the first current angle with the second current angle to generate a verification result includes:

[0027] Calculate the second position deviation between the first current position and the second current position, and / or calculate the second angle deviation between the first current angle and the second current angle;

[0028] The second position deviation and / or the second angle deviation are compared with preset position thresholds and angle thresholds, respectively.

[0029] If the second position deviation does not exceed the position threshold and / or the second angle deviation does not exceed the angle threshold, a verification result that passes the verification is generated; otherwise, a verification result that fails the verification is generated.

[0030] In one example, after acquiring a frontal image of the chip, the process also includes:

[0031] Batch verification and identity recognition are performed based on the frontal image;

[0032] And / or, perform surface defect detection based on the frontal image.

[0033] In one example, after acquiring an image of the reverse side of the chip, the process also includes:

[0034] Batch verification and identity recognition based on the reverse image; and / or,

[0035] Pin inspection is performed based on the reverse image, including at least one of integrity inspection, damage and defect inspection.

[0036] In one example, the method further includes a teaching step:

[0037] The initial material placement coordinates obtained from visual positioning are adjusted based on human visual feedback to generate fine-tuned coordinates;

[0038] The actuator, controlled by the fine-tuned coordinates, carries the reference chip and descends to the preset height of the test stand.

[0039] The actuator is controlled to descend at a speed below the speed threshold, while real-time pressure data associated with the test seat is collected simultaneously.

[0040] If the real-time pressure data is within the preset safe pressure threshold range, the fine-tuned coordinates will be used as the initial photo position of the test station; if the real-time pressure data is not within the preset safe pressure threshold range, the movement will be stopped and an alarm will be triggered.

[0041] It should be further noted that the technical features corresponding to the above examples can be combined or replaced to form new technical solutions.

[0042] The present invention also includes a multi-station vision-guided automatic loading and unloading control system for executing the loading and unloading method formed by any or a combination of the above examples, the system comprising:

[0043] The visual inspection module includes a top station unit, a front station unit, a back station unit, and a test socket station unit. The top station unit is used to acquire a top image of the chip, determine and filter the front and back sides based on the top image, and perform preliminary positioning on the filtered front-facing chips. The front station unit is used to acquire a front image of the chip, calculate a first positional deviation between the chip's first current position and a standard position based on the front image, and / or calculate a first angular deviation between the chip's first current angle and a standard angle based on the front image. The back station unit is used to acquire a back image of the chip, independently calculate a second current position and / or a second current angle based on the back image, cross-validate the first current position and the second current position, and / or cross-validate the first current angle and the second current angle to generate a verification result. The test socket station unit is used to acquire a test socket image, detect residual chips on the test socket based on the test socket image, and calculate the current position of the test socket.

[0044] The motion control module, connected to the vision inspection module, is used to control the actuator to pick up the chip according to the preliminary positioning result and move the chip to the front station and the back station. The motion control module is also used to control the actuator to compensate and correct the placement position and / or angle of the chip according to the first deviation result, the verification result and the current position of the test socket, and to put the chip into the test socket. The first deviation result includes the first position deviation and / or the first angle deviation.

[0045] In one example, the system further includes:

[0046] The pressure sensing module is located below the test stand and is used to collect real-time pressure data of the test stand.

[0047] The teaching subsystem, connected to the motion control module and pressure sensing module, adjusts the initial feeding coordinates, generates fine-tuned coordinates, and feeds them back to the motion control module. The motion control module then controls the actuator to move the reference chip according to these fine-tuned coordinates, lowering it to the preset height of the test stand. Simultaneously, the motion control module controls the actuator to descend at a speed below a threshold. The teaching subsystem simultaneously collects real-time pressure data associated with the test stand. If the real-time pressure data is within the preset safe pressure threshold range, the fine-tuned coordinates are used as the initial image capture position for the test stand. If the real-time pressure data is outside the preset safe pressure threshold range, the teaching subsystem feeds back the abnormal result to the motion control module, causing the motion control module to stop the actuator's movement and trigger an alarm.

[0048] In one example, the system further includes:

[0049] The host computer is connected to the motion control module and the vision inspection module. It is used to send task instructions to the motion control module and the vision inspection module, and to receive the result data returned by the motion control module and the vision inspection module, and to perform task management and scheduling. The host computer also stores production parameter files, vision parameter files and hardware board files for different workstations.

[0050] The multi-station parallel processing module is connected to the host computer module and is used to load and independently configure the corresponding production parameter files, vision parameter files and hardware board files from the host computer for the top station, front station and back station respectively. The hardware board file stores the standard position and standard angle.

[0051] It should be further noted that the technical features corresponding to the above system examples can be combined or replaced to form new technical solutions.

[0052] Compared with the prior art, the beneficial effects of the present invention are:

[0053] 1. In one example, through multi-stage and multi-dimensional visual guidance including initial positioning, front-side fine positioning, back-side verification positioning, and test socket fine positioning, the real-time position angle of the chip at each station is accurately compared and verified with the pre-calibrated standard position angle. Combined with high-precision motion control, sub-millimeter-level real-time compensation and correction are achieved for the entire process of chip picking, transfer, and final placement, achieving a repeatability accuracy of ±0.1mm and effectively avoiding damage to chip pins and test sockets. At the same time, the standard position and standard angle are adapted to different chip models, realizing production flexibility. The front, back, and test socket stations all generate recordable first position deviation, first angle deviation, verification results, and the current position of the test socket, providing accurate data traceability for the entire process. This enables rapid diagnosis and location of problem links, thereby significantly improving test yield, equipment reliability, and overall efficiency.

[0054] 2. Through cross-validation, it is possible to effectively identify chips with abnormal positioning caused by misjudgment at a single workstation or mechanical drift, thereby avoiding batch errors and greatly improving the robustness of the system and production safety.

[0055] 3. Through multi-level chip quality inspection (identity verification, surface defects, pin quality inspection), a comprehensive quality screening from appearance to function is formed, and the initial quality screening is completed at the same time as positioning, which further improves the test yield.

[0056] 4. Subpixel-level manual fine-tuning of the initial feeding coordinates is performed through physical observation, eliminating static system errors such as mechanical installation and fixture tolerances. At the same time, real-time pressure monitoring and threshold judgment are performed during the low-speed descent process, which can immediately stop and alarm when alignment deviation or foreign objects occur, effectively preventing damage to chips and expensive test sockets, and significantly improving the reliability, safety and service life of the equipment. Attached Figure Description

[0057] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. The accompanying drawings are provided to provide a further understanding of the present application and constitute a part of the present application. The same reference numerals are used in these drawings to denote the same or similar parts. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation of the present application.

[0058] Figure 1 This is a flowchart illustrating a method provided as an example of the present invention. Detailed Implementation

[0059] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0060] In the description of this invention, the use of ordinal numbers (e.g., "first to third", etc.) is for distinguishing objects and is not limited to that order, and should not be construed as indicating or implying relative importance.

[0061] Furthermore, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0062] In one example, such as Figure 1 As shown, a multi-station vision-guided automatic loading and unloading method includes the following steps:

[0063] S1: Transfer the chip to the top station, acquire the top image of the chip, determine and filter the front and back based on the top image, and perform preliminary positioning on the selected front-facing chips.

[0064] In step S1, the front and back of the chip are quickly identified, the chips with the front facing up that can be picked up are selected, and preliminary positioning is performed. The preliminary positioning results are output to provide a basic position reference for the subsequent fine positioning operation at the front station.

[0065] S2: The control actuator picks up the chip based on the preliminary positioning results and moves the chip to the front workstation, acquires a front image of the chip, calculates the first position deviation between the first current position and the standard position of the chip based on the front image, and / or calculates the first angle deviation between the first current angle and the standard angle of the chip based on the front image.

[0066] In step S2, preferably, the first positional deviation between the chip's first current position and the standard position is calculated based on the frontal image, and the first angular deviation between the chip's first current angle and the standard angle is calculated based on the frontal image. The standard position and standard angle are derived from experiments or historical experience. At this point, the chip's placement position and angle can be compensated and corrected using the first positional deviation and the first angular deviation, further improving positioning accuracy.

[0067] S3: Transfer the chip to the reverse side station, acquire the reverse side image of the chip, independently calculate the second current position and / or the second current angle of the chip based on the reverse side image, and cross-verify the first current position and the second current position, and / or cross-verify the first current angle and the second current angle to generate a verification result.

[0068] In step S3, it is preferable to independently calculate the second current position and the second current angle of the chip based on the reverse image. At this time, the first current position and the second current position, the first current angle and the second current angle are cross-validated. The chip position is then independently verified a second time based on the chip's reverse features to ensure the reliability of the positioning result.

[0069] S4: Acquire test socket images, detect residual chips on the test socket based on the test socket images, and calculate the current position of the test socket.

[0070] In step S4, before unloading, residual chip detection is performed on the test socket to further accurately locate the precise position of the test socket, thereby placing the chip into the test socket without deviation, which further improves the positioning accuracy of unloading and loading.

[0071] S5: Based on the first deviation result, the verification result, and the current position of the test socket, control the actuator to compensate and correct the placement position and / or angle of the chip, and place the chip into the test socket; the first deviation result includes the first position deviation and / or the first angle deviation.

[0072] In step S5, the preferred control actuator compensates and corrects the placement position and angle of the chip to ensure positioning accuracy to the greatest extent. At this time, combining the mechanical model of the actuator with the first deviation result, the verification result, and the current position of the test seat, the micro-compensation amount required during the final unloading is calculated. This compensation amount is used to correct the accumulated errors caused by mechanical errors, thermal drift, or transmission gaps that may accumulate throughout the entire process from picking up to unloading, ensuring positioning stability during the gripping and transfer process.

[0073] This invention utilizes a multi-station visual positioning mechanism to achieve a closed-loop, high-precision guidance process from initial chip screening, precise positioning, reverse verification to test socket alignment. Each visual station has a clearly defined function and works collaboratively, forming a complete quality and precision control closed loop from chip screening and multi-angle positioning to socket status confirmation. This ensures long-term stable system operation and provides multi-stage visual positioning and correction capabilities, achieving a repeatability accuracy of ±0.1mm, guaranteeing reliable contact between QFN and other multi-pin chips and the test socket. Furthermore, this invention adapts to the corresponding standard positions and angles of different chip models, enabling production flexibility. The front, back, and test socket stations all generate recordable first position deviations, first angle deviations, verification results, and the current position of the test socket, providing precise data traceability for the entire process. This allows for rapid diagnosis and location of problematic steps, significantly improving test yield, equipment reliability, and overall efficiency.

[0074] In one example, the front and back sides are determined and filtered based on the image on the top of the chip, and the front-facing chips are initially located, including:

[0075] S11: Process the top image to identify the first chip region, extract the outline of the first chip, and calculate the center coordinates of the first pixel of the chip.

[0076] Specifically, the acquired top image is first preprocessed and binarized. The preprocessing includes image denoising, contrast enhancement, and correction of uneven lighting. The binarization process highlights the difference between the chip and the background. Furthermore, all independent first chip regions in the field of view are identified through Blob analysis (connected component analysis). For each independent first chip region, edge detection is performed to accurately extract the first chip contour. By calculating the minimum bounding rectangle or geometric centroid of the first chip region contour, the first pixel-level center coordinates of the chip are determined, ensuring that even if the chip has slight rotation or misalignment, it can be accurately segmented and located.

[0077] By employing subpixel edge detection and contour feature point analysis, combined with packaging shape features such as characters and special marking dots on the front surface of the chip, stable front character feature extraction is achieved, unaffected by light fluctuations, slight occlusion, or surface contamination.

[0078] S12: Extract the character or texture features on the chip surface and match them with the preset front character texture template to realize the judgment and screening of the front and back of the chip.

[0079] Specifically, after determining the first pixel-level center coordinates of the chip, an image of the region containing characters or specific textures on the chip surface is extracted with these coordinates as the center. This region image is then matched with a pre-learned front-side character texture template using normalized cross-correlation. Chips with a matching score higher than a set threshold are determined to be front-side up and proceed to the next process; chips that fail to match are determined to be back-side up or abnormal, are marked as abnormal, and are guided to the discard process.

[0080] By combining normalized cross-correlation matching with feature point descriptor matching, high-precision matching is ensured even when the chip exhibits slight deformation or attitude changes, achieving sub-pixel accuracy and providing reliable input for subsequent motion compensation. Furthermore, this step integrates localization and screening, improving processing efficiency.

[0081] S13: Based on the camera calibration model, the center coordinates of the first pixel of the front-facing chip are converted into preliminary world coordinates in the mechanical coordinate system, thus completing the initial positioning.

[0082] Specifically, for chips determined to be facing upwards, the first pixel-level center coordinates of the chip are input into a pre-completed nine-point calibration model. This calibration model establishes a high-precision mapping relationship between the image pixel coordinate system and the robot's world coordinate system. Through model calculation and transformation, the precise world coordinates of the chip in the robot's coordinate system are finally output. These precise world coordinates serve as both the initial position of the center point of the dual suction nozzles (actuators) and the reference position for planning the path to move the chip to the front vision station for precise positioning.

[0083] Preferably, in step S11, only the top image is processed and the first chip area is identified. Subsequently, based on the front and back determination result in step S12, only the center coordinates of the first pixel of the chip facing upwards are calculated to reduce computational overhead.

[0084] The above-mentioned connected component analysis, template matching and filtering, and coordinate transformation collaborative processing not only efficiently completed the front and back screening, but also output a high-precision preliminary spatial position, laying a solid first-stage foundation for the entire system to achieve a comprehensive positioning accuracy of ±0.1mm.

[0085] In one example, calculating the chip's first current position based on the frontal image includes:

[0086] The front image is processed to identify the second chip region, extract the outline of the second chip, and calculate the center coordinates of the second pixel of the chip as the first current position;

[0087] Specifically, after the chip is initially positioned at the top workstation and moved to the front workstation, a high-resolution front camera acquires a front image of the chip under optimal lighting conditions. The top image is first preprocessed, and then the second chip region is quickly separated through connected component analysis. Subpixel-level edge detection is performed on this region to accurately extract the second chip contour, and its geometric center or the centroid of the main contour is calculated to obtain the precise second pixel center coordinates of the chip in the current field of view, which serves as the first current position.

[0088] Furthermore, the calculated first current position is compared with the standard position calibrated during the teaching phase to calculate the pixel-level deviation of the chip in the X and Y directions. This deviation is converted into a physical position deviation in the mechanical coordinate system using the pre-calibrated camera pixel resolution (the actual physical size represented by each pixel), resulting in the first position deviation. Based on this first position deviation and the mechanical offset parameters of the dual suction nozzles, the motion control module can calculate in real time the precise position compensation required by the nozzle end effector, thereby ensuring that the nozzles can pick up or place the chip with zero deviation.

[0089] In one example, calculating a first angular deviation between a first current angle of the chip and a standard angle based on a frontal image includes:

[0090] Identify a first preset feature mark on the frontal image, and calculate the angle of the first preset feature mark in the frontal image as the first current angle;

[0091] Specifically, a first preset feature mark, such as character information or a dedicated marker, is identified on the front side of the chip. Using template matching or feature point matching algorithms, the current angle of the first preset feature mark is compared with a standard angle to calculate the chip's rotational deviation angle around the Z-axis, which is then used as the first current angle. This angular deviation value is sent to the motion control module in real time to drive the U-axis (rotation axis) for automatic correction before picking up or placing the chip, ensuring strict alignment between the chip pins and the test socket.

[0092] Combining the above examples of front-facing image processing steps at the front-facing workstation, we can obtain the corrected precise center coordinates, rotation angle, character recognition results, and defect detection marks. The center coordinates and rotation angle are used to drive the motion system to complete high-precision correction and placement; the character and defect results are used for quality judgment, and defective products will be sorted and rejected in subsequent processes. This front-facing image processing method ensures that the chip can enter the next stage with a repeatability accuracy better than ±0.1mm and the correct posture, which is key to achieving the overall system accuracy and reliability goals.

[0093] In one example, the second current position of the chip is calculated independently based on the reverse image, including:

[0094] The reverse image is processed to identify the third chip region, extract the outline of the third chip, and calculate the center coordinates of the third pixel of the chip as the second current position;

[0095] Specifically, after the chip is transferred to the reverse side station, a complete positioning process is restarted based on the reverse side image. First, the reverse side image is preprocessed, and then the third chip region is quickly separated through connected component analysis. Subpixel-level edge detection is performed on this third chip region to accurately extract the third chip contour, and its geometric center or the center of the main contour is calculated to obtain the precise third pixel center coordinates of the chip in the current field of view, which is used as the second current position.

[0096] In one example, the second current angle of the chip is calculated independently based on the reverse image, including:

[0097] Identify a second preset feature mark on the reverse image, and calculate the angle of the second preset feature mark in the image as the second current angle.

[0098] Specifically, a second preset feature mark, such as character information or special marker point, is identified on the front side of the chip. The current angle of the second preset feature mark is compared with the standard angle through template matching or feature point matching algorithm, and the rotation deviation angle of the chip around the Z-axis is calculated as the second current angle.

[0099] Furthermore, this example is combined with the previous example. The second current position and second current angle are then compared with the expected position and angle (first current position and first current angle) transmitted from the front station to achieve cross-validation. During cross-validation, the center of the two nozzle stations is used to determine the positional and angular deviations of the nozzles picking up the chip's center. Compensation is then performed based on these deviations to ensure the dual-station nozzles pick up the chip's center. Theoretically, the center of the reverse camera should coincide with the chip's center. In practice, the reverse station detects a deviation in the chip's center point. If the difference from the camera center point is greater than 0.2mm, or the angle is greater than 1 degree, the chip is considered unqualified and placed in the waste bin. This effectively avoids batch errors caused by misjudgment at a single station or mechanical drift.

[0100] In one example, it is preferable to cross-validate the first current position with the second current position and the first current angle with the second current angle, including:

[0101] (1) Calculate the second position deviation between the first current position and the second current position, and calculate the second angle deviation between the first current angle and the second current angle;

[0102] (2) Compare the second position deviation and the second angle deviation with the preset position threshold and angle threshold, respectively;

[0103] (3) If the second position deviation does not exceed the position threshold and the second angle deviation does not exceed the angle threshold, a verification result of passing the verification is generated; otherwise, it is determined that the positioning process is abnormal or the chip attitude is unstable, a verification result of failing the verification is generated, and an alarm is triggered.

[0104] In one example, after acquiring a frontal image of the chip, the process also includes:

[0105] Batch verification and identity recognition are performed based on the frontal image.

[0106] Specifically, optical character recognition or specific character template matching technology is used to read the laser-engraved characters on the chip surface. The recognition results are compared with preset correct characters to determine whether the chip's model, batch, and other information are correct. Defective products will be sorted out and removed, achieving preliminary traceability and screening.

[0107] In one example, after acquiring a frontal image of the chip, the process also includes:

[0108] Surface defect detection based on frontal images.

[0109] Specifically, based on the same high-resolution frontal image, a defect detection algorithm is run to detect scratches, stains, and defects. By analyzing image texture, brightness anomalies, or contour integrity, possible defective areas are marked, and non-conforming products are sorted out and removed.

[0110] In one example, after acquiring an image of the reverse side of the chip, the process also includes:

[0111] Batch verification and identity recognition are performed based on the reverse image.

[0112] Specifically, the pin array texture or specific layout features on the reverse side of the chip are image sampled, and high-precision template matching is performed with a preset standard pin texture template for this chip model to calculate the matching similarity. This step is not only used to assist in positioning, but more importantly, it serves as a batch conformity verification: chips with high matching scores are confirmed as qualified products in the current batch; chips with low matching scores may be mixed in with chips of different models or abnormal batches, and will be marked and rejected.

[0113] In one example, after acquiring an image of the reverse side of the chip, the process also includes:

[0114] Pin inspection is performed based on the reverse image, including at least one of integrity inspection, damage and defect inspection.

[0115] Specifically, while completing the positioning, parallel pin quality inspection is performed on the same high-resolution reverse image: (1) Integrity inspection: by analyzing the connectivity and contour integrity of the pin area, it is determined whether there are fatal defects such as missing pins, bending or bridging; (2) Damage and defect analysis: using grayscale morphology or texture analysis algorithms, defects such as scratches, oxidation, stains or plating abnormalities on the pin surface are detected. This example performs simultaneous detection and classification of multiple defects such as missing pins, bending, bridging, scratches, etc. The detection results are bound and recorded with the positioning data, providing a complete data chain for subsequent quality traceability; using contour edge grayscale co-occurrence matrix texture analysis, feature modeling and matching of the pin array on the reverse side of the chip are performed to realize damage and defect analysis, which not only assists in positioning, but can also be used as a batch consistency verification method to effectively identify mixed materials, wrong materials or packaging abnormalities, and eliminate test confusion from the source.

[0116] Combining the above examples of reverse image processing steps involving the reverse side of the workstation, we can obtain the verified chip position and angle, pin texture matching score (for batch verification), final micro-compensation amount for material placement, and a detailed pin quality inspection report. This reverse image processing method ensures that only chips with accurate positioning, correct identification, and intact pins are allowed to be placed in the expensive test socket, fundamentally guaranteeing test reliability and equipment safety.

[0117] In one example, the method also includes a teaching step:

[0118] The initial material placement coordinates obtained from visual positioning are adjusted based on human visual feedback to generate fine-tuned coordinates;

[0119] The actuator, controlled by the fine-tuned coordinates, carries the reference chip and descends to the preset height of the test stand.

[0120] The actuator is controlled to descend at a speed below the speed threshold, while real-time pressure data associated with the test seat is collected simultaneously.

[0121] If the real-time pressure data is within the preset safe pressure threshold range, the fine-tuned coordinates will be used as the initial photo position of the test station; if the real-time pressure data is not within the preset safe pressure threshold range, the movement will be stopped and an alarm will be triggered.

[0122] After the operator completes initial coordinate fine-tuning based on the adjusted coordinates, the guide nozzle carrying the chip descends to a detection height (preset height) approximately 1mm from the target test socket, and then enters the pressure monitoring unloading mode. In this mode, real-time pressure data on the test socket on the hardware board is collected, and the pressure curve is monitored simultaneously. The nozzle descends slowly at an extremely low speed (below the speed threshold) to complete the final stroke, dynamically judging whether the pressure value is within the preset safety threshold. If the pressure data is stable and does not exceed the limit, it is determined that the chip is stress-free and accurately positioned, and this position is automatically saved as a high-precision unloading point, serving as the initial photo position for the test socket station to eliminate positional differences between test sockets on different hardware boards; if the pressure rises abnormally, it stops immediately and an alarm is triggered, indicating possible alignment deviation or socket abnormality. This step ensures that the deviation between the actual unloading position and the theoretical standard position is stable within ±0.1mm, and fundamentally prevents mechanical damage to the chip or socket caused by inaccurate placement.

[0123] Preferably, before the teaching step, the following steps are also included:

[0124] S00: Each motion axis of the actuator performs a high-precision zero-reset operation and establishes a precise coordinate system reference;

[0125] S01: Select the corresponding production parameter file, vision parameter file, and hardware board file for each workstation. The production parameter file defines the motion flow of each workstation, such as chip picking, conveying, and placement. The vision parameter file stores the algorithm templates, image acquisition light source settings, and detection thresholds required for image processing, feature matching, and defect detection at each vision workstation. The hardware board file records the standard position obtained after high-precision teaching calibration for each test socket on a specific physical hardware board, serving as a spatial reference for achieving ±0.1mm accuracy.

[0126] Combining the above examples, we obtain a preferred example of the present invention, in which the method includes the following steps:

[0127] S100: Each motion axis of the actuator performs a high-precision zero-reset operation and establishes a precise coordinate system reference;

[0128] S200: Select the corresponding production parameter file, vision parameter file, and hardware board file for each workstation;

[0129] S300: Adjusts the initial feeding coordinates obtained from visual positioning based on human visual feedback to generate fine-tuned coordinates; controls the actuator to carry the reference chip and descend to the preset height of the test stand based on the fine-tuned coordinates; controls the actuator to descend at a speed lower than the speed threshold, and simultaneously collects real-time pressure data associated with the test stand; if the real-time pressure data is within the preset safe pressure threshold range, the fine-tuned coordinates are used as the initial image position of the test stand station; if the real-time pressure data is not within the preset safe pressure threshold range, the movement is stopped and an alarm is triggered.

[0130] S400: Transfer the chip to the top station, acquire the top image of the chip, determine and filter the front and back based on the top image, and perform preliminary positioning on the filtered front-facing chips;

[0131] S500: The control actuator picks up the chip according to the preliminary positioning results and moves the chip to the front work station. It collects the front image of the chip and calculates the first position deviation between the first current position and the standard position and the first angle deviation between the first current angle and the standard angle based on the front image.

[0132] S600: Transfer the chip to the reverse side station, acquire the reverse side image of the chip, independently calculate the second current position and the second current angle of the chip based on the reverse side image, and cross-verify the first current position with the second current position and the first current angle with the second current angle to generate a verification result;

[0133] S700: Acquires test socket images, performs residual chip detection based on test socket images, and calculates the current position of the test socket;

[0134] S800: Based on the first position deviation, the first angle deviation, the verification result, and the current position of the test socket, control the actuator to compensate and correct the placement position and angle of the chip, and place the chip into the test socket.

[0135] The present invention also includes a multi-station vision-guided automatic loading and unloading control system for executing the loading and unloading method composed of any or more of the above examples. The system includes a vision detection module and a motion control module, and the vision detection module is connected to the motion control module.

[0136] The visual inspection module includes a top station unit, a front station unit, a back station unit, and a test socket station unit. Specifically, the top station unit is used to acquire a top image of the chip, determine and filter the front and back sides based on the top image, and perform preliminary positioning on the selected front-facing chips. The front station unit is used to acquire a front image of the chip, calculate a first positional deviation between the chip's first current position and a standard position based on the front image, and / or calculate a first angular deviation between the chip's first current angle and a standard angle based on the front image. The back station unit is used to acquire a back image of the chip, independently calculate a second current position and / or a second current angle based on the back image, cross-validate the first current position and the second current position, and / or cross-validate the first current angle and the second current angle to generate a verification result. The test socket station unit is used to acquire a test socket image, detect residual chips on the test socket based on the test socket image, and calculate the current position of the test socket.

[0137] The motion control module, as the core of the system, employs high-precision servo motors and drivers to control the X, Y, Z, and U-axis motion units of the actuator, ensuring a repeatability accuracy better than ±0.1mm. Specifically, the motion control module controls the actuator to pick up the chip based on the preliminary positioning results and move the chip to the front and back workstations. The motion control module also controls the actuator to compensate and correct the chip's placement position and / or angle based on the first deviation result, the verification result, and the current position of the test socket, and then places the chip into the test socket. The first deviation result includes a first positional deviation and / or a first angular deviation.

[0138] Preferably, the motion control module provides a manual device operation interface, allowing users to make fine adjustments to each motion axis by preset small movement distances, in order to help achieve and maintain a positioning accuracy of ±0.1mm.

[0139] This example achieves a repeatability accuracy of better than ±0.1mm for the grasping, transporting, and placing of chips through a high-precision motion control module and a vision inspection module.

[0140] In one example, the system also includes a pressure sensing module and a teaching subsystem, which are connected to the motion control module and the pressure sensing module. The pressure sensing module is located below the test stand and is used to collect real-time pressure data from the test stand. The teaching subsystem adjusts the initial feeding coordinates, generates fine-tuned coordinates, and feeds them back to the motion control module. The motion control module then controls the actuator to carry the reference chip and descend to a preset height on the test stand based on these fine-tuned coordinates. The motion control module also controls the actuator to descend at a speed below a speed threshold. Simultaneously, the teaching subsystem collects real-time pressure data associated with the test stand. If the real-time pressure data is within the preset safe pressure threshold range, the fine-tuned coordinates are used as the initial image capture position for the test stand. If the real-time pressure data is outside the preset safe pressure threshold range, the teaching subsystem feeds back the abnormal result to the motion control module, causing the motion control module to stop the actuator's movement and trigger an alarm.

[0141] Preferably, the teaching subsystem guides the user step by step through a graphical interface to automatically calibrate and save the material picking position, various visual image capture positions, and the material placement position, and provides a coordinate fine-tuning function based on visual feedback. At this time, the operator uses the coordinate fine-tuning function provided by the teaching subsystem, combined with the up and down movement of the nozzle, to perform fine compensation in the X and Y directions of the material placement position until the chip can be placed into the socket without deviation, and saves the high-precision position coordinates after compensation to the system.

[0142] In one example, the system also includes a host computer and a multi-station parallel processing module. The host computer is connected to the motion control module, the vision inspection module, and the multi-station parallel processing module. The host computer sends task instructions to the motion control module and the vision inspection module, and receives the result data returned by the motion control module and the vision inspection module, performing task management and scheduling. The host computer also stores production parameter files, vision parameter files, and hardware board files for different workstations. Preferably, the host computer is used for overall system scheduling, task management, and data interaction between modules. The multi-station parallel processing module supports simultaneous production at each independent workstation. The top workstation, front workstation, and back workstation load and independently configure their corresponding production parameter files, vision parameter files, and hardware board files from the host computer. The hardware board files store standard positions and standard angles.

[0143] In one example, the system also includes a parameter management subsystem for centralized management of the system's core parameter files throughout their entire lifecycle. This includes motion flow and control parameter files, hardware board position and pressure sensing threshold parameter files, and vision model and detection parameter files. This parameter management subsystem supports the creation, modification, deletion, version management, and cloning / reuse of these files. When importing a new chip model, engineers can quickly clone the parameter files based on similar models and only need to calibrate and adjust key differences in the teaching subsystem (such as updating the vision template, recalibrating the material placement position, and setting new pressure thresholds), thereby significantly shortening changeover and debugging time and improving the equipment's flexible production capabilities.

[0144] Combining the above system examples, we obtain a preferred example of the present invention, in which the system works as follows:

[0145] (1) After the system is started, the operator first runs the motion process and control parameter file. After the first use or after changing the model, the operator needs to click the zero reset button on the main interface to return each motion axis to the mechanical origin and establish a high-precision reference.

[0146] (2) The operator selects the preset motion flow and control parameter file, hardware board position and pressure sensing threshold parameter file, vision model and detection parameter file for each activated workstation, and enters the batch number. After confirmation, the system enters the teaching interface.

[0147] If the selected hardware board location and pressure sensor threshold parameter file are new models, a complete teaching process is required. The system guides you step-by-step through the graphical interface:

[0148] a. First, guide the device to the starting test position, call the Socket vision to take a picture, create an empty Socket material template, and accurately locate the Socket position.

[0149] b. Then, the camera moves sequentially to the top, front, and back view positions. At the top vision position, the light source and camera parameters are adjusted to establish a front / back judgment model and a preliminary positioning template. At the front vision position, the parameters are finely adjusted to establish a template for precise positioning and angle correction. At the back vision position, a template for pin detection and position verification is established.

[0150] c. The system calls the calibrated Socket vision for position verification. The operator can manually fine-tune the interface to perform sub-pixel-level compensation of the X / Y coordinates, while simultaneously observing the alignment of the chip and the Socket by moving the nozzle up and down, until the chip can be accurately placed without any obstruction. This high-precision coordinate is then saved as the standard position. After the teaching is completed, the main interface will display the batch information panel for each workstation.

[0151] (3) Before production, the operator can set the planned material feeding quantity for each workstation on the main interface. After clicking the start button, the system will enter the fully automatic operation mode. Its high-precision production process is as follows:

[0152] a. Material picking and initial screening: The system picks up the material from the vibratory feeder and transports it to the top station. The camera at this station captures a top image of the chip and quickly determines the front and back of the chip based on the top image. Only chips with the front facing up are selected and initially positioned to provide initial position data for subsequent precise operations.

[0153] b. Frontal Precision Positioning: Qualified chips with their front faces facing up are sent to the frontal vision station, where high-resolution frontal images are acquired. Through algorithms such as multi-fitting, the precise position and angle of the chip are calculated and identified. The position and angle deviations are then sent to the motion control system for real-time compensation.

[0154] c. Reverse side verification and inspection: The chip is then flipped or moved to the reverse side vision station. This station performs reverse side imaging of the chip, performs pin inspection, damage analysis, etc., and performs secondary fine positioning verification of the chip position based on the reverse side features to ensure the accuracy of the front side positioning.

[0155] d. Final placement: The chip is transported above the test socket. The socket vision station first takes a picture of the target socket to confirm that there are no chip residues inside and accurately locates the current position of the socket. The motion control module combines the final position information of the chip and the precise position of the socket to drive the suction nozzle to place the chip into the socket with high precision (±0.1mm), and puts the defective chip into the waste bin.

[0156] Throughout the process, the administrator can manually and precisely control the motion axis or air path of any actuator through the equipment operation interface for debugging or fault recovery. Vision parameter engineers can independently adjust and test the camera, light source, and detection algorithm parameters of each station through the vision parameter setting interface. The accuracy of these parameters is the cornerstone of maintaining the overall positioning accuracy of ±0.1mm of the system.

[0157] This invention achieves high-precision, closed-loop guidance throughout the entire process, from chip initial screening, precise positioning, reverse verification to socket alignment, through a multi-station, multi-sensor, and multi-algorithm fusion visual positioning system. This invention not only achieves sub-pixel-level positioning and multi-dimensional detection within each station, but also constructs a high-precision loading and unloading system with self-calibration, self-adaptation, and self-diagnosis capabilities through cross-validation of data between stations and dynamic error compensation. This overcomes the bottlenecks of insufficient positioning accuracy, poor reliability, and cumbersome changeover processes in existing technologies, demonstrating significant technological advancement and industrial application value.

[0158] The present invention also provides a computer program product, comprising a computer program that, when executed by a processor, implements the steps of the multi-station vision-guided automatic loading and unloading method formed by any or a combination of the above examples. The processor may be a single-core or multi-core central processing unit or a specific integrated circuit, or one or more integrated circuits configured to implement the present invention.

[0159] The present invention also provides a storage medium having the same inventive concept as a multi-station vision-guided automatic loading and unloading method formed by any or more of the above examples, wherein computer instructions are stored thereon, and the computer instructions, when executed, perform the steps of the multi-station vision-guided automatic loading and unloading method formed by any or more of the above examples.

[0160] Based on this understanding, the technical solution of this embodiment, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0161] This invention also provides a terminal having the same inventive concept as any example or combination of examples corresponding to the above-described multi-station vision-guided automatic loading and unloading method, including a memory and a processor. The memory stores computer instructions executable on the processor, and the processor executes the steps of the above-described multi-station vision-guided automatic loading and unloading method when executing the computer instructions. The processor may be a single-core or multi-core central processing unit or a specific integrated circuit, or one or more integrated circuits configured to implement this invention.

[0162] In one example, the terminal, i.e., the electronic device, is represented in the form of a general-purpose computing device. The components of the electronic device may include, but are not limited to: at least one processing unit (processor) mentioned above, at least one storage unit mentioned above, and a bus connecting different system components (including storage units and processing units).

[0163] The storage unit stores program code that can be executed by the processing unit, causing the processing unit to perform the steps described in the "Exemplary Methods" section above, based on various exemplary embodiments of the present invention. For example, the processing unit can execute the aforementioned multi-station vision-guided automatic loading and unloading method.

[0164] The storage unit may include a readable medium in the form of a volatile storage unit, such as a random access memory (RAM) 3201 and / or a cache storage unit, and may further include a read-only memory (ROM).

[0165] The storage unit may also include a program / utility having a set (at least one) of program modules, including but not limited to: an operating system, one or more application programs, other program modules, and program data, each or some combination of these examples may include an implementation of a network environment.

[0166] A bus can represent one or more of several types of bus structures, including a memory cell bus or memory cell controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus that uses any of the various bus structures.

[0167] The electronic device can also communicate with one or more external devices (e.g., keyboards, pointing devices, Bluetooth devices, etc.), one or more devices that enable a user to interact with the electronic device, and / or any device that enables the electronic device to communicate with one or more other computing devices (e.g., routers, modems, etc.). This communication can be achieved through input / output (I / O) interfaces. Furthermore, the electronic device can communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter. The network adapter communicates with other modules of the electronic device via a bus. It should be understood that other hardware and / or software modules can be used in conjunction with the electronic device, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0168] Through the above description, those skilled in the art will readily understand that the exemplary embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solution according to this exemplary embodiment can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.) or on a network, including several instructions to cause a computing device (such as a personal computer, server, terminal device, or network device, etc.) to execute the method of the exemplary embodiment of this application.

[0169] The above detailed embodiments are a description of the present invention. It should not be considered that the specific embodiments of the present invention are limited to these descriptions. For those skilled in the art, several simple deductions and substitutions can be made without departing from the concept of the present invention, and all of these should be considered to fall within the protection scope of the present invention.

Claims

1. A multi-station vision-guided automatic loading and unloading method, characterized in that, Includes the following steps: The chip is transferred to the top station, a top image of the chip is acquired, the front and back sides are determined and screened based on the top image, and the front-facing chips are initially positioned. The control actuator picks up the chip based on the preliminary positioning results and moves the chip to the front workstation. It then acquires a front image of the chip, calculates the chip's first current position based on the front image, and calculates the first position deviation between the first current position and the standard position. And / or, calculate the first current angle of the chip based on the front image, and calculate the first angle deviation between the first current angle and the standard angle; The chip is transferred to the reverse side station, the reverse side image of the chip is acquired, the second current position and / or the second current angle of the chip are independently calculated based on the reverse side image, and the first current position and the second current position are cross-validated, and / or the first current angle and the second current angle are cross-validated to generate a verification result; Acquire test socket images, detect residual chips on the test socket based on the test socket images, and calculate the current position of the test socket; Based on the first deviation result, the verification result, and the current position of the test socket, the control actuator compensates and corrects the placement position and / or angle of the chip, and places the chip into the test socket; the first deviation result includes the first position deviation and / or the first angle deviation.

2. The multi-station vision-guided automatic loading and unloading method according to claim 1, characterized in that, The step of determining and filtering the front and back sides based on the top image, and initially locating the filtered front-facing chips, includes: The top image is processed to identify the first chip region, extract the outline of the first chip, and calculate the center coordinates of the first pixel of the chip. Extract character or texture features from the chip surface and match them with a preset front character texture template to determine and filter the front and back of the chip; Based on the camera calibration model, the center coordinates of the first pixel of the front-facing chip are converted into preliminary world coordinates in the mechanical coordinate system, thus completing the initial positioning.

3. The multi-station vision-guided automatic loading and unloading method according to claim 1, characterized in that, The calculation of the chip's first current position based on the frontal image includes: The front image is processed to identify the second chip region, extract the outline of the second chip, and calculate the center coordinates of the second pixel of the chip as the first current position; And / or, calculate the first current angle of the chip based on the frontal image, including: Identify a first preset feature mark on the frontal image, and calculate the angle of the first preset feature mark in the frontal image as the first current angle; And / or, independently calculate the second current position of the chip based on the reverse image, including: The reverse image is processed to identify the third chip region, extract the outline of the third chip, and calculate the center coordinates of the third pixel of the chip as the second current position; And / or, independently calculate the chip's second current angle based on the reverse image, including: Identify a second preset feature mark on the reverse image, and calculate the angle of the second preset feature mark in the image as the second current angle.

4. The multi-station vision-guided automatic loading and unloading method according to claim 1, characterized in that, The step of cross-validating the first current position with the second current position, and / or cross-validating the first current angle with the second current angle to generate a verification result includes: Calculate the second position deviation between the first current position and the second current position, and / or calculate the second angle deviation between the first current angle and the second current angle; The second position deviation and / or the second angle deviation are compared with preset position thresholds and angle thresholds, respectively. If the second position deviation does not exceed the position threshold and / or the second angle deviation does not exceed the angle threshold, a verification result that passes the verification is generated; otherwise, a verification result that fails the verification is generated.

5. The multi-station vision-guided automatic loading and unloading method according to claim 1, characterized in that, After acquiring a frontal image of the chip, the following is also included: Batch verification and identity recognition are performed based on the frontal image; And / or, perform surface defect detection based on the frontal image.

6. The multi-station vision-guided automatic loading and unloading method according to claim 1, characterized in that, After acquiring an image of the reverse side of the chip, the process also includes: Batch verification and identity recognition based on the reverse image; and / or, Pin inspection is performed based on the reverse image, including at least one of integrity inspection, damage and defect inspection.

7. The multi-station vision-guided automatic loading and unloading method according to claim 1, characterized in that, The method also includes a teaching step: The initial material placement coordinates obtained from visual positioning are adjusted based on human visual feedback to generate fine-tuned coordinates; The actuator, controlled by the fine-tuned coordinates, carries the reference chip and descends to the preset height of the test stand. The actuator is controlled to descend at a speed below the speed threshold, while real-time pressure data associated with the test seat is collected simultaneously. If the real-time pressure data is within the preset safe pressure threshold range, the fine-tuned coordinates will be used as the initial photo position of the test station. If the real-time pressure data is not within the preset safe pressure threshold range, the movement will be stopped and an alarm will be triggered.

8. A multi-station vision-guided automatic loading and unloading control system, used to execute the method according to any one of claims 1-7, characterized in that, The system includes: The visual inspection module includes a top station unit, a front station unit, a back station unit, and a test socket station unit. The top station unit is used to acquire a top image of the chip, determine and filter the front and back sides based on the top image, and perform preliminary positioning on the filtered front-facing chips. The front station unit is used to acquire a front image of the chip, calculate a first positional deviation between the chip's first current position and a standard position based on the front image, and / or calculate a first angular deviation between the chip's first current angle and a standard angle based on the front image. The back station unit is used to acquire a back image of the chip, independently calculate a second current position and / or a second current angle based on the back image, cross-validate the first current position and the second current position, and / or cross-validate the first current angle and the second current angle to generate a verification result. The test socket station unit is used to acquire a test socket image, detect residual chips on the test socket based on the test socket image, and calculate the current position of the test socket. The motion control module, connected to the vision inspection module, is used to control the actuator to pick up the chip according to the preliminary positioning result and move the chip to the front station and the back station. The motion control module is also used to control the actuator to compensate and correct the placement position and / or angle of the chip according to the first deviation result, the verification result and the current position of the test socket, and to put the chip into the test socket. The first deviation result includes the first position deviation and / or the first angle deviation.

9. The multi-station vision-guided automatic loading and unloading control system according to claim 8, characterized in that, The system also includes: The pressure sensing module is located below the test stand and is used to collect real-time pressure data of the test stand. The teaching subsystem, connected to the motion control module and pressure sensing module, adjusts the initial feeding coordinates, generates fine-tuned coordinates, and feeds them back to the motion control module. The motion control module then controls the actuator to move the reference chip according to these fine-tuned coordinates, lowering it to the preset height of the test stand. Simultaneously, the motion control module controls the actuator to descend at a speed below a threshold. The teaching subsystem simultaneously collects real-time pressure data associated with the test stand. If the real-time pressure data is within the preset safe pressure threshold range, the fine-tuned coordinates are used as the initial image capture position for the test stand. If the real-time pressure data is outside the preset safe pressure threshold range, the teaching subsystem feeds back the abnormal result to the motion control module, causing the motion control module to stop the actuator's movement and trigger an alarm.

10. The multi-station vision-guided automatic loading and unloading control system according to claim 8, characterized in that, The system also includes: The host computer is connected to the motion control module and the vision inspection module. It is used to send task instructions to the motion control module and the vision inspection module, and to receive the result data returned by the motion control module and the vision inspection module, and to perform task management and scheduling. The host computer also stores production parameter files, vision parameter files and hardware board files for different workstations. The multi-station parallel processing module is connected to the host computer module and is used to load and independently configure the corresponding production parameter files, vision parameter files and hardware board files from the host computer for the top station, front station and back station respectively. The hardware board file stores the standard position and standard angle.

Images