A Method and System for Controlling the Processing Accuracy of MicroLED Pixel Units Based on Online Feedback

Abstract

This invention relates to the field of precision control technology, and more specifically, to a method and system for precision control of Micro LED pixel unit processing based on online feedback. The method involves placing a Micro LED chip and a micro-driver IC at designated positions on a pad, and performing a first-stage curing process on the Micro LED chip and the micro-driver IC to obtain the first-stage processed pixel unit. An optical inspection station identifies the position of the Micro LED chip and the micro-driver IC relative to the pad, obtaining position offset data for the pixel unit. The position offset data is compared according to a preset standard, and based on the comparison result, a corresponding work instruction is generated from three levels of instructions: qualified, fine-tuning, and rework, to execute the corresponding processing work. A second-stage curing process and subsequent processing steps are then performed on the first-stage processed pixel unit to prepare the finished pixel unit.

Description

Technical Field

[0001] This invention relates to the field of precision control technology, and more specifically, to a method and system for controlling the processing precision of Micro LED pixel units based on online feedback. Background Technology

[0002] Micro LED chips and micro driver ICs require precise electrical connections to the pads on the pads to ensure stable signal transmission and normal operation of the pixel units. Positional misalignment can lead to poor electrical connections, such as increased contact resistance or open circuits, affecting the performance and reliability of the pixel units and even causing the entire display to malfunction. Traditional processing methods struggle to meet the high-precision requirements when handling micro-sized Micro LED pixel units. Traditional methods often lack real-time detection and feedback mechanisms, making it impossible to promptly detect and correct precision issues during processing. Final inspection is usually performed after processing is complete. Once a problem is discovered, extensive rework or scrapping is required, increasing production costs and time. Summary of the Invention

[0003] In view of this, embodiments of the present invention provide a method and system for adjusting the processing accuracy of Micro LED pixel units based on online feedback, so as to realize the automatic real-time adjustment of the chip layout of pixel units.

[0004] To achieve the above objectives, the embodiments of the present invention provide the following technical solutions: According to one aspect of the present invention, a method for controlling the processing accuracy of Micro LED pixel units based on online feedback is provided, comprising: The Micro LED chip and the micro driver IC are placed at designated positions on the pads, and the Micro LED chip and the micro driver IC are subjected to the first stage of curing to obtain the first stage of processed pixel units. The positional offset data of the pixel unit is obtained by identifying the position of the Micro LED chip and the micro driver IC relative to the pad through the optical inspection station. The position offset data is compared according to a preset standard, and a corresponding work instruction is generated from three levels of instructions: qualified, fine-tuning, and rework, based on the comparison result, to perform the corresponding processing work. The pixel unit processed in the first stage is subjected to a second stage of curing and subsequent processing steps to prepare the finished pixel unit.

[0005] According to another aspect of the present invention, a Micro LED pixel unit processing precision control system based on online feedback is provided, comprising: The preliminary processing module is used to place the Micro LED chip and the micro driver IC at designated positions on the pads, and to perform the first stage of curing processing on the Micro LED chip and the micro driver IC to obtain the first stage of processed pixel units; The offset recognition module is used to identify the position of the Micro LED chip and the micro driver IC relative to the pad of the pixel unit through the optical detection station, and to obtain the position offset data of the pixel unit. The instruction processing module is used to compare the position offset data according to a preset standard, and select from the three levels of instructions (qualified, fine-tuning, and rework) based on the comparison result to generate the corresponding work instruction to perform the corresponding processing work. The subsequent processing module is used to perform a second-stage curing process and subsequent processing steps on the pixel units processed in the first stage to prepare the finished pixel units.

[0006] As can be seen from the above technical solutions, the Micro LED pixel unit processing accuracy control method based on online feedback provided by the present invention has the following beneficial effects: This invention acquires position offset data through optical detection, enabling real-time monitoring of processing accuracy. Based on preset standards, it generates three levels of instructions, allowing for targeted handling of different accuracy conditions. Qualified products can directly proceed to subsequent processes, fine-tuning can salvage products with minor deviations, and rework can disassemble and reuse severely deviated units. This not only ensures product quality, improves display performance and electrical connection reliability, but also increases production efficiency, reduces scrap rate and costs, and adapts to the needs of large-scale production. Attached Figure Description

[0007] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are merely embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort: Figure 1 A schematic diagram illustrating the steps of the Micro LED pixel unit processing accuracy control method based on online feedback provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of a Micro LED pixel unit processing accuracy control system based on online feedback, provided in an embodiment of the present invention. Detailed Implementation

[0008] The technical solutions of the embodiments 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.

[0009] Micro LED chips and micro driver ICs require precise electrical connections to the pads on the pads to ensure stable signal transmission and normal operation of the pixel units. Positional misalignment can lead to poor electrical connections, such as increased contact resistance or open circuits, affecting the performance and reliability of the pixel units and even causing the entire display to malfunction. Traditional processing methods struggle to meet the high-precision requirements when handling micro-sized Micro LED pixel units. Traditional methods often lack real-time detection and feedback mechanisms, making it impossible to promptly detect and correct precision issues during processing. Final inspection is usually performed after processing is complete. Once a problem is discovered, extensive rework or scrapping is required, increasing production costs and time.

[0010] In view of this, the present invention provides a method for controlling the processing accuracy of Micro LED pixel units based on online feedback, the steps of which are as follows: Figure 1 As shown, it includes: The first step is to place the Micro LED chip and the micro driver IC on the designated positions on the pads, and then perform the first stage of curing on the Micro LED chip and the micro driver IC to obtain the first stage of processed pixel units.

[0011] Specifically, in the first step of the embodiment provided by the present invention, a high-precision die bonder or other automated equipment is used. The specified position coordinates on the pad are preset by the program. The MicroLED chip and the micro driver IC are accurately picked up by the robotic arm or suction cup and placed accurately at the corresponding position on the pad.

[0012] More specifically, during the placement process, the equipment is often equipped with a high-precision vision positioning system. This system acquires real-time images of the pads and chips, and through image recognition technology, accurately determines the target position on the pads as well as the actual position of the chip and driver IC, thereby guiding the mechanical components to be placed precisely.

[0013] More specifically, Micro LED chips and micro driver ICs need to be accurately electrically connected to circuit structures such as pads on the pads. Only by placing them in the designated positions can the pins of the chip and driver IC be correctly aligned with the pads to achieve good electrical conduction and ensure the normal operation of the entire pixel unit. Micro LED displays have extremely high requirements for pixel arrangement accuracy. Precisely placing the chip and driver IC in the designated positions helps to ensure the positional accuracy and consistency of each pixel unit, thereby achieving high-quality, high-resolution display effects.

[0014] More specifically, a common method of heat curing is to place the pads containing the chip and driver IC into a heating device, such as an oven or reflow oven. Depending on the characteristics of the curing material used (such as glue, solder, etc.), a suitable temperature profile and time parameters are set to allow the material to gradually cure during the heating process, thus fixing the chip and driver IC onto the pads. In some cases where photocurable materials are used, a light irradiation device with a specific wavelength and intensity is used to irradiate the pads, causing the material to undergo a photochemical reaction and cure.

[0015] More specifically, the main purpose of the first curing process is to initially fix the Micro LED chip and the micro driver IC. During subsequent processing, transportation and testing, the chip and driver IC will be affected by external forces such as vibration and collision. Through curing, they can maintain a stable position on the pads, reducing processing errors and electrical connection problems caused by position movement. The pixel unit after initial curing has a certain degree of stability, which facilitates subsequent testing, adjustment and further processing steps. For example, it can more easily perform optical testing and position fine-tuning, while also reducing the risk of damage to the chip and driver IC during subsequent processing.

[0016] The second step involves using an optical inspection station to identify the position of the Micro LED chip and the micro driver IC relative to the pads in the pixel unit, thereby obtaining the position offset data of the pixel unit.

[0017] Specifically, in the second step of the embodiment provided by the present invention, after the pixel unit is transported to the optical inspection station, the station will perform image acquisition on the pixel unit at a specified resolution to obtain an optical inspection image. The optical inspection station is usually equipped with multiple cameras or optical sensors at different angles to acquire images of the pixel unit from multiple perspectives. This can obtain more comprehensive pixel unit information and avoid some features not being accurately identified due to occlusion or blind spots in a single perspective. According to the characteristics of the pixel unit and the inspection requirements, the parameters of the image acquisition device, such as exposure time, focal length, aperture, etc., are adjusted to ensure that the acquired image is clear and has appropriate contrast, and can accurately reflect the characteristics of the Micro LED chip, micro driver IC and pads.

[0018] More specifically, Micro LED chips and micro driver ICs need to be precisely electrically connected to the pads to ensure the normal operation of the pixel units. By obtaining position offset data through position identification, the positional deviation of the chip and driver IC relative to the pads can be detected in a timely manner. If the offset exceeds the allowable range, it will lead to poor electrical connection, such as increased contact resistance and unstable signal transmission, thereby affecting the performance and reliability of the pixel units. Therefore, accurately identifying position offset and making adjustments is the key to ensuring the quality of electrical connection.

[0019] More specifically, the acquired optical detection images are used to identify the features of the Micro LED chip, the micro driver IC, and the pads on the pads to obtain their positioning and identification data. The acquired raw images are preprocessed, including noise reduction, filtering, and contrast enhancement, to improve image quality and facilitate subsequent feature recognition.

[0020] More specifically, image processing and computer vision algorithms, such as edge detection, template matching, and feature point detection, are used to extract feature information of Micro LED chips, micro driver ICs, and pads from preprocessed images, such as contours, corners, and center positions. Through trained machine learning or deep learning models, the features in the image are classified and recognized to determine the specific positions and orientations of Micro LED chips, micro driver ICs, and pads.

[0021] More specifically, in Micro LED displays, the positional accuracy of each pixel unit has a significant impact on display quality. Positional offset can lead to problems such as uneven spacing between pixels and inconsistent color mixing, thereby affecting the resolution, contrast, and color accuracy of the display. By detecting and controlling positional offset, the positional accuracy and consistency of each pixel unit can be ensured, thus improving the overall display quality.

[0022] More specifically, based on the positioning and identification data, the positional offset of the Micro LED chip and the micro driver IC relative to the pads is analyzed to obtain the positional offset data of the pixel unit. The positioning and identification data of the Micro LED chip, the micro driver IC and the pads are unified into the same coordinate system to enable accurate position comparison. By comparing the actual position of the Micro LED chip and the micro driver IC with the ideal position of the pads, their positional offsets in various directions are calculated, including translational offsets and rotational offsets in the X and Y directions. The calculated positional offsets are organized into standardized positional offset data, including information such as the direction, magnitude and angle of the offset, for subsequent analysis and processing.

[0023] More specifically, position offset data provides an important basis for subsequent adjustments and optimizations. If the position offset is detected to exceed the acceptable range, corresponding adjustment measures can be taken based on the offset data, such as driving the die bonder or fine-tuning mechanism to correct the position of the chip and driver IC. At the same time, by analyzing a large amount of position offset data, problems and trends in the processing can be discovered, thereby optimizing the processing technology and improving processing accuracy and production efficiency.

[0024] The third step is to compare the position offset data according to the preset standard, and select the corresponding work instruction from the three levels of instructions (qualified, fine-tuning, and rework) based on the comparison result to perform the corresponding processing work.

[0025] Specifically, in the third step of the embodiment provided by this invention, the design drawings of the corresponding pixel units are obtained, the ideal positional relationship between the Micro LED chip and the micro driver IC relative to the pads is clarified, and the offset situations that occur under different conditions are analyzed by referring to previous calibration records. For various possible offset situations, corresponding visual recognition logic and theoretical calibration strategies are generated. For example, when a translational offset occurs in a certain direction, it is determined how to accurately identify the offset through visual detection and formulate corresponding adjustment strategies. The design drawings and calibration records are preliminarily analyzed by an automated data summarization algorithm to extract key information and patterns. At the same time, manual review and supplementation are carried out to ensure the accuracy and comprehensiveness of the preset standards.

[0026] More specifically, the obtained pixel unit position offset data is matched one by one with various offset conditions in the preset standard. The direction, magnitude, angle and other parameters of the offset are checked to see if they match a certain condition in the preset standard. For cases where there is no complete match, the similarity between the position offset data and various conditions in the preset standard is calculated. An appropriate similarity calculation method is used, such as Euclidean distance or cosine similarity, and the closest preset condition is determined based on the degree of similarity.

[0027] More specifically, if the position offset data matches the acceptable range in the preset standard, that is, the offset is within the allowable error range, a qualified instruction is generated, indicating that the position accuracy of the pixel unit meets the requirements and can directly proceed to the subsequent processing steps; when the position offset data exceeds the acceptable range, but can be adjusted to meet the acceptable standard, a fine-tuning instruction is generated. The fine-tuning instruction contains the theoretical adjustment strategies required for the fine-tuning operation, such as the adjustment direction, distance, angle, and other parameters; if the position offset data seriously exceeds the preset standard and cannot be fine-tuned to meet the acceptable requirements, or if the cost or difficulty of fine-tuning is too high, a rework instruction is generated, meaning that the pixel unit needs to enter the scrap process for dismantling and reuse.

[0028] More specifically, when the work instruction is a qualified instruction, it directly jumps to the second stage of curing and subsequent processing steps for the pixel unit processed in the first stage, and continues to complete the entire pixel unit preparation process. According to the fine-tuning instruction, the die bonding head or fine-tuning mechanism is driven to physically correct the position of the Micro LED chip and the micro driver IC. While performing the first calibration stage, the calibration effect is monitored through the optical inspection station. Based on the feedback results obtained from the monitoring, the calibration method to be performed in the subsequent calibration stage is determined. When the work instruction is a rework instruction, the pixel unit processed in the first stage is arranged to enter the waste process, the pixel unit is disassembled, and the reusable components are recycled and processed for reuse in production.

[0029] More specifically, by comparing position offset data with preset standards and generating corresponding work instructions based on the comparison results, it can be ensured that only pixel units that meet quality requirements enter the subsequent processing and production stages. Unqualified products can be adjusted or processed in a timely manner to prevent defective products from entering the market, thereby improving the overall quality and reliability of the products.

[0030] More specifically, by adopting a three-level instruction approach, different processing measures can be taken according to the degree of positional offset. For pixel units with small offsets, fine-tuning can bring them up to the qualified standard, avoiding unnecessary rework and saving time and costs. For seriously unqualified products, rework can be carried out in a timely manner to avoid wasting more resources in subsequent processing and improve the efficiency of the entire production process.

[0031] More specifically, for the pixel units involved in the rework instructions, disassembly and reuse can minimize the waste of raw materials and reduce production costs. At the same time, by analyzing and summarizing the data in the production process, the preset standards and processing technology can be continuously optimized to further improve the efficiency of resource utilization.

[0032] More specifically, the entire comparison and instruction generation process can be realized through computer systems and automated equipment, reducing manual intervention and improving the automation level of the production process. At the same time, through the analysis and learning of a large amount of data, preset standards and processing strategies can be continuously optimized to achieve intelligent upgrading of the production process.

[0033] The fourth step involves performing a second-stage curing process and subsequent processing steps on the pixel units processed in the first stage to prepare the finished pixel units.

[0034] Specifically, in the fourth step of the embodiment provided by the present invention, if the first stage of curing is carried out by heating, the second stage of curing can further adjust the temperature curve based on the first stage. For example, the first stage allows the material to be initially cured at a lower temperature, while the second stage raises the temperature to a suitable range and maintains it for a certain time to promote the complete curing of the material. High-precision heating equipment, such as an oven or reflow oven with precise temperature control function, is used, and the operation is carried out strictly in accordance with the preset temperature-time curve.

[0035] More specifically, for cases using photocurable materials, the second curing stage involves adjusting the intensity, duration, and wavelength of the light. For example, the first stage uses weaker light to partially cure the material, while the second stage increases the light intensity and duration to ensure complete curing. Professional lighting equipment is used to precisely control the lighting parameters.

[0036] More specifically, the first curing stage only initially fixes the Micro LED chip and the micro driver IC. The second curing stage enables the curing material to more tightly bond the chip and driver IC to the pads, improving the stability and reliability of the connection between them and preventing displacement due to factors such as vibration and temperature changes during subsequent use. After the second curing stage, the physical and chemical properties of the curing material will be further optimized. For example, the material's hardness, wear resistance, and corrosion resistance will be improved, thereby improving the overall performance and lifespan of the pixel unit.

[0037] More specifically, professional electrical testing equipment, such as multimeters and oscilloscopes, is used to test various electrical parameters of the pixel unit. The tests include voltage, current, resistance, signal transmission, etc. The probes of the testing equipment are accurately connected to the corresponding test points of the pixel unit, and the test is performed according to the preset test procedure. The test data is recorded. Through electrical performance testing, problems in the electrical connection and signal transmission of the pixel unit, such as short circuits, open circuits, and signal distortion, can be detected in time. If these problems are not detected and repaired in time, the pixel unit will not work properly, affecting the display effect of the entire screen.

[0038] More specifically, suitable encapsulation materials, such as epoxy resin and silicone, are used to encapsulate the pixel unit. The encapsulation process can use automated encapsulation equipment to precisely coat or inject the encapsulation material onto the surface of the pixel unit, forming a protective coating. After encapsulation, a curing process is required to completely harden the encapsulation material. Encapsulation can protect the pixel unit from the influence of the external environment, such as dust, moisture, and oxygen. These external factors can negatively affect the performance and lifespan of the pixel unit. Encapsulation can improve the stability and reliability of the pixel unit and extend its service life.

[0039] More specifically, a combination of manual inspection and machine vision inspection can be used. Manual inspection mainly involves visually inspecting the appearance of pixel units for obvious defects, such as scratches, cracks, and foreign objects. Machine vision inspection uses high-precision cameras and image processing algorithms to conduct a comprehensive and detailed inspection of the appearance of pixel units. It can detect some minute defects that are difficult to detect with the naked eye. Appearance defects affect the optical performance and overall aesthetics of pixel units. For example, scratches can cause light scattering, affecting the display effect; cracks can allow external substances to enter the pixel unit and damage the internal structure. Through appearance inspection, pixel units with appearance defects can be screened out, thereby improving the product yield.

[0040] More specifically, the pixel unit is placed in a specific aging test environment, such as high temperature and high humidity, and continuously powered on for a period of time. During the aging test, the performance changes of the pixel unit are monitored in real time and relevant data is recorded. The aging test can simulate the long-term operation of the pixel unit in actual use and expose some potential quality problems in advance. For example, some pixel units that perform normally in the early stages of normal use will experience performance degradation or failure after a period of operation. Through aging tests, these potential defective products can be screened out, thereby improving the reliability and stability of the product.

[0041] As can be seen from the above technical solutions, the Micro LED pixel unit processing accuracy control method based on online feedback provided by the present invention has the following beneficial effects: This invention acquires position offset data through optical detection, enabling real-time monitoring of processing accuracy. Based on preset standards, it generates three levels of instructions, allowing for targeted handling of different accuracy conditions. Qualified products can directly proceed to subsequent processes, fine-tuning can salvage products with minor deviations, and rework can disassemble and reuse severely deviated units. This not only ensures product quality, improves display performance and electrical connection reliability, but also increases production efficiency, reduces scrap rate and costs, and adapts to the needs of large-scale production.

[0042] Furthermore, the step of identifying the position of the Micro LED chip and the micro driver IC relative to the pads in the pixel unit through an optical inspection station to obtain the position offset data of the pixel unit includes: S21: The pixel unit is transported to the optical inspection station to acquire an image of the pixel unit at a specified resolution through the optical inspection station, so as to obtain an optical inspection image; S22: Perform feature recognition on the optical detection image of the Micro LED chip, micro driver IC and pads to obtain positioning recognition data of the Micro LED chip, micro driver IC and pads; S23: Based on the positioning and identification data, perform position offset analysis of the Micro LED chip and the micro driver IC relative to the pads to obtain the position offset data of the pixel unit.

[0043] Specifically, automated conveying equipment, such as conveyor belts and robotic arms, is used to accurately transport the pixel units that have undergone the first stage of curing to the optical inspection station. During the transport process, it is necessary to ensure the stability of the pixel units to avoid affecting the quality of subsequent image acquisition due to vibration or shaking. At the optical inspection station, the parameters of the optical inspection equipment are adjusted according to the preset specified resolution requirements. For example, the focal length, aperture, and exposure time of the camera are set to ensure that the acquired images are clear and accurate, and can clearly present the characteristics of the Micro LED chip, the micro driver IC, and the pads. In order to obtain more comprehensive information, the optical inspection station can acquire images of the pixel units from multiple different perspectives. Multiple cameras or rotatable camera devices are used to capture images of the pixel units from different angles to obtain multiple sets of optical inspection images.

[0044] More specifically, images are the foundation for subsequent feature recognition and positional offset analysis. By acquiring images at specified resolutions, sufficiently clear and detailed image information can be obtained, providing the necessary conditions for accurately identifying the features of Micro LED chips, micro driver ICs, and pads. Multi-view acquisition can avoid the problem of some features not being accurately identified due to occlusion or blind spots in a single viewpoint, thereby obtaining more comprehensive pixel unit information and improving the accuracy of subsequent analysis.

[0045] More specifically, preprocessing operations are performed on the acquired optical inspection images, including noise reduction, filtering, and contrast enhancement. For example, Gaussian filtering is used to remove noise from the image, and histogram equalization is used to enhance the image contrast, making the features in the image more obvious. Image processing and computer vision algorithms are then used to extract features from the preprocessed images. Common algorithms include edge detection algorithms (such as Canny edge detection), corner detection algorithms (such as Harris corner detection), and template matching algorithms. Through these algorithms, key features of Micro LED chips, micro driver ICs, and pads, such as edges, corners, and contours, are identified. For some complex feature recognition tasks, trained machine learning or deep learning models can be used. For example, convolutional neural networks (CNNs) can be used to classify and recognize images, determine the specific location and orientation of Micro LED chips, micro driver ICs, and pads, and obtain their localization and recognition data.

[0046] More specifically, feature recognition can accurately determine the position and orientation of Micro LED chips, micro driver ICs, and pads in an image. This positioning and recognition data is the key input for position offset analysis. Only by accurately identifying their positions can the position offset of the chip and driver IC relative to the pads be calculated. In actual production, the appearance and features of pixel units are affected by a variety of factors, such as surface defects and uneven lighting. Machine learning and deep learning models have strong adaptability and robustness, and can accurately identify features under complex conditions, thereby improving the accuracy and reliability of recognition.

[0047] More specifically, the positioning and identification data of the Micro LED chip, micro driver IC, and pads are unified into a single coordinate system. Typically, a fixed coordinate system is established with the pads as a reference. The position data of the chip and driver IC are transformed into this coordinate system for accurate position comparison. Based on the positioning and identification data in the unified coordinate system, the positional offset of the Micro LED chip and micro driver IC relative to the pads is calculated. This calculation includes translational offsets in the X and Y directions and rotational offsets around specific axes. For example, the translational offset is obtained by calculating the coordinate difference between the center point of the chip and driver IC and the corresponding reference point on the pad; the rotational offset is obtained by analyzing the difference between the chip and driver IC's attitude angle and the ideal attitude angle. The calculated positional offset data is then organized and output in a standardized format. This offset data can be stored in a database for subsequent analysis and processing, and can also be fed back to the control system to provide a basis for subsequent fine-tuning operations.

[0048] More specifically, the purpose of position offset analysis is to quantify the positional deviation of the Micro LED chip and the micro driver IC relative to the pads. By accurately calculating the offset, we can intuitively understand the positional accuracy of the chip and the driver IC and determine whether it meets production requirements. Position offset data is an important basis for subsequent fine-tuning or rework decisions. If the offset exceeds the allowable error range, the position of the chip and the driver IC needs to be adjusted to ensure the quality and performance of the pixel unit. Therefore, accurate position offset analysis is crucial for ensuring product quality and production efficiency.

[0049] Furthermore, the optical inspection station performs multi-view image acquisition on the pixel unit to obtain an optical inspection image, and determines whether the pixel unit needs to be supplemented for inspection based on the position offset data of the optical inspection image. If supplementary inspection is required, the viewing angle of the second round of inspection is obtained by analyzing the position offset data, so as to determine the optical inspection work to adjust the viewing angle of image acquisition according to the viewing angle to obtain the second round of optical inspection image.

[0050] Specifically, multiple cameras or optical sensors are arranged around the optical inspection station to capture images of the pixel unit from different angles. For example, cameras are set up above, to the sides, and other directions of the pixel unit to ensure that all surfaces and angles of the pixel unit are covered. The control system controls multiple cameras synchronously to ensure that images of the pixel unit are captured from different perspectives at the same time. During the acquisition process, the system operates according to preset parameters such as resolution and exposure time to obtain clear and accurate optical inspection images.

[0051] More specifically, the acquired optical inspection images are processed and analyzed to obtain the position offset data of the pixel units. This position offset data is compared with the preset pass standards to evaluate whether the position accuracy of the pixel units meets the requirements. If the position offset data shows that the position deviation of some parts of the pixel unit is close to or exceeds the pass range, or if the position offset situation is more complicated and difficult to judge accurately with the current image data, then it is determined that the pixel unit needs to be supplemented for inspection.

[0052] More specifically, the positional offset data is analyzed in depth to identify the specific locations and characteristics of the pixel unit positional offset. For example, it is analyzed whether a large offset has occurred in a corner of a chip or whether the entire driver IC has rotated. Based on the results of the offset feature analysis, the viewing angle for the second round of detection is determined. If the offset of a certain part is not clear from the current viewing angle and needs to be observed from other angles, then the viewing angle that can better observe that part is selected as the viewing angle for the second round of detection. For example, if it is found that the side offset of the chip is difficult to judge, then a specific angle on the side is selected for the second round of detection.

[0053] More specifically, the optical inspection station adjusts the angle of the camera or optical sensor according to the determined viewing angle of the second round of inspection through mechanical structure or optical system so that it is aligned with the specified viewing angle. After adjusting the viewing angle, the pixel unit is image acquired again to obtain the second round of optical inspection images. The acquisition process is also carried out according to the preset parameters to ensure image quality.

[0054] More specifically, single-view image acquisition can result in occlusion and blind spots, making it impossible to accurately obtain positional information for certain parts. Multi-view image acquisition allows for a comprehensive observation of all parts of the pixel unit, reducing detection errors caused by occlusion and blind spots. For some complex positional offset situations, single-view images cannot provide enough information for accurate judgment. Multi-view images and supplementary detection can observe the offset from different angles, more accurately analyze the degree and cause of the offset, and improve the accuracy of detection.

[0055] More specifically, supplementary testing can promptly identify positional misalignment issues missed in the initial testing. Pixel units that do not meet quality requirements can be adjusted or processed in a timely manner, preventing defective products from entering subsequent production stages and thus ensuring the overall quality of the product. Based on the results of supplementary testing, the production process can be optimized. For example, if positional misalignment issues are found to be frequent at a specific viewpoint, the corresponding production stage can be inspected and improved to enhance the stability and reliability of the production process.

[0056] More specifically, by determining whether supplementary detection is needed and the viewing angle of the second round of detection based on the positional offset data, targeted detection can be achieved, avoiding unnecessary comprehensive detection of all pixel units, reducing detection time and resource waste, and improving detection efficiency and resource utilization efficiency.

[0057] Furthermore, the steps for generating corresponding work instructions based on the comparison results to execute the corresponding processing work include: S31: Generate corresponding work instructions based on the comparison results, wherein the work instructions include any one of the following: a qualified instruction, a fine-tuning instruction, and a rework instruction; S32: When the work instruction is a qualified instruction, the process jumps to the second stage of solidification processing and subsequent processing steps for the pixel unit processed in the first stage. S33: When the working instruction is a fine-tuning instruction, the die bonding head or fine-tuning mechanism is driven to physically correct the position of the Micro LED chip and the micro driver IC according to the fine-tuning instruction. S34: When the work instruction is a rework instruction, the pixel unit processed in the first stage is arranged to enter the waste process to disassemble and reuse the pixel unit.

[0058] Specifically, based on the comparison results, corresponding work instructions are generated. These work instructions include any one of the following: qualified instructions, fine-tuning instructions, and rework instructions. In other words, the positioning accuracy of the pixel unit is determined by the comparison results, and the corresponding requirements are judged based on the actual situation to generate the corresponding work instructions to allow it to enter the next stage.

[0059] More specifically, when the system determines that the position offset data is qualified based on the comparison with the preset standard, it automatically sends a signal through the control system. This signal triggers the conveyor on the production line to transport the pixel unit processed in the first stage from the current detection station to the second stage curing station. During the transport process, the pixel unit is ensured to move smoothly to avoid affecting its existing position and state due to vibration or collision. After arriving at the second stage curing station, the pixel unit is cured according to the preset second stage curing process parameters, such as temperature, time, and light intensity. Then, the subsequent processing steps are carried out in sequence.

[0060] More specifically, a qualified instruction means that the positional offset of the Micro LED chip and the micro driver IC of the pixel unit relative to the pad is within the allowable error range, and its positional accuracy can meet the product quality requirements. It can directly enter the second stage of curing and subsequent steps, which can maintain the continuity of the production process, improve production efficiency, avoid unnecessary stoppages and processing, and ensure that the product can be completed on time and efficiently.

[0061] More specifically, when the working instruction is a fine-tuning instruction, the control system first parses the fine-tuning instruction and extracts the theoretical calibration strategies contained therein, such as submicron-level parameter information like the direction, distance, and angle of adjustment. Based on the calibration parameters obtained from the parsing, the control system drives the die bonder or fine-tuning mechanism. The die bonder or fine-tuning mechanism is usually equipped with high-precision motors, sensors, and motion control algorithms, which can achieve precise submicron-level displacement and angle adjustments. For example, by precisely controlling the movement of the die bonder in the X, Y, and Z directions, as well as its rotation around a specific axis, the position of the Micro LED chip and the micro driver IC can be corrected.

[0062] More specifically, while performing the first calibration stage, the calibration effect is monitored in real time through the optical inspection station. The optical inspection station collects images of the pixel units again, analyzes the positional changes of the Micro LED chip and the micro driver IC, and transmits the feedback results to the control system in a timely manner. The control system determines the calibration method to be performed in the subsequent calibration stage based on the feedback results. If the current calibration effect does not meet the expectations, the subsequent calibration parameters will be adjusted according to the actual situation, and corrections will continue until the positional offset meets the qualified standard.

[0063] More specifically, the fine-tuning instruction indicates that the position offset of the pixel unit exceeds the acceptable range, but it can still be adjusted to meet the acceptable standard. By physically correcting the position of the chip and driver IC, products that might otherwise be judged as unqualified can be salvaged, reducing the scrap rate, improving the product yield and the utilization rate of production resources. At the same time, the real-time monitoring and dynamic adjustment mechanism can ensure the accuracy and effectiveness of the fine-tuning operation, ensuring that the corrected pixel unit meets the quality requirements.

[0064] More specifically, when the work instruction is a rework instruction, the system will mark the pixel unit processed in the first stage, indicating that it needs to enter the scrap process. At the same time, it will be isolated from the normal production line by an automated conveyor or robotic arm and transported to a dedicated rework processing area. In the rework processing area, professional operators or automated disassembly equipment will disassemble the pixel unit. The disassembly process requires careful handling to avoid damaging reusable components. For example, appropriate tools will be used to separate the Micro LED chip and micro driver IC from the pads. The disassembled components will be fully inspected to determine whether they can still be used. For reusable components, they will be cleaned, tested, and recalibrated, and then stored in a dedicated inventory area for reuse in production. For components that cannot be reused, they will be scrapped.

[0065] More specifically, the rework instruction targets pixel units whose positional deviations significantly exceed preset standards and cannot be adjusted to meet the qualification requirements, or whose adjustment costs are too high or too difficult. By arranging for these units to enter the waste process for dismantling and reuse, the waste of raw materials can be minimized and production costs reduced. At the same time, the recycling and reuse of reusable parts helps improve resource utilization efficiency, which is in line with the principle of sustainable development. In addition, the management and recording of the rework process can provide data support for the improvement of subsequent production processes.

[0066] Furthermore, the fine-tuning instruction includes the theoretical calibration strategy required for the fine-tuning operation. The theoretical calibration strategy includes several calibration stages. While executing the first calibration stage, the calibration effect is monitored through the optical inspection station. Based on the feedback results obtained from the monitoring, the calibration method to be executed in the subsequent calibration stages is determined.

[0067] Specifically, the several adjustment stages in the theoretical adjustment strategy are divided according to the complexity of the positional offset and the possible adjustment methods. For example, for a simple translational offset, only one or two adjustment stages are needed to complete the correction; while for a complex offset that involves both translation and rotation, multiple adjustment stages are needed to make adjustments step by step. Each adjustment stage has a clear goal and operation parameters, such as the direction, distance, and angle of adjustment.

[0068] More specifically, the design of the calibration phase is based on a large amount of historical data and experimental results. By analyzing similar offset cases in the past, the optimal adjustment sequence and parameter range for different types of offsets are summarized. For example, for an offset of a certain direction and size, after many experiments, it was found that making small-scale translation adjustments first and then making small-scale angle adjustments can more efficiently achieve the qualified standard. Therefore, the corresponding calibration phases will be set in this order in the theoretical calibration strategy.

[0069] More specifically, during the first calibration stage, the control system will drive the die bonder or fine-tuning mechanism to make preliminary adjustments to the positions of the Micro LED chip and the micro driver IC according to the theoretical calibration strategies in the fine-tuning instructions. For example, if the goal of the first calibration stage is to make translational adjustments in the X direction, the die bonder will precisely move the chip and the driver IC a specified distance in the X direction.

[0070] More specifically, the optical inspection station monitors the calibration effect in real time. It re-acquires images of the pixel units, performs feature recognition and position analysis, compares the current position data with the target position in the first calibration stage, and evaluates the adjustment effect. For example, it calculates the deviation between the adjusted position and the target position to determine whether the adjustment has been made in the expected direction.

[0071] More specifically, the optical inspection station transmits the feedback results obtained from the supervision to the control system. The control system will conduct a detailed analysis of the feedback data to determine whether the adjustment in the first calibration stage is effective. If the adjusted position is closer to the qualified range but still does not meet the standard, it means that the current calibration method is basically correct, but further parameter adjustments are needed. If the adjusted position deviates from the expected direction or the adjustment effect is not obvious, the calibration strategy needs to be re-evaluated.

[0072] More specifically, based on the feedback analysis results, the control system will determine the execution method for subsequent calibration phases. If the first calibration phase is successful, fine-tuning will continue as planned in subsequent phases, but the parameter values ​​will be adjusted appropriately, such as reducing the adjustment step size, to more accurately achieve the pass standard. If the first calibration phase is unsuccessful, the control system will change the calibration method, such as adjusting the direction or sequence of adjustments, or using different adjustment methods. For example, if the original plan was to perform translation adjustments first, but the results were found to be unsatisfactory, then angle adjustments will be performed first instead.

[0073] More specifically, by conducting real-time monitoring and feedback adjustments at each calibration stage, deviations during the adjustment process can be corrected in a timely manner, ensuring that the final positional accuracy meets the requirements. This avoids the problems of over-adjustment or under-adjustment caused by a one-time adjustment, improving the accuracy of fine-tuning. Different pixel units have different types and degrees of positional offsets, and even the same type of offset may have different specific situations. This dynamic adjustment method can flexibly adjust the calibration strategy according to the actual feedback results, adapting to various complex offset situations and improving the success rate of fine-tuning. By analyzing the feedback results and making dynamic adjustments, unnecessary repeated adjustments can be avoided, reducing the number of adjustments and time, improving the efficiency of fine-tuning, and thus improving the efficiency of the entire production process.

[0074] Furthermore, qualified instructions, fine-tuning instructions, and rework instructions are recorded to construct processing supervision information for feedback on the first segment of the pixel unit's solidification process, and the processing method of the first segment of the pixel unit's solidification process is optimized based on the processing supervision information.

[0075] Specifically, a dedicated data recording system is set up on the production line and connected to the control system that issues pass, fine-tuning, and rework instructions. When an instruction is generated, the system automatically collects relevant information, including the instruction type (pass, fine-tuning, rework), the corresponding pixel unit number, position offset data, and the time the instruction was generated. The collected instruction information is then organized and stored in a database according to certain formats and rules. A relational database, such as MySQL or PostgreSQL, can be used. Different types of instruction information are stored in different tables to facilitate subsequent queries and analysis. At the same time, a unique record is created for each pixel unit, linking its related instruction information and position offset data.

[0076] More specifically, by recording different types of instructions, we can gain a comprehensive understanding of the quality status of pixel units after the first stage of curing. A qualified instruction indicates that the positional accuracy of the pixel unit meets the requirements, a fine-tuning instruction indicates that there is a certain offset but it can be resolved through adjustment, and a rework instruction indicates that the offset is serious and needs to be reprocessed. This information can reflect the potential problems and trends in the processing. Constructing processing supervision information provides a data foundation for subsequent optimization of the first stage of curing. By analyzing a large amount of instruction information, we can discover the patterns and potential problems in the processing, providing a basis for optimization.

[0077] More specifically, based on processing supervision information, the proportions of qualified, fine-tuning, and rework instructions under different processing methods are calculated. For example, when using certain curing temperature, time, or pressure parameters, the proportion of qualified pixel units to the total number of processed units, as well as the proportion of pixel units requiring fine-tuning or rework, are statistically analyzed. At the same time, the distribution of position offset data is analyzed to assess the degree of influence of processing methods on position accuracy. Based on these statistical and analytical results, the benchmark value parameters of processing methods, such as the pass rate, fine-tuning rate, and rework rate, are calculated. By evaluating the effectiveness of processing methods, the impact of different processing parameters and methods on pixel unit quality can be understood. If the pass rate is low and the rework rate is high under a certain processing method, it indicates that there is a problem with the processing method and it needs to be optimized.

[0078] More specifically, the benchmark value parameters for each time period are evaluated, their changing trends are observed, and the fine-tuning instructions of the processing supervision information within a specified time range are selected as feedback analysis objects. For example, a period with large fluctuations in benchmark value parameters within the most recent month is selected, and the fine-tuning instructions within that period are analyzed in depth. The fine-tuning instructions within the specified time range can reflect the small problems that frequently occur in the processing process. These problems are caused by unreasonable details in the processing method. By analyzing these fine-tuning instructions, potential problems and improvement directions in the processing process can be discovered.

[0079] More specifically, the feedback parsing object is analyzed in the form of fine-tuning, and the specific content of the fine-tuning operation is analyzed, such as the direction, distance, and angle of adjustment. By statistically analyzing a large number of fine-tuning instructions, the trend characteristics of pixel unit displacement can be identified. For example, it is found that most fine-tuning is carried out in a certain direction, or the displacement distance has a certain pattern. Understanding the trend characteristics of pixel unit displacement helps to deeply understand the problems existing in the processing. By analyzing these characteristics, the causes of displacement can be identified, such as uneven stress distribution during the curing process and equipment precision problems.

[0080] More specifically, based on trend characteristics, the processing method of the first stage of curing is decomposed into process elements, such as curing temperature, time, pressure, and material properties. A probability analysis of processing error is performed on each process element, and the degree of influence of each element on positional offset is calculated. For example, through experiments and data analysis, it is determined that the fluctuation of curing temperature has a greater probability of affecting positional offset, so the optimization weight of this element is relatively high. Analyzing the optimization weight of process elements can help determine the focus of optimization, concentrate resources and efforts on elements that have a greater impact on positional offset, and more effectively improve processing accuracy and product quality.

[0081] More specifically, based on the optimization weights of each processing element, the processing method is simulated and its feasibility is evaluated by replacing the processing element. For example, the setting range of curing temperature is changed to simulate the processing effect at different temperatures and evaluate its feasibility. Based on the simulation and evaluation results, an optimized processing method is generated, including new processing parameters and process flow. Through replacement simulation and feasibility evaluation, the effect of the optimized scheme can be predicted before actual application, avoiding the risks of blind adjustment. The generated optimized processing method can more effectively improve the processing accuracy and quality of pixel units, reduce the fine-tuning rate and rework rate, and improve production efficiency and economic benefits.

[0082] Furthermore, the step of optimizing the processing method of the first segment of the solidification process of the pixel unit based on the processing supervision information includes: S101: Based on the processing supervision information, evaluate the effectiveness of the processing method of the first segment of the pixel unit solidification process to obtain the benchmark value parameter of the processing method; S102: Evaluate the benchmark value parameters for each time period, and select the fine-tuning instructions of the processing supervision information within the specified time range as the feedback analysis objects; S103: Perform fine-tuned analysis on the feedback analysis object to obtain the trend characteristics of the displacement of the pixel unit; S104: Based on the trend characteristics, the processing method of the first stage of curing is decomposed into processing elements, and a probability analysis of the processing error of each processing element is performed to obtain the optimized weight of each processing element. S105: Based on the optimization weights of each mode element, perform mode element replacement simulation and feasibility assessment on the processing mode to generate an optimized processing mode.

[0083] Specifically, data related to the first stage of solidification processing is extracted from the processing supervision information database, including the number of qualified, fine-tuning, and rework instructions, as well as the position offset data of the corresponding pixel units. Based on the extracted data, a series of indicators reflecting the effectiveness of the processing method are calculated, such as the pass rate (number of qualified pixel units / total number of processed pixel units), the fine-tuning rate (number of pixel units requiring fine-tuning / total number of processed pixel units), and the rework rate (number of pixel units requiring rework / total number of processed pixel units). These indicators constitute the benchmark value parameters of the processing method. Weighted averaging and other methods can be used to comprehensively evaluate the various indicators to obtain a comprehensive benchmark value parameter to fully reflect the effectiveness of the processing method.

[0084] More specifically, by calculating benchmark value parameters, the effectiveness of processing methods can be quantified, making the evaluation results more intuitive and accurate. This helps companies clearly understand the actual effect of the current processing methods and provides basic data for subsequent optimization. Benchmark value parameters can serve as a basis for comparison between different processing methods or different time periods of the same processing method. Through comparison, the room for improvement and existing problems of the processing methods can be identified.

[0085] More specifically, the processing supervision information is grouped according to time sequence, such as by day, week, month, etc. The benchmark value parameters of each time period are analyzed to observe their changing trends. Visualization tools such as line charts can be used to show the changes of parameters over time, so as to more intuitively discover fluctuations and anomalies. Based on the results of trend analysis, the fine-tuning instructions in the processing supervision information within the time period where the benchmark value parameters fluctuate greatly or perform poorly are selected as feedback analysis objects. These time periods often reflect the stages in the processing process where problems may exist.

[0086] More specifically, selecting fine-tuning instructions within a specific time period as the feedback analysis target allows for a focus on stages where problems frequently occur during processing, enabling more targeted analysis and improvement. By evaluating the benchmark value parameters for different time periods, periodic or phased patterns in the processing can be identified, helping to take preventative measures to avoid the recurrence of problems.

[0087] More specifically, the selected feedback parsing object (fine-tuning instruction) is organized, and detailed information about the fine-tuning operation is extracted, such as the direction, distance, and angle of adjustment. Statistical methods are used to analyze the organized data to find the trend characteristics of pixel unit displacement, such as calculating the frequency of displacement in different directions and the distribution of displacement distance. Machine learning or data mining algorithms can be used to perform pattern recognition on the data to discover the hidden patterns and trends in the data.

[0088] More specifically, by analyzing the fine-tuning instructions, we can gain a deeper understanding of the specific circumstances and trends of pixel unit displacement, providing a basis for finding the root cause of the problem. Trend characteristics can provide direction for subsequent processing optimization. For example, if the displacement in a certain direction is found to be more frequent, we can focus on the processing factors in that direction.

[0089] More specifically, the first stage of curing is decomposed into multiple process elements, such as curing temperature, curing time, pressure, and material properties. Based on trend characteristics, a probability analysis of processing error is performed on each process element. This can be achieved through experimental design and historical data statistics to determine the correlation and probability of influence between each element and displacement. Based on the results of the probability analysis, the optimization weight of each process element is calculated. A higher weight indicates a greater impact of that element on processing error, and thus requires more optimization.

[0090] More specifically, by breaking down the processing methods and analyzing their weights, we can identify the key elements that need to be optimized, avoid making blind adjustments, improve the efficiency and focus of optimization, and provide a basis for resource allocation, allowing more resources to be invested in elements that have a greater impact on processing errors, thus achieving the rational use of resources.

[0091] More specifically, based on the optimization weights of each factor in the processing method, replacement simulations are performed on the factors in the processing method. For example, the setting value of the curing temperature is tried to simulate the processing effect at different temperatures. The feasibility assessment of the result of the factor replacement simulation is conducted, considering factors such as technical feasibility, cost feasibility, and the impact on other processing stages. This assessment can be carried out through experimental verification, cost-benefit analysis, and other methods. Based on the results of the feasibility assessment, an optimized processing method is generated. The optimized processing method should comprehensively consider the adjustment schemes of each factor to ensure that while improving processing accuracy, production efficiency and cost rationality are guaranteed.

[0092] More specifically, through element replacement simulation and feasibility assessment, the effects and potential problems of the optimization scheme can be predicted before actual application, reducing the risks in the optimization process. The optimized processing method generated based on the assessment results can more effectively improve the processing accuracy and quality of pixel units, while taking into account production efficiency and cost, and achieving continuous improvement of the processing method.

[0093] Furthermore, the feature is that the design drawings and calibration records of the corresponding pixel units are obtained, and several offset states are generated for the pixel units after the first stage of processing based on the design drawings and calibration records. Corresponding visual recognition logic and theoretical calibration strategies are generated for each offset state. The visual recognition logic and theoretical calibration strategies corresponding to each offset state are combined as a preset standard for comparing the position offset data.

[0094] Specifically, design drawings for the corresponding pixel units are obtained from the enterprise's design management system. These drawings contain detailed information such as the ideal position, size, and spacing of the Micro LED chip and micro driver IC on the pads. At the same time, calibration records are extracted from the quality management system or historical database of the production process. These records cover various offsets of pixel units that occurred in previous production processes, as well as the corresponding calibration methods and results. The obtained design drawings and calibration records are then organized and cleaned to remove duplicate, erroneous, or irrelevant data, ensuring the accuracy and completeness of the data. The design drawings are converted into a digital format for subsequent analysis and processing. The calibration records are categorized and labeled for easy retrieval and use.

[0095] More specifically, based on the ideal position information in the design drawings and the actual offsets that appear in the calibration records, computer simulation technology is used to generate several possible offset situations. For example, the chip's translational offset in the X, Y, and Z directions, as well as rotational offsets around different axes, can be simulated. Different offset amplitudes and angles can be set to cover various possible situations. Referring to actual cases in the calibration records, some common or representative offset situations are included in the generated offset situation list. At the same time, considering new offset situations that may occur during the production process, the simulated offset situations can be appropriately expanded and supplemented.

[0096] More specifically, for each generated offset condition, we analyze how to accurately identify the offset using optical image detection, determine the features that need to be identified, such as the edges, corners, and center positions of the chip and driver IC, and the variation patterns of these features under different offset conditions. Based on the analysis results, we formulate corresponding visual recognition algorithms and rules, such as using specific image processing algorithms for edge detection and template matching, to achieve accurate identification of the offset condition. According to the characteristics of the offset condition and the requirements of the production process, we formulate corresponding theoretical adjustment strategies. For translational offset, we can determine the direction and distance to be adjusted; for rotational offset, we determine the rotation angle and axis. At the same time, considering the feasibility and accuracy requirements in actual operation, we optimize and refine the adjustment strategies to ensure that the offset can be effectively corrected.

[0097] More specifically, the visual recognition logic and theoretical calibration strategies corresponding to various deviation conditions are integrated to form a complete preset standard. This information can be stored in a database to establish a mapping relationship between deviation conditions, visual recognition logic, and theoretical calibration strategies, facilitating subsequent queries and comparisons. The preset standard is then verified through testing with actual production data and experiments to ensure its accuracy and effectiveness. As production processes improve and new problems arise, the preset standard is updated and improved in a timely manner to ensure it can adapt to ever-changing production needs.

[0098] More specifically, by combining design drawings and calibration records to generate preset standards, various possible offset situations can be accurately defined, and corresponding visual recognition logic can be formulated. This enables more accurate identification of pixel unit position offsets during actual inspection, avoiding misjudgments and omissions, and improving the accuracy and reliability of inspection. The theoretical calibration strategies in the preset standards provide clear guidance for subsequent fine-tuning operations. When a pixel unit position offset is detected, it can be quickly and accurately adjusted according to the corresponding calibration strategies in the preset standards, improving the efficiency and accuracy of calibration and reducing product quality problems caused by improper calibration.

[0099] More specifically, establishing preset standards helps optimize the production process. By analyzing and summarizing various deviations, problems and potential risks in the production process can be identified. Corresponding improvement measures can be taken to address these problems, such as adjusting the parameters of production equipment and optimizing the process flow, thereby improving the stability of the production process and product quality. Integrating various deviations and their corresponding handling methods into preset standards achieves standardized management of the production process. This helps improve production efficiency, reduce production costs, and also facilitates the monitoring and evaluation of the production process, ensuring that product quality meets unified standards and requirements.

[0100] Furthermore, the design drawings and calibration records are analyzed and processed through both automated data summarization algorithms and manual processing to analyze visual recognition logic and theoretical calibration strategies.

[0101] Specifically, the automated algorithm first performs preliminary cleaning and transformation on the design drawings and calibration records. For the design drawings, they are converted from specific graphic formats (such as CAD format) into data matrices that the algorithm can process, and key geometric information, such as the coordinates and dimensions of the chip and driver IC, is extracted. For the calibration records, they are processed into standardized tabular data, the data format is unified, and duplicate and invalid data are removed. Feature extraction algorithms are used to mine useful information from the preprocessed data. For example, edge detection algorithms are used to identify the contour features of the chip and driver IC from the design drawings. For the calibration records, features such as offset type and calibration parameters are extracted, and the frequency and distribution patterns of different features are calculated through statistical analysis.

[0102] More specifically, machine learning algorithms such as cluster analysis and decision trees are used to perform pattern recognition on features, automatically classifying the offset conditions in design drawings and calibration records. For example, translational offsets and rotational offsets of different degrees and directions are classified separately. Based on the features and classification results, visual recognition rules corresponding to different offset conditions are generated. For example, for a certain translational offset, the algorithm can determine which regions in the optical inspection image to focus on feature changes to identify the offset.

[0103] More specifically, based on the classified offset status and historical calibration records, the algorithm can learn corresponding theoretical calibration strategies. Through regression analysis, it can establish a mathematical model between offset parameters and calibration parameters, and predict the best calibration method to be adopted for different offsets, such as the direction and distance of adjustment.

[0104] More specifically, design drawings and calibration records typically contain a large amount of data. Automated data summarization algorithms can process this data quickly, improving analysis efficiency. They can perform feature extraction and pattern recognition on massive amounts of data in a short time, providing timely support for subsequent decision-making. Through complex mathematical operations and analysis, algorithms can discover potential patterns and rules in the data that are not easily noticed by humans. For example, the correlation between some subtle offset features and calibration methods is difficult to discover directly by humans, but algorithms can reveal these relationships through learning from a large amount of data. Automated algorithms can perform analysis according to preset rules and processes, reducing interference from human factors and ensuring the accuracy and consistency of analysis results. Each time the same data is analyzed, the same results are obtained, avoiding the subjectivity and uncertainty of human judgment.

[0105] More specifically, experienced engineers or technicians review the design drawings and calibration records. With their professional knowledge and industry experience, they can intuitively understand the intent of the design drawings and the actual situation of the calibration records. For example, they can determine whether certain offsets conform to the physical laws in actual production, or conduct preliminary analysis of special offsets based on past practical experience. The results of the classification and analysis by the automated algorithm are checked and screened manually to identify key issues that the algorithm has missed or misjudged. For example, some complex offsets are difficult for the algorithm to classify accurately due to the special nature of the data. In this case, the human can make corrections and supplements according to the actual situation.

[0106] More specifically, by combining the analysis results of the automation algorithm with other factors in actual production, such as equipment performance and process limitations, a comprehensive judgment is made to further optimize and improve the visual recognition logic and theoretical calibration strategies. For example, in actual production, some calibration operations are limited by the accuracy of the equipment. Humans can adjust the theoretical calibration strategies generated by the algorithm based on these factors to make them more operable.

[0107] More specifically, manual processing can fully leverage the expertise and practical experience of engineers and technicians. They have in-depth understanding of the industry's specific requirements and the details of production processes, which can compensate for the shortcomings of automated algorithms in this regard. During the analysis process, they can make more reasonable judgments and adjustments to the results from the perspective of practical application. In actual production, some complex and special deviations may occur, which may not be accurately handled by predefined algorithms. Manual processing can flexibly analyze and judge according to specific circumstances and propose targeted solutions.

[0108] More specifically, manual review and optimization of the analysis results can ensure that the visual recognition logic and theoretical adjustment strategies are operable and practical in actual production. Taking into account various limitations and actual needs in the production process, the analysis results can better serve production practice and improve product quality and production efficiency.

[0109] Based on the technical content of the online feedback-based Micro LED pixel unit processing accuracy control method described in the above-disclosed embodiments, the present invention provides an online feedback-based Micro LED pixel unit processing accuracy control system, the structure of which is as follows: Figure 2 The method for controlling the processing accuracy of Micro LED pixel units based on online feedback as described in any one of the first aspects includes: The preliminary processing module is used to place the Micro LED chip and the micro driver IC at designated positions on the pads, and to perform the first stage of curing processing on the Micro LED chip and the micro driver IC to obtain the first stage of processed pixel units; The offset recognition module is used to identify the position of the Micro LED chip and the micro driver IC relative to the pad of the pixel unit through the optical detection station, and to obtain the position offset data of the pixel unit. The instruction processing module is used to compare the position offset data according to a preset standard, and select from the three levels of instructions (qualified, fine-tuning, and rework) based on the comparison result to generate the corresponding work instruction to perform the corresponding processing work. The subsequent processing module is used to perform a second-stage curing process and subsequent processing steps on the pixel units processed in the first stage to prepare the finished pixel units.

[0110] In this embodiment, the specific implementation of each module in the above system embodiment is described in the above method embodiment, and will not be repeated here.

[0111] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0112] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0113] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for controlling the processing accuracy of Micro LED pixel units based on online feedback, characterized in that, include: The Micro LED chip and the micro driver IC are placed at designated positions on the pads, and the Micro LED chip and the micro driver IC are subjected to the first stage of curing to obtain the first stage of processed pixel units. The positional offset data of the pixel unit is obtained by identifying the position of the Micro LED chip and the micro driver IC relative to the pad through the optical inspection station. The position offset data is compared according to a preset standard, and corresponding work instructions are generated based on the comparison results to perform the corresponding processing work. The pixel unit processed in the first stage is subjected to a second stage of curing and subsequent processing steps to prepare the finished pixel unit.

2. The method for controlling the processing accuracy of Micro LED pixel units based on online feedback as described in claim 1, characterized in that, The step of identifying the position of the Micro LED chip and the micro driver IC relative to the pads in the pixel unit through an optical inspection station to obtain the position offset data of the pixel unit includes: The pixel unit is transported to the optical inspection station, where the pixel unit is image acquired at a specified resolution to obtain an optical inspection image. The optical detection image is subjected to feature recognition of the Micro LED chip, micro driver IC and pads to obtain positioning recognition data of the Micro LED chip, micro driver IC and pads; Based on the positioning and identification data, the position offset of the Micro LED chip and the micro driver IC relative to the pads is analyzed to obtain the position offset data of the pixel unit.

3. The method for controlling the processing accuracy of Micro LED pixel units based on online feedback as described in claim 2, characterized in that, The optical inspection station performs multi-view image acquisition on the pixel unit to obtain an optical inspection image. Based on the positional offset data of the optical inspection image, it determines whether the pixel unit needs to be supplemented for inspection. If supplementary inspection is required, the viewing angle of the second round of inspection is obtained by analyzing the positional offset data. The optical inspection station then adjusts the viewing angle of the image acquisition according to the viewing angle to obtain the second round of optical inspection image.

4. The method for controlling the processing accuracy of Micro LED pixel units based on online feedback as described in claim 1, characterized in that, The steps for generating corresponding work instructions based on the comparison results to execute the corresponding processing work include: Based on the comparison results, corresponding work instructions are generated, wherein the work instructions include any one of the following: a qualified instruction, a fine-tuning instruction, and a rework instruction; When the work instruction is a qualified instruction, the process jumps to the second stage of solidification processing and subsequent processing steps for the pixel units processed in the first stage. When the working instruction is a fine-tuning instruction, the die bonding head or fine-tuning mechanism is driven to physically correct the position of the MicroLED chip and the micro driver IC according to the fine-tuning instruction. When the work instruction is a rework instruction, the pixel unit processed in the first stage is arranged to enter the waste process to disassemble and reuse the pixel unit.

5. The method for controlling the processing accuracy of Micro LED pixel units based on online feedback as described in claim 4, characterized in that, The fine-tuning instruction includes theoretical calibration strategies required for the fine-tuning operation. These theoretical calibration strategies include several calibration stages. While executing the first calibration stage, the calibration effect is monitored through the optical inspection station. Based on the feedback obtained from the monitoring, the calibration method to be executed in subsequent calibration stages is determined.

6. The method for controlling the processing accuracy of Micro LED pixel units based on online feedback as described in claim 4, characterized in that, The qualified instructions, fine-tuning instructions, and rework instructions are recorded to construct processing supervision information for feedback on the first segment of the pixel unit's solidification process, and the processing method of the first segment of the pixel unit's solidification process is optimized based on the processing supervision information.

7. The method for controlling the processing accuracy of Micro LED pixel units based on online feedback as described in claim 6, characterized in that, The steps for optimizing the processing method of the first segment of the solidification process of the pixel unit based on the processing supervision information include: The effectiveness of the processing method for the first segment of the pixel unit solidification process is evaluated based on the processing supervision information to obtain the benchmark value parameter of the processing method. The baseline value parameters for each time period are evaluated, and the fine-tuning instructions of the processing supervision information within the specified time range are selected as the feedback analysis objects. The feedback parsing object is finely analyzed to obtain the trend characteristics of the displacement of the pixel unit; Based on the aforementioned trend characteristics, the processing method of the first stage of curing is decomposed into processing elements, and a probability analysis of processing error is performed on each processing element to obtain the optimized weight of each processing element. Based on the optimization weights of each factor, the processing method is simulated and its feasibility is evaluated by replacing the factor, so as to generate an optimized processing method.

8. The method for controlling the processing accuracy of Micro LED pixel units based on online feedback as described in claim 1, characterized in that, Obtain the design drawings and calibration records of the corresponding pixel units. Based on the design drawings and calibration records, generate several offset states for the pixel units after the first stage of processing. Analyze and generate corresponding visual recognition logic and theoretical calibration strategies for each offset state. Combine the visual recognition logic and theoretical calibration strategies corresponding to each offset state as a preset standard for comparing the position offset data.

9. The method for controlling the processing accuracy of Micro LED pixel units based on online feedback as described in claim 8, characterized in that, The design drawings and calibration records are analyzed and processed through both automated data summarization algorithms and manual processing to analyze visual recognition logic and theoretical calibration strategies.

10. A Micro LED pixel unit processing precision control system based on online feedback, characterized in that, include: The preliminary processing module is used to place the Micro LED chip and the micro driver IC at designated positions on the pads, and to perform the first stage of curing processing on the Micro LED chip and the micro driver IC to obtain the first stage of processed pixel units; The offset recognition module is used to identify the position of the Micro LED chip and the micro driver IC relative to the pad of the pixel unit through the optical detection station, and to obtain the position offset data of the pixel unit. The instruction processing module is used to compare the position offset data according to a preset standard, and select from the three levels of instructions (qualified, fine-tuning, and rework) based on the comparison result to generate the corresponding work instruction to perform the corresponding processing work. The subsequent processing module is used to perform a second-stage curing process and subsequent processing steps on the pixel units processed in the first stage to prepare the finished pixel units.

Images