A method and system for multiple equal-division prediction alignment of an exposure machine

Abstract

The present invention relates to the technical field of printed circuit board manufacturing, and specifically to a multiple averaging prediction alignment method and system of an exposure machine, comprising the following steps: S1: acquiring multiple image data under different exposure conditions through an exposure machine; S2: performing image correction on the identified blurred area through an inverse motion compensation algorithm; S3: calculating the superposition error between each exposure image, and making a preliminary correction to the error; S4: dynamically adjusting the optical parameters of the exposure machine using an adaptive optical operator; S5: iteratively fusing the corrected image, and performing multiple iterative averaging processes; S6: transmitting the final alignment image data obtained in S5 to the control process of the exposure machine. The present invention significantly improves the image alignment accuracy and clarity during the PCB exposure process by integrating motion data analysis, inverse motion compensation, multiple averaging prediction alignment algorithms, and optical parameter adaptive adjustment technology, thereby ensuring high-quality PCB manufacturing.

Description

Technical Field

[0001] The present invention relates to the technical field of printed circuit board manufacturing, and in particular to a multiple equal-division prediction alignment method and system for an exposure machine. Background Art

[0002] In the modern PCB (printed circuit board) manufacturing process, the exposure machine is a key device for image transfer. It forms a circuit pattern by accurately transferring the design image from the photoresist film to the printed circuit board. In order to ensure high-quality image transfer, the requirements for the light source intensity, exposure time and image alignment accuracy of the exposure machine are very strict. However, in actual operation, due to the slight movement of the PCB during the exposure process and the possible superposition errors during multiple exposures, problems such as image blur and misalignment often occur. These factors directly affect the accuracy and performance of the final product. Therefore, how to ensure high-precision alignment and clarity of the image under complex exposure conditions has become an important challenge in PCB manufacturing.

[0003] In the prior art, although a variety of alignment correction methods have been proposed, such as static parameter adjustment and single exposure correction, the effects of these methods are still limited when dealing with dynamic motion and multiple exposure superposition errors. Due to the lack of an effective recognition and correction mechanism for image blur caused by PCB movement, and the inability to effectively correct superposition errors under multiple exposure conditions, the prior art is difficult to meet application scenarios with high precision and high consistency requirements.

[0004] Therefore, there is an urgent need for a multiple equal-division prediction alignment method and system for an exposure machine to achieve accurate alignment and high-quality transfer of PCB images, thereby improving the overall performance and manufacturing accuracy of the product. Summary of the invention

[0005] Based on the above objectives, the present invention provides a multiple equal-division prediction alignment method and system for an exposure machine.

[0006] A multiple equal-division prediction alignment method for an exposure machine includes the following steps:

[0007] S1: Acquire multiple image data under different exposure conditions through the exposure machine, and use the sensor to monitor the motion data of the PCB in real time, record the motion trajectory and speed change of the PCB, and associate the motion data with the corresponding image data for storage;

[0008] S2: Based on the PCB motion data obtained in S1, the associated image data is fuzzy identified, and the image is corrected on the identified fuzzy area through an inverse motion compensation algorithm;

[0009] S3: The image corrected in S2 is subjected to a superposition process of multiple exposure images, and the superposition error between the exposure images is calculated through a multiple average prediction alignment algorithm, and the error is preliminarily corrected;

[0010] S4: According to the superposition error correction result in S3, the optical parameters of the exposure machine, including the light source intensity and the exposure time, are dynamically adjusted using an adaptive optical operator to optimize the average predicted alignment effect of each exposure image;

[0011] S5: Based on the parameter optimization of S4, iterative fusion is performed on the corrected image, and the residual alignment error is corrected through multiple iterative averaging processes;

[0012] S6: The final alignment image data obtained in S5 is transmitted to the control process of the exposure machine to achieve the final transfer of the PCB image. At the same time, the error data generated during the iteration process is recorded and analyzed, and fed back to the image acquisition and correction process for optimization.

[0013] Optionally, the S1 specifically includes:

[0014] S11: setting a plurality of exposure conditions, including parameters of light source intensity, exposure time, and exposure angle, through a control unit of the exposure machine, wherein each set exposure condition corresponds to an image acquisition operation, and is used to obtain a plurality of image data under different exposure conditions;

[0015] S12: During the PCB processing, a multi-channel high-precision sensor network is configured. The sensors include laser displacement sensors and inertial measurement units to collect the motion data of the PCB in real time, including dynamic information of motion trajectory and speed change;

[0016] S13: During each image acquisition operation, the acquired image data is synchronized with the motion data recorded by the sensor using the synchronization control mechanism of the exposure machine to ensure that the time correlation between each image data and the corresponding motion data is consistent;

[0017] S14: Associating the synchronized image data with the corresponding motion data to generate association identifiers, and storing the association identifiers and the corresponding image and motion data in a data storage unit.

[0018] Optionally, the S2 specifically includes:

[0019] S21: Based on the PCB motion data obtained in step S1, the motion trajectory and speed change of the PCB are analyzed by the data processing unit, and the motion parameters of the PCB at different time points are extracted, including displacement, speed and acceleration information;

[0020] S22: according to the motion parameters obtained by the analysis in S21, the motion influence on each image pixel during the exposure process is calculated, the influence is accumulated to identify potential blur areas, and a corresponding blur area map is generated;

[0021] S23: Based on the blurred area identified in S22, an inverse motion compensation algorithm is applied to calculate each blurred area, and the motion trajectory of the PCB is reversely deduced to perform displacement correction on each affected pixel in the image, so as to restore the blurred pixels to their original positions;

[0022] S24: Output the image after compensation correction in S23 and replace the original blurred image data for subsequent image processing steps.

[0023] Optionally, the S22 specifically includes:

[0024] S221: Based on the PCB motion parameters extracted in S21, for each image pixel, calculate the displacement caused by the PCB motion during the exposure process. Specifically, for each pixel In time Movement displacement and The formula is:

[0025] ;

[0026] ,in, and Respectively represent the horizontal and vertical positions of the pixel in the image coordinate system; Indicates the time point in the exposure process; and Represents the pixel at time lateral and longitudinal displacements; and Respectively expressed in time The lateral and longitudinal velocities; and Respectively expressed in time lateral acceleration and longitudinal acceleration; Indicates the time interval between two adjacent time points;

[0027] S222: Based on the cumulative displacement of each pixel point calculated in S221, calculate its motion blur metric, and the calculation formula is: , Represents pixel The total fuzzy measure of and Represents pixel points In time lateral and longitudinal displacements; and Respectively indicate the time points at which the exposure starts and ends;

[0028] S223: The fuzzy metric calculated in S222 is With the preset blur threshold Compare and identify all Pixels that meet the conditions are marked as fuzzy areas;

[0029] S224: Generate a corresponding blurred area map according to the blurred area identified in S223, where the pixel points in the blurred area are assigned a value of 1 in the binary image, while the pixel points in the non-blurred area are assigned a value of 0, thereby generating a binary blurred area map that clearly identifies the blurred area .

[0030] Optionally, the S23 specifically includes:

[0031] S231: Based on the fuzzy area identified in step S22, reversely calculate the motion trajectory of the PCB, specifically for each affected pixel , according to the speed calculated in S221 and And acceleration and , reversely deduce the pixel point at the initial exposure moment location;

[0032] S232: using the initial position reversely derived in S231, correcting the displacement of each blurred pixel;

[0033] S233: Apply the pixel positions corrected in S232 to the image data to replace the original blurred pixel positions, so that all blurred pixels in the image are restored to their original clear positions.

[0034] Optionally, the S3 specifically includes:

[0035] S31: The corrected image data in S2 are sequentially stacked in chronological order. The image generated by each exposure is divided into a number of sub-areas. The image sub-areas corresponding to each exposure are stacked according to the corresponding positions. The stacked image area The formula is:

[0036] ,in, Indicates The first exposure generated sub-regions; Indicates the first sub-regions; Indicates the number of superimposed exposures;

[0037] S32: Perform error analysis on the image sub-regions after superposition in step S31 by using a multiple equal-division prediction alignment algorithm, specifically by calculating the displacement difference between each pixel before and after superposition. As superposition error;

[0038] S33: The stacking error calculated in S32 , a preliminary error correction algorithm is applied to each sub-region to adjust the horizontal and vertical displacements of the image so that the error between all superimposed sub-regions is minimized.

[0039] Optionally, the S4 specifically includes:

[0040] S41: Generate optical parameter adjustment data based on the superposition error correction result calculated in step S3, specifically, the superposition error of each sub-area Optical parameters The relationship between them is expressed as: ,in, Indicates The optical parameter adjustment amount of each sub-region; Indicates The superposition error of each sub-area; represents the initial setting value of the optical parameter; is the adjustment factor; represents the maximum stacking error among all sub-areas;

[0041] S42: Adaptive optics operators are used to adjust the optical parameters of each sub-region Make dynamic adjustments, including light intensity and exposure time , the optical operator automatically adjusts its optical parameters according to the superposition error of each sub-region, so that the image average prediction alignment effect of all sub-regions is optimized;

[0042] S43: Based on the optical parameters adjusted in S42, multiple iterations of optimization are performed so that the superposition errors of all sub-areas are gradually reduced to below a preset threshold. The optimization formula in the iteration process is: ,in, Indicates The sub-region is Optical parameters after iterations; Indicates The sub-region is Optical parameters after iterations; and Respectively represent The sub-region is Second and The stacking error after iterations; It is the iteration adjustment coefficient, which is used to control the adjustment amplitude of each iteration.

[0043] Optionally, the S5 specifically includes:

[0044] S51: After completing the optical parameter optimization in S4, an initial error evaluation is performed on the corrected image data. Specifically, for each sub-area Analyze the image and calculate the residual alignment error , the calculation formula is: ,in, Indicates Residual alignment error in each sub-region; and Respectively represent The corrected horizontal and vertical coordinates of each sub-region; and They represent the corresponding positions in the reference coordinate system respectively;

[0045] S52: residual alignment error calculated in S51 , the image is processed multiple times through the iterative averaging algorithm, and the formula for iterative averaging is: ,in, Indicates The sub-region is Image data after iterations; Indicates The sub-region is Image data after iterations; represents the reference image data, and the data of each sub-region is gradually adjusted to the reference position through averaging processing;

[0046] S53: Based on S52, error correction is performed on the image after each iteration of averaging, the coordinate value of each sub-region is updated, and the residual alignment error is recalculated;

[0047] S54: After each iteration, check whether the residual alignment errors of all sub-regions have converged to a preset threshold. Below, if the residual alignment errors of all sub-regions satisfy , the iteration is stopped; otherwise, the iterative averaging process is continued until the error converges.

[0048] Optionally, the S6 specifically includes:

[0049] S61: transmitting the final alignment image data obtained in S5 to the control process of the exposure machine through the data interface, and the transmission process includes encoding the corrected image data of each sub-area according to a predetermined data format and sending it to the exposure machine through a high-speed data bus;

[0050] S62: After the alignment image data is transmitted to the exposure machine control process, the received data is verified, specifically by comparing it with the original image data to verify whether the received data is consistent with the sent data;

[0051] S63: During the iteration process of S5, residual error data after each iteration is recorded in real time, where the recorded data includes the error value of each sub-region, the corresponding number of iterations, and the adjusted optical parameter value;

[0052] S64: Analyze the error data recorded in S63, identify the trend and law of error change, and generate a feedback model to guide subsequent image acquisition and correction processes;

[0053] S65: Pre-adjust exposure conditions and optical parameters using the feedback model to reduce initial errors.

[0054] A multiple averaging prediction alignment system for an exposure machine, used to implement the multiple averaging prediction alignment method for an exposure machine, comprises the following modules:

[0055] Image acquisition module: used to obtain multiple image data under different exposure conditions through the exposure machine, and use sensors to monitor the motion data of the PCB in real time, record the motion trajectory and speed changes of the PCB, and associate the motion data with the corresponding image data for storage;

[0056] Blur recognition module: Based on the PCB motion data provided by the image acquisition module, it calculates the motion impact of each image pixel during the exposure process, identifies potential blur areas, and generates corresponding blur area maps;

[0057] Inverse motion compensation module: Based on the identified blurred areas, the inverse motion compensation algorithm is applied to perform displacement correction on each blurred area to restore the blurred pixels to their original positions;

[0058] Image stacking correction module: The corrected image data is processed by stacking multiple exposure images, and the stacking error between each exposure image is calculated through the multiple average prediction alignment algorithm, and the error is preliminarily corrected;

[0059] Optical parameter optimization module: Based on the superposition error correction results, the adaptive optical operator is used to dynamically adjust the optical parameters of the exposure machine, including the light source intensity and exposure time, to optimize the average predicted alignment effect of each exposure image, and perform multiple iterations to optimize the optical parameters;

[0060] The iterative fusion convergence module corrects the residual alignment error through multiple iterative averaging processes based on the image data after optical parameter optimization, and makes a convergence judgment on the corrected image to ensure that the alignment error of all sub-areas meets the preset accuracy requirements;

[0061] Control feedback optimization module: transmits the final alignment image data to the control process of the exposure machine to achieve the final transfer of the PCB image, and records and analyzes the error data generated during the iteration process in real time to generate a feedback model to guide the optimization of optical parameters in the subsequent image acquisition and correction process.

[0062] Beneficial effects of the present invention:

[0063] The present invention effectively solves the problems of image blur and multiple exposure superposition error caused by tiny motion during PCB exposure by integrating real-time analysis of motion data, inverse motion compensation technology, multiple averaging prediction alignment algorithm, and adaptive adjustment of optical parameters. In particular, the inverse motion compensation technology allows accurate correction of image offset caused by dynamic motion of PCB, while the multiple averaging prediction alignment algorithm ensures accurate alignment of images during multiple exposures.

[0064] The present invention, by introducing adaptive optical parameter adjustment, can dynamically adjust the light source intensity and exposure time of the exposure machine according to the real-time error analysis results, thereby optimizing the image quality and alignment accuracy. This adaptive adjustment mechanism not only reduces the alignment error, but also improves the flexibility and efficiency of the entire exposure system. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0066] Figure 1 A schematic diagram of a multiple equal-division prediction alignment system according to an embodiment of the present invention;

[0067] Figure 2 Schematic diagram of the process flow of the multiple equal-division prediction alignment method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0068] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments, and those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0069] It should be noted that the references to "one embodiment", "embodiment", "exemplary embodiments", "some embodiments" and the like in the specification indicate that the embodiments described may include specific features, structures or characteristics, but not every embodiment may include the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the art to implement such feature, structure or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0070] In general, a term can be understood, at least in part, from its use in context. For example, depending, at least in part, on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending, at least in part, on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0071] like Figure 1 As shown, a multiple equal-division prediction alignment method of an exposure machine includes the following steps:

[0072] S1: Acquire multiple image data under different exposure conditions through the exposure machine, and use the sensor to monitor the motion data of the PCB in real time, record the motion trajectory and speed change of the PCB, and associate the motion data with the corresponding image data for storage;

[0073] S2: Based on the PCB motion data obtained in S1, the associated image data is fuzzy identified, and the image is corrected on the identified fuzzy area through an inverse motion compensation algorithm to reduce the image blur caused by the PCB motion;

[0074] S3: The image corrected in S2 is subjected to a superposition process of multiple exposure images, and the superposition error between each exposure image is calculated through a multiple average prediction alignment algorithm, and the error is preliminarily corrected to improve the alignment accuracy of the image superposition;

[0075] S4: According to the superposition error correction result in S3, the adaptive optical operator (AO) is used to dynamically adjust the optical parameters of the exposure machine, including the light source intensity and exposure time, so as to optimize the average predicted alignment effect of each exposure image and ensure the accurate alignment of subsequent exposure images;

[0076] S5: Based on the parameter optimization of S4, iterative fusion is performed on the corrected image. Through multiple iterative averaging processes, the residual alignment error is corrected to obtain a high-precision final image alignment result.

[0077] S6: The final alignment image data obtained in S5 is transmitted to the control process of the exposure machine to realize the final transfer of the PCB image. At the same time, the error data generated in the iterative process is recorded and analyzed, and fed back to the image acquisition and correction process for optimization, so as to further optimize the accuracy and stability of subsequent image alignment.

[0078] S1 specifically includes:

[0079] S11: setting a plurality of exposure conditions, including parameters of light source intensity, exposure time, and exposure angle, through a control unit of the exposure machine, wherein each set exposure condition corresponds to an image acquisition operation, and is used to obtain a plurality of image data under different exposure conditions;

[0080] S12: During the PCB processing, a multi-channel high-precision sensor network is configured. The sensors include laser displacement sensors and inertial measurement units (IMUs) to collect the motion data of the PCB in real time, including dynamic information of motion trajectory and speed changes;

[0081] S13: During each image acquisition operation, the acquired image data is synchronized with the motion data recorded by the sensor using the synchronization control mechanism of the exposure machine to ensure that the time correlation between each image data and the corresponding motion data is consistent;

[0082] The synchronization control mechanism includes the following technical steps:

[0083] Unified clock source: A unified high-precision clock source is configured in the control unit and sensor network of the exposure machine to provide a global time reference. All image acquisition and motion data recording operations are triggered and marked based on this unified clock to ensure that the timestamp of each data point is consistent.

[0084] Trigger synchronization signal: During each exposure, the exposure machine control unit will send a synchronization trigger signal to start the image sensor and motion sensor at the same time. This trigger signal can be implemented by hardware or software to ensure that image acquisition and motion data recording start at the same time;

[0085] Real-time data tagging: Every time the image sensor captures an image, the control unit assigns a timestamp to the image. Similarly, when the motion sensor records motion data, the same timestamp is also added to the data. In this way, each set of image data and its corresponding motion data can be accurately matched through the timestamp;

[0086] Delay compensation algorithm: Due to differences in hardware response time, there may be slight delays in the acquisition process of image sensors and motion sensors. Delay compensation can correct these time differences to ensure that the final data synchronization is accurate.

[0087] Synchronous data queue: During the data storage process, a synchronous data queue is used to store the collected images and motion data in timestamp order. Each set of image data will be automatically associated with its corresponding motion data to form a one-to-one matching relationship, ensuring data consistency in subsequent processing.

[0088] S14: Associating the synchronized image data with its corresponding motion data to generate an associated identifier, and storing these associated identifiers and their corresponding image and motion data in a data storage unit to ensure data consistency and traceability in subsequent processing; through the above-mentioned step design, not only can the image data be accurately acquired under different exposure conditions, but also the motion trajectory and speed change of the PCB can be monitored and recorded in real time, and these motion data can be accurately associated and stored with the corresponding image data, so that in the subsequent image processing and alignment correction, the error caused by the motion can be better compensated, further improving the alignment accuracy and image quality of the exposure machine.

[0089] S2 specifically includes:

[0090] S21: Based on the PCB motion data obtained in step S1, the motion trajectory and speed change of the PCB are analyzed by the data processing unit, and the motion parameters of the PCB at different time points are extracted, including displacement, speed and acceleration information;

[0091] S22: according to the motion parameters analyzed in S21, the motion influence on each image pixel during the exposure process is calculated, the influence is accumulated to identify the potential blur area, and a corresponding blur area map is generated. The identification process determines the blur area by calculating the motion displacement of the pixel and comparing it with a preset blur threshold.

[0092] S23: Based on the blurred area identified in S22, an inverse motion compensation algorithm is applied to calculate each blurred area, and the motion trajectory of the PCB is reversely deduced to perform displacement correction on each affected pixel in the image, so as to restore the blurred pixels to their original positions, thereby restoring the clarity of the image;

[0093] S24: The image after compensation correction in S23 is output and processed, and the original blurred image data is replaced to ensure that the corrected image has higher clarity and alignment accuracy for subsequent image processing steps; through the design of the above steps, the impact of PCB movement on the image can be accurately analyzed and calculated, and the clarity of the blurred area can be restored through an effective correction algorithm. The final correction processing ensures high-precision alignment of the image, which significantly improves the image quality and overall performance of the exposure machine under complex motion conditions.

[0094] S22 specifically includes:

[0095] S221: Based on the PCB motion parameters extracted in S21, for each image pixel, calculate the displacement caused by the PCB motion during the exposure process. Specifically, for each pixel In time Movement displacement and The formula is:

[0096] ;

[0097] ,in, and Respectively represent the horizontal and vertical positions of the pixel in the image coordinate system; Indicates the time point in the exposure process; and Represents the pixel at time lateral and longitudinal displacements; and Respectively expressed in time The lateral and longitudinal velocities; and Respectively expressed in time lateral acceleration and longitudinal acceleration; Represents the time interval between two adjacent time points; by calculating the displacement at each time point during the entire exposure process, these displacements are accumulated to determine the total motion impact of the pixel;

[0098] S222: Based on the cumulative displacement of each pixel point calculated in S221, calculate its motion blur metric, and the calculation formula is: , Represents pixel The total fuzzy measure of and Represents pixel points In time lateral and longitudinal displacements; and They represent the time points at which the exposure starts and ends, respectively. The blur metric of the pixel is calculated by accumulating the displacement during the entire exposure period.

[0099] S223: The fuzzy metric calculated in S222 is With the preset blur threshold Compare and identify all Pixels that meet the conditions are marked as fuzzy areas;

[0100] S224: Generate a corresponding blurred area map according to the blurred area identified in S223, where the pixel points in the blurred area are assigned a value of 1 in the binary image, while the pixel points in the non-blurred area are assigned a value of 0, thereby generating a binary blurred area map that clearly identifies the blurred area , the expression is:

[0101] ,in, Represents the pixel value in the fuzzy area map; Represents pixel The total fuzzy measure of represents the preset blur threshold; through the above steps, the consistency and accuracy of the parameters are ensured in the calculation step. This method can accurately calculate the cumulative displacement of each pixel due to the movement of the PCB during the exposure process, identify possible blur areas, and generate a clear binary blur area map. This precise identification and representation method provides a solid data foundation for subsequent image correction, and significantly improves the image alignment accuracy and overall image quality of the exposure machine.

[0102] S23 specifically includes:

[0103] S231: Based on the fuzzy area identified in step S22, reversely calculate the motion trajectory of the PCB, specifically for each affected pixel , according to the speed calculated in S221 and And acceleration and , reversely deduce the pixel point at the initial exposure moment The position of; the derivation formula is: ; ,in, and Respectively represent the horizontal and vertical positions of the pixel at the initial moment of exposure; and Respectively represent the horizontal and vertical positions of the pixel at the end of exposure; and denote the lateral and longitudinal velocities, respectively; and denote the lateral and longitudinal accelerations respectively; is the time differential;

[0104] S232: Using the initial position reversely derived in S231, the displacement of each blurred pixel is corrected. The correction formula is as follows: ; ,in, and Respectively represent the corrected horizontal and vertical displacements of the pixel;

[0105] S233: Apply the pixel position corrected in S232 to the image data, replace the original blurred pixel position, so that all blurred pixels in the image are restored to their original clear positions, thereby improving the overall clarity and alignment accuracy of the image; through the above-mentioned inverse motion compensation step, the motion trajectory of the PCB during the exposure process can be accurately reversed, and this information can be used to perform precise displacement correction on each pixel point in the blurred area. The corrected pixel points can be restored to their original positions, ensuring that the blurred area in the image is clearly restored. This process significantly improves the image processing accuracy of the exposure machine, especially in applications with high precision requirements, ensuring the quality and consistency of the final product.

[0106] S3 specifically includes:

[0107] S31: The corrected image data in S2 are sequentially stacked in chronological order. The image generated by each exposure is divided into a number of sub-areas. The image sub-areas corresponding to each exposure are stacked according to the corresponding positions. The stacked image area The formula is:

[0108] ,in, Indicates The first exposure generated sub-regions; Indicates the first sub-regions; Indicates the number of superimposed exposures;

[0109] S32: Perform error analysis on the image sub-regions after superposition in step S31 by using a multiple equal-division prediction alignment algorithm, specifically by calculating the displacement difference between each pixel before and after superposition. As the superposition error, the calculation formula is: ,in, and Respectively represent The horizontal and vertical coordinates of each sub-region at the first exposure; and Respectively represent the horizontal and vertical coordinates after superposition; Indicates The superposition error of each sub-area;

[0110] S33: The stacking error calculated in S32 , a preliminary error correction algorithm is applied to each sub-region to adjust the horizontal and vertical displacement of the image so that the error between all superimposed sub-regions is minimized; the corrected coordinates It is expressed as: ; ,in, and They represent the lateral and longitudinal error correction values ​​obtained by preliminary correction calculation respectively; and Respectively represent The corrected horizontal and vertical coordinates of each sub-region; through the design of the above steps, the multiple exposure images can be accurately superimposed, and the superposition error between the exposure images can be effectively calculated through the multiple average prediction alignment algorithm. Subsequently, the superposition error is corrected using the preliminary error correction algorithm to ensure that the superimposed image maintains consistency and high precision. This process significantly improves the alignment accuracy of the exposure machine when processing multiple exposure images, thereby enhancing the quality and reliability of the final image.

[0111] S4 specifically includes:

[0112] S41: Generate optical parameter adjustment data based on the superposition error correction result calculated in step S3, specifically, the superposition error of each sub-area Optical parameters The relationship between them is expressed as: ,in, Indicates The optical parameter adjustment amount of each sub-region; Indicates The superposition error of each sub-area; represents the initial setting value of the optical parameter; is the adjustment coefficient, which determines the influence of the superposition error on the adjustment of optical parameters; Indicates the maximum superposition error in all sub-areas and is used to standardize the superposition error; optical parameters According to the superposition error When the stacking error is large, the optical parameter adjustment amount will increase accordingly to optimize the exposure effect;

[0113] S42: Adaptive optics operators are used to adjust the optical parameters of each sub-region Make dynamic adjustments, including light intensity and exposure time , the optical operator automatically adjusts its optical parameters according to the superposition error of each sub-region, so that the image average prediction alignment effect of all sub-regions is optimized. The specific adjustment formula is:

[0114] ;

[0115] ,in, and Respectively represent Adjusted light intensity and exposure time for each sub-area; and Respectively represent the light source intensity and exposure time before adjustment; and is the adaptive adjustment coefficient;

[0116] S43: Based on the optical parameters adjusted in S42, multiple iterations of optimization are performed so that the superposition errors of all sub-areas are gradually reduced to below a preset threshold. The optimization formula in the iteration process is: ,in, Indicates The sub-region is Optical parameters after iterations; Indicates The sub-region is Optical parameters after iterations; and Respectively represent The sub-region is Second and The stacking error after iterations; It is an iterative adjustment coefficient used to control the adjustment amplitude of each iteration. Through the combination of the above steps, the relationship between optical parameters and superposition error can be described more accurately, and the optical parameters can be dynamically adjusted using adaptive optical operators. Through multiple iterative optimizations, the superposition error can be minimized, which significantly improves the image alignment accuracy of the exposure machine and the final image quality.

[0117] S5 specifically includes:

[0118] S51: After completing the optical parameter optimization in S4, an initial error evaluation is performed on the corrected image data. Specifically, for each sub-area Analyze the image and calculate the residual alignment error , the calculation formula is: ,in, Indicates Residual alignment error in each sub-region; and Respectively represent The corrected horizontal and vertical coordinates of each sub-region; and They represent the corresponding positions in the reference coordinate system respectively;

[0119] S52: residual alignment error calculated in S51 , the image is processed multiple times through the iterative averaging algorithm, and the formula for iterative averaging is: ,in, Indicates The sub-region is Image data after iterations; Indicates The sub-region is Image data after iterations; Represents the reference image data. Through averaging, the data of each sub-region is gradually adjusted to the reference position, and the residual error is gradually reduced;

[0120] S53: Based on S52, error correction is performed on the image after each iteration, the coordinate value of each sub-region is updated, and the residual alignment error is recalculated. The iteration formula is:

[0121] ,in, Indicates The sub-region is Residual alignment error after iterations; and Respectively represent The sub-region is The horizontal and vertical coordinates after iterations;

[0122] S54: After each iteration, check whether the residual alignment errors of all sub-regions have converged to a preset threshold. Below, if the residual alignment errors of all sub-regions satisfy , the iteration is stopped; otherwise, the iterative averaging process is continued until the error converges; through the above steps, on the basis of the optical parameter optimization in step S4, the residual alignment error of the image can be gradually reduced through multiple iterative averaging processes and error correction, and finally through the convergence judgment step, it is ensured that the alignment errors of all sub-areas meet the preset accuracy requirements, thereby significantly improving the image alignment accuracy and overall image quality of the exposure machine.

[0123] S6 specifically includes:

[0124] S61: The final alignment image data obtained in S5 is transmitted to the control process of the exposure machine through the data interface. The transmission process includes encoding the corrected image data of each sub-area according to a predetermined data format and sending it to the exposure machine control system through a high-speed data bus for subsequent PCB image transfer. During the transmission process, the data interface uses CRC (cyclic redundancy check) technology to verify the integrity of the data to ensure the accuracy of data transmission;

[0125] S62: After the alignment image data is transmitted to the exposure machine control process, the received data is verified, specifically by comparing it with the original image data to verify whether the received data is consistent with the sent data. During the verification process, if a data error is detected, a data retransmission mechanism is triggered to ensure that the data used for the final PCB image transfer is correct;

[0126] S63: During the iteration process of S5, the residual error data after each iteration is recorded in real time, and the recorded data includes the error value of each sub-region, the corresponding iteration number and the adjusted optical parameter value. These data are stored in the central data recording unit as the basis for subsequent analysis and optimization;

[0127] S64: Analyze the error data recorded in S63 to identify the trend and law of error change. The analysis process uses regression analysis and curve fitting technology to determine the relationship between the error and the adjustment of optical parameters, and generate a feedback model to guide subsequent image acquisition and correction processes.

[0128] The specific steps to generate the feedback model include:

[0129] S641: Collect the residual error data of each sub-region recorded in S63 , the corresponding number of iterations , and the adjusted optical parameter values ,These data points are organized into a data set for subsequent model training and analysis;

[0130] S642: Perform regression analysis on the data set collected in step S641 to determine the relationship between the error data and the optical parameters. Specifically, a multivariate linear regression model is used to establish a mathematical relationship between the error and the optical parameters. The regression model is expressed as: ,in: Indicates The sub-region is Residual error after iterations; Indicates The sub-region is Optical parameters after iterations; Indicates the number of iterations; is the regression coefficient, estimated by the least squares method; is the error term; after regression analysis, the regression results are subjected to curve fitting to generate a curve of the relationship between the error and the optical parameters, which is used to construct the feedback model;

[0131] S643: Based on the regression analysis and curve fitting results in step S642, a feedback model is generated. The feedback model is used to predict the error conditions under different optical parameter settings and provide optimization suggestions for the optical parameters. The expression of the feedback model is expressed as: ,in, Indicates The optical parameters after feedback optimization of each sub-region; The function representing the feedback model is obtained based on the above regression analysis; the feedback model adjusts the optical parameters by analyzing the current error level , to minimize the initial error generated in subsequent image acquisition;

[0132] S644: Apply the feedback model generated in step S643 to the subsequent image acquisition and correction process. Specifically, before a new round of image acquisition begins, adjust the initial optical parameter values ​​according to the prediction of the feedback model. , making it closer to the optimized optical parameters ,In the subsequent image acquisition, the exposure machine operates according to the ,adjusted initial parameters, thereby reducing the initial alignment ,error and optimizing the quality of image acquisition.

[0133] S65: Apply the feedback model obtained in S64 to the image acquisition and correction process. In the subsequent image acquisition and correction, the feedback model is used to pre-adjust the exposure conditions and optical parameters to reduce the initial error and optimize the image alignment effect. The specific formula for feedback optimization is: ,in, represents the optimized optical parameter value; Indicates the optical parameter value set initially; is the feedback adjustment coefficient; is the output value of the feedback model generated according to the error data; through the above steps, the present invention can safely and reliably transmit and verify the final alignment image data obtained in step S5 to ensure data integrity. At the same time, by recording and analyzing the error data generated in the iterative process, and feeding back the analysis results to the image acquisition and correction process, the image processing effect is further optimized. This process significantly improves the image alignment accuracy of the exposure machine and the overall performance of the system, ensuring the high-quality realization of PCB image transfer.

[0134] like Figure 2As shown, a multiple averaging prediction alignment system of an exposure machine is used to implement the above-mentioned multiple averaging prediction alignment method of an exposure machine, including the following modules:

[0135] Image acquisition module: used to obtain multiple image data under different exposure conditions through the exposure machine, and use sensors to monitor the motion data of the PCB in real time, record the motion trajectory and speed changes of the PCB, and associate the motion data with the corresponding image data for storage;

[0136] Blur recognition module: Based on the PCB motion data provided by the image acquisition module, it calculates the motion impact of each image pixel during the exposure process, identifies potential blur areas, and generates corresponding blur area maps;

[0137] Inverse motion compensation module: Based on the identified blurred areas, the inverse motion compensation algorithm is applied to perform displacement correction on each blurred area, restoring the blurred pixels to their original positions to improve the clarity and alignment accuracy of the image;

[0138] Image stacking correction module: The corrected image data is processed by stacking multiple exposure images, and the stacking error between each exposure image is calculated through a multiple average prediction alignment algorithm, and the error is preliminarily corrected to improve the alignment accuracy of image stacking;

[0139] Optical parameter optimization module: Based on the superposition error correction results, the adaptive optical operator is used to dynamically adjust the optical parameters of the exposure machine, including the light source intensity and exposure time, to optimize the average predicted alignment effect of each exposure image, and perform multiple iterations to optimize the optical parameters;

[0140] The iterative fusion convergence module corrects the residual alignment error through multiple iterative averaging processes based on the image data after optical parameter optimization, and makes a convergence judgment on the corrected image to ensure that the alignment error of all sub-areas meets the preset accuracy requirements;

[0141] Control feedback optimization module: transmits the final alignment image data to the control process of the exposure machine to achieve the final transfer of the PCB image, and records and analyzes the error data generated during the iteration process in real time, generates a feedback model to guide the optimization of optical parameters in the subsequent image acquisition and correction process, and improves the overall system performance and alignment accuracy.

[0142] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0143] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A multiple equal-division prediction alignment method for an exposure machine, characterized in that: The following steps are involved: S1: Acquire multiple image data under different exposure conditions through the exposure machine, and use the sensor to monitor the motion data of the PCB in real time, record the motion trajectory and speed change of the PCB, and associate the motion data with the corresponding image data for storage; S2: Based on the PCB motion data obtained in S1, the associated image data is fuzzy identified, and the image is corrected on the identified fuzzy area through an inverse motion compensation algorithm; S3: The image corrected in S2 is subjected to a superposition process of multiple exposure images, and the superposition error between the exposure images is calculated through a multiple average prediction alignment algorithm, and the error is preliminarily corrected; S4: According to the superposition error correction result in S3, the optical parameters of the exposure machine, including the light source intensity and the exposure time, are dynamically adjusted using an adaptive optical operator to optimize the average predicted alignment effect of each exposure image; S5: Based on the parameter optimization of S4, iterative fusion is performed on the corrected image, and the residual alignment error is corrected through multiple iterative averaging processes; S6: The final alignment image data obtained in S5 is transmitted to the control process of the exposure machine to achieve the final transfer of the PCB image. At the same time, the error data generated in the iteration process is recorded and analyzed, and fed back to the image acquisition and correction process for optimization; The S2 specifically includes: S21: Based on the PCB motion data obtained in step S1, the motion trajectory and speed change of the PCB are analyzed by the data processing unit, and the motion parameters of the PCB at different time points are extracted, including displacement, speed and acceleration information; S22: according to the motion parameters obtained by the analysis in S21, the motion influence on each image pixel during the exposure process is calculated, the influence is accumulated to identify potential blur areas, and a corresponding blur area map is generated; S23: Based on the blurred area identified in S22, an inverse motion compensation algorithm is applied to calculate each blurred area, and the motion trajectory of the PCB is reversely deduced to perform displacement correction on each affected pixel in the image, so as to restore the blurred pixels to their original positions; S24: Output the image after compensation correction in S23 and replace the original blurred image data for subsequent image processing steps.

2. The multiple equal-division prediction alignment method of an exposure machine according to claim 1, characterized in that: The S1 specifically includes: S11: setting a plurality of exposure conditions, including parameters of light source intensity, exposure time, and exposure angle, through a control unit of the exposure machine, wherein each set exposure condition corresponds to an image acquisition operation, and is used to obtain a plurality of image data under different exposure conditions; S12: During the PCB processing, a multi-channel high-precision sensor network is configured. The sensors include laser displacement sensors and inertial measurement units to collect the motion data of the PCB in real time, including dynamic information of motion trajectory and speed change; S13: During each image acquisition operation, the acquired image data is synchronized with the motion data recorded by the sensor using the synchronization control mechanism of the exposure machine to ensure that the time correlation between each image data and the corresponding motion data is consistent; S14: Associating the synchronized image data with the corresponding motion data to generate association identifiers, and storing the association identifiers and the corresponding image and motion data in a data storage unit.

3. The multiple equal-division prediction alignment method of an exposure machine according to claim 1, characterized in that: The S22 specifically includes: S221: Based on the PCB motion parameters extracted in S21, for each image pixel, calculate the displacement caused by the PCB motion during the exposure process. Specifically, for each pixel In time Movement displacement and The formula is expressed as: ； ,in, and Respectively represent the horizontal and vertical positions of the pixel in the image coordinate system; Indicates the time point in the exposure process; and Represents the pixel at time lateral and longitudinal displacements; and Respectively expressed in time The lateral and longitudinal velocities; and Respectively expressed in time lateral acceleration and longitudinal acceleration; Indicates the time interval between two adjacent time points; S222: Based on the cumulative displacement of each pixel point calculated in S221, calculate its motion blur metric, and the calculation formula is: , Represents pixel The total fuzzy measure of and Represents pixel points In time lateral and longitudinal displacements; and Respectively indicate the time points at which the exposure starts and ends; S223: The fuzzy metric calculated in S222 is With the preset blur threshold Compare and identify all Pixels that meet the conditions are marked as fuzzy areas; S224: Generate a corresponding blurred area map according to the blurred area identified in S223, where the pixel points in the blurred area are assigned a value of 1 in the binary image, while the pixel points in the non-blurred area are assigned a value of 0, thereby generating a binary blurred area map that clearly identifies the blurred area .

4. The multiple equal-division prediction alignment method of an exposure machine according to claim 3, characterized in that: The S23 specifically includes: S231: Based on the fuzzy area identified in step S22, reversely calculate the motion trajectory of the PCB, specifically for each affected pixel , according to the speed calculated in S221 and And acceleration and , reversely deduce the pixel point at the initial exposure moment location; S232: using the initial position reversely derived in S231, correcting the displacement of each blurred pixel; S233: Apply the pixel position corrected in S232 to the image data to replace the original blurred pixel position, so that all blurred pixels in the image are restored to their original clear positions.

5. The multiple equal-division prediction alignment method of an exposure machine according to claim 1, characterized in that: The S3 specifically includes: S31: The corrected image data in S2 are sequentially stacked in chronological order. The image generated by each exposure is divided into a number of sub-areas. The image sub-areas corresponding to each exposure are stacked according to the corresponding positions. The stacked image area The formula is expressed as: ,in, Indicates The first exposure generated sub-regions; Indicates the first sub-regions; Indicates the number of superimposed exposures; S32: Perform error analysis on the image sub-regions after superposition in step S31 by using a multiple equal-division prediction alignment algorithm, specifically by calculating the displacement difference between each pixel before and after superposition. As superposition error; S33: The stacking error calculated in S32 , a preliminary error correction algorithm is applied to each sub-region to adjust the horizontal and vertical displacements of the image so that the error between all superimposed sub-regions is minimized.

6. The multiple equal-division prediction alignment method of an exposure machine according to claim 1, characterized in that: The S4 specifically includes: S41: Generate optical parameter adjustment data based on the superposition error correction result calculated in step S3, specifically, the superposition error of each sub-area Optical parameters The relationship between them is expressed as: ,in, Indicates The optical parameter adjustment amount of each sub-region; Indicates The superposition error of each sub-area; Indicates the initial setting value of the optical parameter; is the adjustment factor; represents the maximum stacking error among all sub-areas; S42: Adaptive optics operators are used to adjust the optical parameters of each sub-region Make dynamic adjustments, including light intensity and exposure time , the optical operator automatically adjusts its optical parameters according to the superposition error of each sub-region, so that the image average prediction alignment effect of all sub-regions is optimized; S43: Based on the optical parameters adjusted in S42, multiple iterations of optimization are performed so that the superposition errors of all sub-areas are gradually reduced to below a preset threshold. The optimization formula in the iteration process is: ,in, Indicates The sub-region is Optical parameters after iterations; Indicates The sub-region is Optical parameters after iterations; and Respectively represent The sub-region is Second and The stacking error after iterations; It is the iteration adjustment coefficient, which is used to control the adjustment amplitude of each iteration.

7. The multiple equal-division prediction alignment method of an exposure machine according to claim 1, characterized in that: The S5 specifically includes: S51: After completing the optical parameter optimization in S4, an initial error evaluation is performed on the corrected image data. Specifically, for each sub-area Analyze the image and calculate the residual alignment error , the calculation formula is: ,in, Indicates Residual alignment error in each sub-region; and Respectively represent The corrected horizontal and vertical coordinates of each sub-region; and They represent the corresponding positions in the reference coordinate system respectively; S52: residual alignment error calculated in S51 , the image is processed multiple times through the iterative averaging algorithm, and the formula for iterative averaging is: ,in, Indicates The sub-region is Image data after iterations; Indicates The sub-region is Image data after iterations; represents the reference image data, and the data of each sub-region is gradually adjusted to the reference position through averaging processing; S53: Based on S52, error correction is performed on the image after each iteration of averaging, the coordinate value of each sub-region is updated, and the residual alignment error is recalculated; S54: After each iteration, check whether the residual alignment errors of all sub-regions have converged to a preset threshold. Below, if the residual alignment errors of all sub-regions satisfy , the iteration is stopped; otherwise, the iterative averaging process is continued until the error converges.

8. The multiple equal-division prediction alignment method of an exposure machine according to claim 1, characterized in that: The S6 specifically includes: S61: transmitting the final alignment image data obtained in S5 to the control process of the exposure machine through the data interface, and the transmission process includes encoding the corrected image data of each sub-area according to a predetermined data format and sending it to the exposure machine through a high-speed data bus; S62: After the alignment image data is transmitted to the exposure machine control process, the received data is verified, specifically by comparing it with the original image data to verify whether the received data is consistent with the sent data; S63: During the iteration process of S5, residual error data after each iteration is recorded in real time, where the recorded data includes the error value of each sub-region, the corresponding number of iterations, and the adjusted optical parameter value; S64: Analyze the error data recorded in S63, identify the trend and law of error change, and generate a feedback model to guide subsequent image acquisition and correction processes; S65: Pre-adjust exposure conditions and optical parameters using the feedback model to reduce initial errors.

9. A multiple equal-division prediction alignment system for an exposure machine, used to implement a multiple equal-division prediction alignment method for an exposure machine as claimed in any one of claims 1 to 8, characterized in that: Includes the following modules: Image acquisition module: used to obtain multiple image data under different exposure conditions through the exposure machine, and use sensors to monitor the motion data of the PCB in real time, record the motion trajectory and speed changes of the PCB, and associate the motion data with the corresponding image data for storage; Blur recognition module: Based on the PCB motion data provided by the image acquisition module, it calculates the motion impact of each image pixel during the exposure process, identifies potential blur areas, and generates corresponding blur area maps; Inverse motion compensation module: Based on the identified blurred areas, the inverse motion compensation algorithm is applied to perform displacement correction on each blurred area to restore the blurred pixels to their original positions; Image stacking correction module: The corrected image data is processed by stacking multiple exposure images, and the stacking error between each exposure image is calculated through the multiple average prediction alignment algorithm, and the error is preliminarily corrected; Optical parameter optimization module: Based on the superposition error correction results, the adaptive optical operator is used to dynamically adjust the optical parameters of the exposure machine, including the light source intensity and exposure time, to optimize the average predicted alignment effect of each exposure image, and perform multiple iterations to optimize the optical parameters; The iterative fusion convergence module corrects the residual alignment error through multiple iterative averaging processes based on the image data after optical parameter optimization, and makes a convergence judgment on the corrected image to ensure that the alignment error of all sub-areas meets the preset accuracy requirements; Control feedback optimization module: transmits the final alignment image data to the control process of the exposure machine to achieve the final transfer of the PCB image, and records and analyzes the error data generated during the iteration process in real time to generate a feedback model to guide the optimization of optical parameters in the subsequent image acquisition and correction process.

Images