A method for splicing and calibration of multi-galvanometer collaborative laser processing

Abstract

The present application discloses a multi-galvanometer collaborative laser processing splicing calibration method, which relates to the field of additive manufacturing technology. This method divides the image to be processed into multiple sub-unit images through image segmentation processing technology and then completes the splicing process. Then, the calibration matrix of the nominal parameters of each galvanometer is used for calibration, which can take into account the optimal processing area of each galvanometer and the overall rotation and translation calibration of a single galvanometer while meeting the requirements of the overall processing area. Next, the global error threshold obtained by converting the processing results of the evaluation area using the virtual large field lens method is used to evaluate the processing quality to adjust the splicing calibration method, so that multiple galvanometers can use the calibration strategy of a single equivalent large field lens, and then consider the splicing problem and calibration correction problem between adjacent galvanometers one by one, thereby reducing the dependence of the overall system splicing accuracy on individual calibration, facilitating the elimination of calibration errors within the entire processing area and improving the processing effect.

Description

Technical Field

[0001] The present application relates to the technical field of additive manufacturing, and in particular to a stitching calibration method for multi-galvanometer collaborative laser processing. Background Art

[0002] 3D printing is the common name of Additive Manufacturing (AM). AM technology integrates computer-aided design, material processing and forming technology. Based on digital model files, through software and numerical control systems, special metal materials, non-metal materials and medical biological materials are stacked layer by layer in ways such as extrusion, sintering, melting, photocuring, and spraying to manufacture solid objects.

[0003] Galvanometer laser processing is a commonly used additive manufacturing technology. With the pursuit of large sizes and higher printing efficiency, the application of multi-galvanometer partition scanning and collaborative processing is becoming more and more common, which can be used to manufacture larger-sized workpieces and has high practical significance.

[0004] However, in a multi-galvanometer stitching system, the problems of multi-galvanometer synchronous control and precision processing of overlapping areas have become the key factors restricting the efficiency of the multi-galvanometer stitching system. These problems not only involve complex adjustment and testing processes, but may also lead to deviations in laser processing results. When multiple galvanometers work simultaneously, any position deviation, such as being not horizontal or deviating from the center point, will affect the final laser processing quality. The traditional single-unit independent calibration method is not only time-consuming and laborious, but also lacks overall consideration. Especially when the number of galvanometers is large, manual adjustment is not only inefficient, but also difficult to ensure the accuracy of calibration and the consistency of errors, resulting in limited stitching accuracy and performance of the entire multi-galvanometer stitching system. Summary of the Invention

[0005] In view of the above problems and technical requirements, the present application proposes a stitching calibration method for multi-galvanometer collaborative laser processing. The technical solution of the present application is as follows:

[0006] A stitching calibration method for multi-galvanometer collaborative laser processing, the method comprising:

[0007] Dividing the image to be processed into multiple sub-unit images according to the scanning range of each galvanometer in the multi-galvanometer stitching system, each sub-unit image corresponding to one galvanometer in the multi-galvanometer stitching system and each sub-unit image not exceeding the scanning range of the corresponding galvanometer;

[0008] Extracting the edge pixels of each sub-unit image respectively, and performing interpolation fitting on the edge pixels of two adjacent sub-unit images to correct the stitching process of the two sub-unit images;

[0009] Calibrate the sub-unit images after stitching corresponding to the galvanometers using the nominal parameter calibration matrix of each galvanometer to obtain the calibrated sub-unit images;

[0010] Control each galvanometer to perform processing according to the corresponding calibrated sub-unit images, and evaluate the processing quality of the image to be processed using the global error threshold when it is determined that the calibration errors of all calibrated sub-unit images are qualified;

[0011] When the processing quality of the image to be processed is evaluated as unqualified using the global error threshold, adjust the interpolation fitting method and re-execute the step of interpolating and fitting the edge pixels of two adjacent sub-unit images to correct the stitching process of the two sub-unit images until the processing quality of the image to be processed is evaluated as qualified using the global error threshold, and then complete the stitching calibration of the multi-galvanometer stitching system; wherein, the global error threshold is obtained by converting the processing result of the evaluation area based on the virtual large field lens method by the multi-galvanometer stitching system.

[0012] A further technical solution thereof is that the multi-galvanometer collaborative laser processing stitching calibration method further includes:

[0013] Copy the rectangular evaluation area into a multi-stitching evaluation area in a k*k multi-stitching form, determine that the focal length of the virtual large field lens corresponding to the multi-stitching evaluation area is k*k times the focal length of the field lens corresponding to the original evaluation area, and the data representation accuracy of the multi-stitching evaluation area is k times the data representation accuracy of the original evaluation area; wherein, there are multiple galvanometers arranged in a k*k pattern in the multi-galvanometer stitching system, the evaluation area does not exceed the scanning range of a single galvanometer, and the integer parameter k≥2;

[0014] Calibrate the evaluation area corresponding to the galvanometer using the nominal parameter calibration matrix of each galvanometer to obtain the calibrated evaluation area, and control each galvanometer in the multi-galvanometer stitching system to perform processing according to the corresponding calibrated evaluation area respectively;

[0015] Convert the processing result of the multi-galvanometer stitching system for the multi-stitching evaluation area to obtain the global error threshold.

[0016] A further technical solution thereof is that converting the processing result of the multi-galvanometer stitching system for the multi-stitching evaluation area to obtain the global error threshold includes:

[0017] Reduce the processing result of the multi-stitching evaluation area and the maximum processing error of the multi-stitching evaluation area by k*k as the global error threshold.

[0018] A further technical solution thereof is that controlling each galvanometer to perform processing according to the corresponding calibrated sub-unit images includes:

[0019] All galvanometers are divided into multiple galvanometer splicing groups in sequence according to the processing order of the sub-unit images corresponding to each galvanometer. Each galvanometer splicing group includes multiple galvanometers;

[0020] Instruction data corresponding to each galvanometer is generated according to the pixel line processing trajectory of the calibrated sub-unit image of each sub-unit image. The instruction data of multiple galvanometers in the same galvanometer splicing group is encapsulated in the same galvanometer splicing group control instruction, and the galvanometer splicing group control instruction is sent to the multi-galvanometer splicing system to control multiple galvanometers in the same galvanometer splicing group. The number of galvanometers in the same galvanometer splicing group is determined based on the constraint that the update period of the galvanometer splicing group control instruction is less than the requirement of the galvanometer instruction update period.

[0021] A further technical solution thereof is that each galvanometer includes an x-axis and a y-axis. The instruction data of each galvanometer includes x-axis instruction data and y-axis instruction data. Each galvanometer splicing group constructed includes two galvanometers, and the obtained galvanometer splicing group control instruction sequentially includes a start bit, the x-axis instruction data of the first galvanometer in the galvanometer splicing group, the y-axis instruction data of the first galvanometer in the galvanometer splicing group, the x-axis instruction data of the second galvanometer in the galvanometer splicing group, the y-axis instruction data of the second galvanometer in the galvanometer splicing group, and a parity check bit, a total of 82 bits. The update frequency of each bit is 1 / 12.4 MHZ, and the update period of the galvanometer splicing group control instruction is 6.61 μs.

[0022] A further technical solution thereof is that interpolating and fitting the edge pixels of two adjacent sub-unit images to correct the splicing process of the two sub-unit images includes:

[0023] Using the bilinear interpolation algorithm to interpolate the edge pixels of two adjacent sub-unit images to obtain the edge pixel interpolation result, and merging the interpolated edge pixel interpolation result into one of the sub-unit images to correct the two sub-unit images.

[0024] A further technical solution thereof is that dividing the image to be processed into multiple sub-unit images includes:

[0025] Performing grid division on the rectangular area where the image to be processed is located. The sizes of each grid are the same and the range of each grid does not exceed the scanning range of the corresponding galvanometer;

[0026] Taking the image to be processed within each grid as a candidate sub-image, and correcting the candidate sub-image based on the image features of the image to be processed to obtain a sub-unit image.

[0027] A further technical solution thereof is that correcting the candidate sub-image based on the image features of the image to be processed to obtain a sub-unit image includes:

[0028] The K-Means clustering algorithm is used to perform pixel clustering on the image to be processed to obtain several clustered sub-images;

[0029] For any candidate sub-image, when there is a clustered sub-image completely within the range of the candidate sub-image, the clustered sub-image is used as the sub-unit image; otherwise, the candidate sub-image is directly used as the sub-unit image.

[0030] A further technical solution thereof is that the multi-galvo collaborative laser processing stitching calibration method further includes:

[0031] When the processing result of any calibrated sub-unit image and the processing error of the calibrated sub-unit image reach the sub-unit error threshold, adjust or replace the galvo corresponding to the calibrated sub-unit image; until the processing results of all calibrated sub-unit images and the processing errors of the corresponding calibrated sub-unit images are all less than the sub-unit error threshold, it is determined that the calibration errors of each calibrated sub-unit image are all evaluated as qualified. A further technical solution thereof is that evaluating the processing quality of the image to be processed by using the global error threshold includes:

[0032] When the maximum processing error between the processing result of the image to be processed and the image to be processed is less than the global error threshold, it is determined that the processing quality of the image to be processed is qualified; otherwise, it is determined that the processing quality of the image to be processed is unqualified.

[0033] The beneficial technical effects of this application are:

[0034] This application discloses a multi-galvo collaborative laser processing stitching calibration method. This method makes constraints and settings on the effective and optimal working areas of individual galvos through preliminary image segmentation processing, and then performs stitching calibration to meet the requirements of the overall processing format while taking into account the optimal processing areas of each galvo and the overall rotation and translation calibration of individual galvos. Then, the overall calibration data is split into local single or grouped galvo calibration compensation data. This method enables multiple galvos to use the calibration strategy of a single equivalent large field lens, and then consider the stitching problem and calibration correction problem between adjacent galvos one by one, thereby reducing the dependence of the overall system stitching accuracy on individual calibration. Utilizing the mapping relationship between matrix operations and multi-head stitching is beneficial to eliminating calibration errors within the entire processing format, improving the processing effect, and the "overall-zone-overall" calibration and evaluation scheme is beneficial to ensuring the consistency of the processing trajectory effects of each part during the processing of large-sized parts.

[0035] This method regards the image to be processed as a two-dimensional digital matrix, performs graphic processing, introduces a coefficient matrix of the same order for each unit, conducts mathematical modeling to include the actual differences such as individual rotation, translation, and scaling of each unit, discretizes the area to be processed, and the discretization method follows the general laws of the field lens optical angle and amplitude distortion. It maps the processed graphic pixel coordinate matrix to each single galvanometer, realizes overall stitching and independent operation of the local area of each single unit, and smoothly transitions to the overall stitching area. Without changing physical parameters such as the position of the galvanometer, it can minimize the deviation of the final laser processing result and achieve the same printing effect as when the galvanometer is located at the center point position.

[0036] This method designs appropriate control instruction groups for different numbers of galvanometers, ensures the synchronous control of each galvanometer and the real-time requirement of a 10us control cycle through the design of the galvanometer control instruction format, and at the same time takes into account the grouping of the control instruction transmission signals of the multi-galvanometer stitching system and the determination of the default rules of the stitching sequence, so as to improve the processing efficiency and control accuracy of the multi-galvanometer stitching system, reduce the demand for the number of galvanometer control cards in the multi-galvanometer stitching system, reduce the system cost and increase the design flexibility of the control system. The laser galvanometer control card can generate large-format continuous single-galvanometer instructions, cooperate with the data cutting and stitching processing algorithms, which is conducive to improving the processing efficiency. Description of the Drawings

[0037] Figure 1 It is a flowchart of the multi-galvanometer collaborative laser processing stitching calibration method according to an embodiment of the present application.

[0038] Figure 2 It is a schematic diagram of the image to be processed and the grid division of the rectangular area where it is located in an example of the present application.

[0039] Figure 3 It is the instruction format of the galvanometer stitching group control instruction in an embodiment of the present application. Detailed Embodiments

[0040] The following further describes the detailed embodiments of the present application with reference to the drawings.

[0041] The present application discloses a multi-galvanometer collaborative laser processing stitching calibration method. Please refer to Figure 1 the flowchart shown. The multi-galvanometer collaborative laser processing stitching calibration method includes:

[0042] Step 1, divide the image to be processed into multiple sub-unit images according to the scanning range of each galvanometer in the multi-galvanometer stitching system.

[0043] A mapping relationship is established between the multiple sub-unit images obtained by division and the multiple galvanometers in the multi-galvanometer stitching system. Each sub-unit image corresponds to one galvanometer in the multi-galvanometer stitching system and each sub-unit image does not exceed the scanning range of the corresponding galvanometer, ensuring that the galvanometer can at least cover and scan the corresponding sub-unit image. However, the number of sub-unit images obtained by division is not necessarily the same as the number of galvanometers in the multi-galvanometer stitching system. The method for obtaining multiple sub-unit images includes:

[0044] (1) First, perform grid division on the rectangular area where the image to be processed is located. The sizes of the grids are the same and the range of each grid does not exceed the scanning range of the corresponding galvanometer. The actual image to be processed may be a regular figure or an irregular figure. Therefore, first use 0 pixels to fill the image to be processed into a regular rectangular area. For example, Figure 2 in the schematic diagram, the image to be processed, as shown in Figure 2 (a), is an irregular figure. The rectangular area where the image to be processed is filled with 0 pixels is shown by the dotted line in Figure 2 (b). Then, perform grid division on the rectangular area where the image to be processed is located to obtain 12 grids.

[0045] All pixels within each grid obtained by division are valid pixels of the image to be processed, or only some pixels within each grid are valid pixels of the image to be processed, or all pixels within each grid are not valid pixels of the image to be processed. For example, in the example of Figure 2 (b), all pixels within the grids a2, a5, a6, a9, a10, and a11 obtained by division are valid pixels of the image to be processed. Only some pixels within the grids a1, a7, a8, and a12 obtained by division are valid pixels of the image to be processed. The grids a3 and a4 obtained by division do not contain valid pixels of the image to be processed.

[0046] (2) Take the image to be processed within each grid as a candidate sub-image, and correct the candidate sub-image based on the image features of the image to be processed to obtain a sub-unit image.

[0047] As described above, the grids are obtained by dividing the regular rectangular area where the image to be processed is located. In addition to the image to be processed, the regular rectangular area also includes filled 0 pixels. Therefore, only extract the image to be processed within each grid, and the area without the image to be processed does not need to be scanned. For example, Figure 2 in the example of (b), the grids a3 and a4 do not contain candidate sub-images. The areas where the valid pixels of the image to be processed are located within the grids a1, a7, a8, and a12 are used as candidate sub-images, and the remaining white areas do not need to be scanned.

[0048] The candidate sub-images obtained by such extraction can be directly used as the sub-unit images for subsequent processing. However, in order to further reduce the amount of data processing, a step of correcting the candidate sub-images based on the image features of the image to be processed is also added, including: using the K-Means clustering algorithm to perform pixel clustering on the image to be processed to obtain several clustered sub-images. Then, for any candidate sub-image, when there is a clustered sub-image that is completely within the range of the candidate sub-image, the clustered sub-image is used to replace the candidate sub-image as the sub-unit image; otherwise, the candidate sub-image is directly used as the sub-unit image. By clustering similar pixel points, multiple clustered sub-images with similar features can be formed. Replacing the candidate sub-images with them can reduce the amount of data, and on the basis of obtaining the graphic data with the highest proximity to the original graphic as much as possible to ensure the graphic quality, the subsequent data processing efficiency can be improved.

[0049] Step 2: Extract the edge pixels of each sub-unit image respectively, and perform interpolation fitting on the edge pixels of two adjacent sub-unit images to correct the splicing process of the two sub-unit images.

[0050] In this step, the Canny edge detection method can be used to obtain the pixel gradient amplitude and direction. The gradient amplitude and direction reflect the changes of the sub-unit image, so as to obtain the edge pixels of the sub-unit image.

[0051] Then, the splicing process is performed using the edge pixels of the sub-unit image, including using the bilinear interpolation algorithm to interpolate the edge pixels of two adjacent sub-unit images to obtain the edge pixel interpolation result, and merging the interpolated edge pixel interpolation result into one of the sub-unit images to correct the two sub-unit images.

[0052] Step 3: Calibrate the sub-unit image after splicing corresponding to the galvanometer using the nominal parameter calibration matrix of each galvanometer to obtain the calibrated sub-unit image.

[0053] The nominal parameter calibration matrix of each galvanometer is used to characterize the distortion caused by the own nominal parameters of the galvanometer. The nominal parameter calibration matrix can be pre-calibrated, including: pre-controlling each galvanometer to process the target pattern points, comparing the processing results of the galvanometer on the target pattern points with the target pattern points, and fitting the transformation matrix between the two by the least squares method as the nominal parameter calibration matrix of the galvanometer. The nominal parameter calibration matrix of each galvanometer is obtained by combining the rotation matrix and the offset matrix. Since the nominal parameters of different galvanometers are different, the nominal parameter calibration matrices of each galvanometer are often different and can be pre-calibrated separately.

[0054] Thus, the calibrated sub-unit image calibrated using the nominal parameter calibration matrix of the galvanometer can realize the overall rotation and translation calibration functions of a single galvanometer without changing the physical parameters such as the position of the galvanometer.

[0055] Step 4: Control each galvanometer to process according to the calibrated sub-unit image corresponding to it.

[0056] First, generate the command data for the corresponding galvanometer according to the pixel line processing trajectory of the calibrated sub-unit image of each sub-unit image. When specifically generating the command data, it is also necessary to consider the matching relationship between the laser power and the speeds of each processing axis of the galvanometer, including the matching in spatial trajectory and laser control timing to achieve spatio-temporal synchronous control. In addition, to ensure that the overall processed pattern does not show discontinuity, start constraints and corner compensation will be added in the commands. This application does not elaborate on this part specifically.

[0057] In order to reduce the demand for the number of galvanometer control cards in the multi-galvanometer stitching system and increase the design flexibility of the control system, in this step, all galvanometers are sequentially divided into multiple galvanometer stitching groups according to the processing order of the sub-unit images corresponding to each galvanometer. Each galvanometer stitching group includes multiple galvanometers. Then, generate the command data for the corresponding galvanometer according to the calibrated sub-unit image of each sub-unit image, encapsulate the command data of multiple galvanometers in the same galvanometer stitching group into the same galvanometer stitching group control command, and send the galvanometer stitching group control command to the multi-galvanometer stitching system to control multiple galvanometers in the same galvanometer stitching group. The number of galvanometers in the same galvanometer stitching group is determined based on the constraint that the update period of the galvanometer stitching group control command is less than the requirement of the galvanometer command update period, that is, still ensuring to meet the synchronous control of each galvanometer and the real-time requirement of a 10 μs control period.

[0058] In one embodiment, each galvanometer includes an x-axis and a y-axis. Then, the command data generated for each galvanometer includes x-axis command data and y-axis command data. Thus, each galvanometer stitching group constructed includes two galvanometers, and the obtained galvanometer stitching group control command sequentially includes a start bit, the x-axis command data of the first galvanometer in the galvanometer stitching group, the y-axis command data of the first galvanometer in the galvanometer stitching group, the x-axis command data of the second galvanometer in the galvanometer stitching group, the y-axis command data of the second galvanometer in the galvanometer stitching group, and a parity check bit, as Figure 3 shown. Among them, the first galvanometer and the second galvanometer refer to the two galvanometers in the galvanometer stitching group and do not specifically refer to a certain galvanometer.

[0059] Among them, the start bit and the parity check bit each occupy one bit, and the command data of each processing axis of each galvanometer in each galvanometer stitching group occupies 20 bits. Therefore, the entire galvanometer stitching group control command occupies 82 bits in total. The update frequency of each bit is 1 / 12.4 MHZ, and the update period of the entire galvanometer stitching group control command is 6.61 μs, meeting the requirement of the 10 μs command update period of a conventional galvanometer.

[0060] When different numbers of galvanometers are included in a multi-galvanometer stitching system, multiple galvanometer stitching groups can be built with every two galvanometers as one galvanometer stitching group. Through the above design of the galvanometer control instruction format, it is ensured to meet the synchronous control of each galvanometer and the real-time requirement of a 10us control cycle. At the same time, it takes into account the grouping of the control instruction transmission signals of the multi-galvanometer stitching system and the determination of the default rules for the stitching sequence, so as to improve the processing efficiency and control accuracy of the multi-galvanometer stitching system, reduce the demand for the number of galvanometer control cards in the multi-galvanometer stitching system, lower the system cost and increase the design flexibility of the control system.

[0061] Step 5: First, evaluate the calibration errors of each calibrated sub-unit image respectively, including: obtaining the processing error by comparing the processing result of each calibrated sub-unit image with the calibrated sub-unit image through an image detection and processing tool. When the processing error between the processing result and the calibrated sub-unit image of any calibrated sub-unit image reaches the sub-unit error threshold, adjust or replace the galvanometer corresponding to this calibrated sub-unit image, and then re-control each galvanometer to process according to its corresponding calibrated sub-unit image, so as to solve the defects of the galvanometer itself or the deviation caused by non-horizontal installation or non-centered skew. Until the processing errors between the processing results and the corresponding calibrated sub-unit images of all calibrated sub-unit images are less than the sub-unit error threshold, it is determined that the calibration errors of all calibrated sub-unit images are evaluated as qualified.

[0062] Step 6: When it is determined that the calibration errors of all calibrated sub-unit images are evaluated as qualified, further evaluate the processing quality of the image to be processed by using the global error threshold, including: obtaining the maximum processing error by comparing the processing result of the image to be processed with the image to be processed through an image detection and processing tool. When the maximum processing error is less than the global error threshold, it is determined that the processing quality of the image to be processed is qualified; otherwise, it is determined that the processing quality of the image to be processed is unqualified.

[0063] When the processing quality of the image to be processed is unqualified, return to Step 2 to adjust the interpolation fitting method and re-execute the interpolation fitting of the edge pixels of two adjacent sub-unit images to correct the stitching process of the two sub-unit images. When adjusting the interpolation fitting method, the weight distribution in the bilinear interpolation algorithm can be adjusted to adjust the stitching effect. When the processing quality of the image to be processed is qualified, the stitching calibration of the multi-galvanometer stitching system is completed, and the multi-galvanometer stitching system can be used for collaborative processing according to the above stitching method.

[0064] The global error threshold used in this step for evaluating the processing quality is pre-converted from the processing result of the evaluation area by the multi-galvanometer stitching system based on the virtual large field lens method. The methods for obtaining the global error threshold include:

[0065] (1) The rectangular evaluation format is copied into a multi-stitching evaluation format in the form of k*k multi-stitching, where the integer parameter k ≥ 2 and is based on the number of galvanometers in the multi-galvanometer stitching system. The multi-galvanometer stitching system contains multiple galvanometers arranged in k*k patterns. The evaluation format is an arbitrarily selected format that does not exceed the scanning range of a single galvanometer.

[0066] The multi-stitched evaluation format can be considered a single large evaluation format, and the focal length of the virtual large field lens corresponding to the multi-stitched evaluation format is k*k times the focal length of the field lens corresponding to the original evaluation format. Furthermore, the data representation accuracy of the multi-stitched evaluation format is k times that of the original evaluation format. For example, if the data representation accuracy of the original evaluation format is 16 bits, the data representation accuracy of the 2*2 multi-stitched evaluation format is increased to 17 bits, a 2x improvement. The stitched multi-stitched evaluation format can be replaced with a virtual large field lens and a virtual large galvanometer model, thereby constructing a virtual large field lens evaluation model.

[0067] (2) Using the nominal parameter calibration matrix of each galvanometer, the evaluation format corresponding to the galvanometer is calibrated to obtain the calibrated evaluation format. Each galvanometer in the multi-galvanometer splicing system is controlled to process according to its corresponding calibrated evaluation format. This part is similar to the method of calibrating and processing the sub-unit image described above and will not be repeated here.

[0068] (3) Then, the global error threshold is obtained based on the processing results of the multi-stitching evaluation format converted by the multi-galvanometer stitching system. This includes: using image detection and processing tools to compare the processing results of the multi-stitching evaluation format with the multi-stitching evaluation format, obtaining the data error between the two as the maximum processing error, and then reducing the maximum processing error by k*k as the global error threshold. This method can magnify the nominal parameters of a single field lens to a certain multiple, and through the "virtual large field lens" method of "first magnification and then reduction", it indirectly associates with the actual single-unit galvanometer and field lens nominal parameter characteristics at the same time.

[0069] The above description is only a preferred embodiment of the present application, and the present application is not limited to the above embodiments. It is understood that other improvements and variations directly derived or imagined by those skilled in the art without departing from the spirit and concept of the present application should be considered to be included in the scope of protection of the present application.

Claims

1. A method for stitching and calibration of multi-galvanometer collaborative laser processing, characterized in that, The multi-galvo collaborative laser processing stitching calibration method includes: Dividing the image to be processed into multiple sub-unit images according to the scanning range of each galvo in the multi-galvo stitching system, where each sub-unit image corresponds to one galvo in the multi-galvo stitching system and each sub-unit image does not exceed the scanning range of the corresponding galvo; Extracting the edge pixels of each sub-unit image respectively, and performing interpolation fitting on the edge pixels of two adjacent sub-unit images to correct the stitching process of the two sub-unit images; Calibrating the stitched sub-unit image corresponding to the galvo using the nominal parameter calibration matrix of each galvo to obtain the calibrated sub-unit image; Controlling each galvo to process according to the calibrated sub-unit image corresponding to it respectively, and evaluating the processing quality of the image to be processed using the global error threshold when it is determined that the calibration errors of all calibrated sub-unit images are qualified; When the processing quality of the image to be processed is evaluated as unqualified using the global error threshold, adjusting the interpolation fitting method and re-executing the step of performing interpolation fitting on the edge pixels of two adjacent sub-unit images to correct the stitching process of the two sub-unit images until the processing quality of the image to be processed is evaluated as qualified using the global error threshold, and then completing the stitching calibration of the multi-galvo stitching system; wherein, the global error threshold is obtained by converting the processing result of the evaluation area based on the virtual large field lens method by the multi-galvo stitching system.

2. The multi-galvanometer collaborative laser processing stitching and calibration method according to claim 1, wherein The multi-galvo collaborative laser processing stitching calibration method further includes: Copying the rectangular evaluation area into a multi-stitching evaluation area in a k*k multi-stitching form, determining that the focal length of the virtual large field lens corresponding to the multi-stitching evaluation area is k*k times the focal length of the field lens corresponding to the original evaluation area, and the data representation accuracy of the multi-stitching evaluation area is k times the data representation accuracy of the original evaluation area; wherein, there are multiple galvos arranged in a k*k pattern in the multi-galvo stitching system, the evaluation area does not exceed the scanning range of a single galvo, and the integer parameter k≥2; Calibrating the evaluation area corresponding to the galvo using the nominal parameter calibration matrix of each galvo to obtain the calibrated evaluation area, and controlling each galvo in the multi-galvo stitching system to process according to the calibrated evaluation area corresponding to it respectively; Converting the processing result of the multi-galvo stitching system for the multi-stitching evaluation area to obtain the global error threshold.

3. The multi-galvanometer collaborative laser processing splicing calibration method according to claim 2, wherein, Converting the processing result of the multi-galvo stitching system for the multi-stitching evaluation area to obtain the global error threshold includes: Reducing the processing result of the multi-stitching evaluation area by k*k times the maximum processing error of the multi-stitching evaluation area as the global error threshold.

4. The multi-galvanometer collaborative laser processing splicing calibration method according to claim 1, characterized in that Controlling each galvo to process according to the calibrated sub-unit image corresponding to it respectively includes: Sequentially dividing all galvos into multiple galvo stitching groups according to the processing order of the sub-unit images corresponding to each galvo, and each galvo stitching group includes multiple galvos; Generate the instruction data for the corresponding galvanometer based on the pixel line processing trajectory of the calibrated subunit image of each subunit image, encapsulate the instruction data of multiple galvanometers in the same galvanometer stitching group into the same galvanometer stitching group control instruction, and send the galvanometer stitching group control instruction to the multi-galvanometer stitching system to control multiple galvanometers in the same galvanometer stitching group. The number of galvanometers in the same galvanometer stitching group is determined based on the constraint that the update period of the galvanometer stitching group control instruction is less than the requirement of the galvanometer instruction update period.

5. The multi-galvanometer collaborative laser processing stitching calibration method according to claim 4, wherein, Each galvanometer includes an x-axis and a y-axis. The instruction data of each galvanometer includes x-axis instruction data and y-axis instruction data. Each galvanometer stitching group constructed includes two galvanometers, and the obtained galvanometer stitching group control instruction sequentially includes a start bit, the x-axis instruction data of the first galvanometer in the galvanometer stitching group, the y-axis instruction data of the first galvanometer in the galvanometer stitching group, the x-axis instruction data of the second galvanometer in the galvanometer stitching group, the y-axis instruction data of the second galvanometer in the galvanometer stitching group, and a parity check bit, totaling 82 bits. The update frequency of each bit is 1 / 12.4 MHZ, and the update period of the galvanometer stitching group control instruction is 6.61 μs.

6. The multi-galvanometer collaborative laser processing stitching and calibration method according to claim 1, wherein, Interpolating and fitting the edge pixels of two adjacent subunit images to correct the stitching process of the two subunit images includes: Using the bilinear interpolation algorithm to interpolate the edge pixels of two adjacent subunit images to obtain the edge pixel interpolation result, and merging the interpolated edge pixel interpolation result into one of the subunit images to correct the two subunit images.

7. The multi-galvanometer collaborative laser processing splicing and calibration method according to claim 1, characterized in that Dividing the image to be processed into multiple subunit images includes: Performing grid division on the rectangular area where the image to be processed is located. The sizes of each grid are the same and the range of each grid does not exceed the scanning range of the corresponding galvanometer; Regarding the image to be processed within each grid as a candidate sub-image, and correcting the candidate sub-image based on the image features of the image to be processed to obtain the subunit image.

8. The multi-galvanometer collaborative laser processing stitching and calibration method according to claim 7, wherein, Correcting the candidate sub-image based on the image features of the image to be processed to obtain the subunit image includes: Using the K-Means clustering algorithm to perform pixel clustering on the image to be processed to obtain several clustered sub-images; For any candidate sub-image, when there is a clustered sub-image completely within the range of the candidate sub-image, regarding the clustered sub-image as the subunit image, otherwise directly regarding the candidate sub-image as the subunit image.

9. The multi-galvanometer collaborative laser processing splicing calibration method according to claim 1, wherein The multi-galvanometer collaborative laser processing stitching calibration method further includes: When the processing result of any calibrated subunit image has a processing error with the calibrated subunit image reaching the subunit error threshold, adjust or replace the galvanometer corresponding to the calibrated subunit image; until the processing results of all calibrated subunit images have processing errors with the corresponding calibrated subunit images less than the subunit error threshold, it is determined that the calibration errors of each calibrated subunit image are all evaluated as qualified.

10. The multi-galvanometer collaborative laser processing splicing and calibration method according to claim 1, characterized in that, Evaluating the processing quality of the image to be processed using the global error threshold includes: When the processing result of the image to be processed has a maximum processing error less than the global error threshold with respect to the image to be processed, it is determined that the processing quality for the image to be processed is qualified; otherwise, it is determined that the processing quality for the image to be processed is unqualified.

Images