Self-calibration method and system for large-format printed circuit board digital photoetching system

Abstract

The invention discloses a self-calibration method and system of a large-format printed circuit board digital photoetching system. The system comprises an industrial camera, a precision motion platform, a camera light source, a color-changing film module, a photoetching lens module and the like. The method comprises three calibration items, CCD position relation calibration, photoetching lens module position relation calibration, offset calibration of an alignment coordinate system and a photoetching coordinate system, photoetching of a circular target array containing calibration data by using a color-changing film module, and obtaining target position coordinates by using a sub-pixel circle recognition algorithm based on an improved Zernike moment. And comparing the data with system calibration reference data, and performing error compensation to realize self-calibration of the large-format printed circuit board digital photoetching system. According to the invention, various aspects of possible offset errors of the digital photoetching system are almost covered, the acquisition and compensation of error data are effectively realized, and meanwhile, the modularized design is adopted, so that other calibration items can be conveniently added in the later period.

Description

Technical Field

[0001] The present invention relates to the field of calibration research of lithography systems, and particularly to a self-calibration method and system for a large-format printed circuit board digital lithography system. Background Art

[0002] As a carrier for mounting electronic components and making electrical connections in circuits, a printed circuit board (PCB) is an important part of various types of electronic products. The development level of its production and manufacturing plays a fundamental role in the technological progress of industries such as computers, consumer electronics, communication equipment, automotive electronics, and industrial equipment. A large-format printed circuit board can further save materials and improve production efficiency. Under the same process flow, more PCB boards can be produced, creating higher benefits.

[0003] The PCB plate-making industry traditionally relies on a lithography process based on polyester masks, which is prone to causing distortion in PCB plate-making and it is difficult to obtain high-resolution PCB plate-making. Although traditional lithography machines can achieve high precision, the manufacturing cost of mask plates is high, the cycle is long, and the flexibility is poor, which is not suitable for large-scale production. Maskless lithography technology that does not require a physical mask plate has been widely studied. Common maskless lithography technologies include DMD (digital micromirror device) digital lithography, electron beam lithography, focused ion beam lithography, interference lithography, and laser direct writing technology, etc. Compared with digital lithography, other technologies are not suitable for large-scale production and manufacturing due to efficiency and equipment reasons. Therefore, digital lithography has been practically applied. However, during long-term operation, due to high temperature, mechanical wear, etc., the originally calibrated digital lithography system will produce offsets, lose precision, affect the lithography result, and need to be recalibrated.

[0004] The calibration of a digital lithography system is essentially the calibration of a projector. Currently, most calibration applications of projectors are for three-dimensional measurement systems. A projector can be regarded as a reversed camera, but the projector itself cannot capture images and can only rely on a camera. Therefore, when a calibrated camera captures a projection distortion image, using the projection camera model to convert the distorted image to the projection plane, it can be regarded as an image "captured" by the projection system. Then, the amount of distortion and the distortion parameters can be detected and obtained for this image, and the camera and the projection system are modeled as a whole. There are mainly three methods. The first method is the phase matching method: using the world coordinates of the feature points on a known calibration object and applying phase technology to find their image coordinates. The second method is the cross-ratio invariance method: using the world coordinates of the feature points on a known calibration object, then projecting a specific pattern onto the calibration object, and using the cross-ratio invariance to calculate the world coordinates of the feature points of the projection pattern. The third method is the back-projection method: given the image coordinates of the feature points of the projection pattern, using the calibrated camera model to back-project to find their world coordinates. The above methods all require a calibration board, and the camera field of view needs to cover both the projection image and the calibration board at the same time, which has problems in terms of accuracy and efficiency, and is not suitable for projection systems for large-area lithography.

[0005] The accuracy of the target recognition algorithm determines the accuracy of the acquired position data. Improving the equipment such as using an industrial camera with higher resolution can obtain higher accuracy, but it also means high equipment costs and huge image data, which affect the efficiency of self-calibration and the working efficiency of the digital lithography system. Therefore, it is necessary to further study the circle recognition algorithm to improve the recognition accuracy while ensuring the operation efficiency. The highest edge detection accuracy that traditional edge detection methods can achieve is only one pixel. In fact, the edges in the image may be at any position within this pixel. The sub-pixel edge detection technology based on Zernike moments has good anti-noise performance, can effectively suppress noise interference, improve the accuracy of positioning, and has rotational invariance and scale invariance. However, the traditional second-order ideal model ignores the edge transition. Therefore, domestic and foreign scholars have established a third-order gray-level transition edge model of the image edge based on the second-order gray-level step model of the ideal Zernike moments to improve its accuracy. Liu Jinming used a 7×7 template for the third-order model formula, but there is no theoretical derivation and it is not universal (Research on the Detection Algorithm for Surface Mounting Defects of Printed Circuit Boards Based on Image Processing, School of Software Engineering, Information Faculty, Liu Jinming).

[0006] The exposure system of the digital lithography system for large-format printed circuit boards has high requirements for accuracy. For the digital lithography system for large-format printed circuit boards, it is unrealistic to pause regularly during equipment operation and manually place the calibration board for measurement. Therefore, a self-calibration method for the digital lithography system for large-format printed circuit boards needs to be designed to make the entire system have closed-loop control and meet the accuracy and efficiency requirements of production at the same time. Summary of the Invention

[0007] The object of the present invention is to provide a self - calibration method for a large - format printed circuit board digital lithography system. This method can effectively solve the problem that during long - term operation, due to high temperature, mechanical wear, etc., the originally calibrated digital lithography system will generate offsets, lose accuracy, affect the lithography result, and require re - calibration. To achieve the above object, the present invention proposes a self - calibration method for a large - format printed circuit board digital lithography system.

[0008] The self - calibration method for a large - format printed circuit board digital lithography system includes the following steps:

[0009] Step 1: Calibrate to establish the coordinate relationship of the entire digital lithography system, that is, establish the conversion relationship between the lithography coordinate system and the alignment coordinate system. After the system runs stably and the error of manual self - calibration meets the requirements, establish a data benchmark.

[0010] Step 2: Perform target pattern lithography on the color - changing film module. Each lens lithographs a group of circles, and different - sized double concentric circles are used to distinguish the front and back rows, and adjust the distance between the front and back rows to ensure that the patterns do not overlap.

[0011] Step 3: Use the circular target lithographed on the color - changing film module in Step 2 to calibrate the position relationship of the CCD, that is, calibrate the position relationship of the left and right alignment industrial cameras. Use the two cameras to identify the same target dot, and obtain the position data of this dot under the two alignment industrial cameras for subsequent comparison.

[0012] Step 4: Use the circular target lithographed on the color - changing film module in Step 2 to calibrate the position relationship of the lithography lens module. Identify a group of circles corresponding to each lithography lens through the alignment industrial camera, obtain the position data of each dot for calculation, and obtain the magnification, angle, and spacing data of the lithography lens.

[0013] Step 5: Use the circular target lithographed on the color - changing film module in Step 2 to calibrate the offset between the alignment coordinate system and the lithography coordinate system, that is, calibrate the origin position of the lithography coordinate system. Identify a group of circles lithographed by the first lithography lens and obtain its position data.

[0014] Step 6: Compare the position data obtained in Steps 3 - 5 with the position data benchmark established in Step 1, perform error compensation, divide the lithography area by stripes, and allocate the compensated stripe data to the lithography lenses to complete self - calibration.

[0015] Further, in step 1, the data benchmark is established after multiple runs of automatic calibration. When the position error of the camera's point capture meets the specified requirements and stable calibration can be achieved, the data benchmark established through manual operation of automatic calibration is used. The data benchmark includes the position data of the industrial camera for aligning the lithography lens group, which is used for subsequent data comparison for calibration compensation.

[0016] Further, in step 2, when lithographing the target pattern, each lens lithographs three circles with a fixed spacing. One of them is the central circle of the lens, which is convenient for calculating the lens magnification, angle, and spacing data. If it is a double-row lens, the front and rear row lenses need to be lithographed at a certain distance in the y direction to avoid interference between the front and rear rows. At the same time, circles of different sizes are used for distinction.

[0017] Further, after being lithographed and changed in color, the color-changing film module will gradually recover and can be reused. The color-changing film module is fixed at the end of the precision motion platform. The color-changing film is attached to the transparent glass and has independent light source illumination.

[0018] Further, in step 2, the height difference between the color-changing film module and the precision motion platform in the z-axis direction needs to be measured. If the height of the color-changing film module is greater than the height of the motion platform, the lithography focal plane of the target pattern = the actual lithography focal plane - the height difference; if the height of the color-changing film module is less than the height of the motion platform, the lithography focal plane of the target pattern = the actual lithography focal plane + the height difference, ensuring that the automatic calibration focal plane and the actual lithography focal plane are the same focal plane.

[0019] Further, in step 2, the light source band for lithographing the target pattern on the color-changing film module uses the 405 band, and the lithography energy and time are controlled.

[0020] Further, in step 3, the left and right industrial cameras for alignment identify the central circle of a group of circles corresponding to the same lithography lens, and the identified lithography lens is the one located in the center, ensuring that both the left and right industrial cameras for alignment have a stroke.

[0021] Further, the recognition in steps 3 to 5 is performed using a circular target recognition algorithm. The circular target recognition algorithm is a circle recognition algorithm based on sub-pixel edge detection of improved Zernike moments. The definition formula of the nth-order and mth-degree Zernike moments of the two-dimensional edge continuous function f(x, y) is as follows:

[0022]

[0023] Under discrete conditions, the nth-order and mth-degree Zernike moments of the digital image f(x, y) within the unit circle can be defined as:

[0024]

[0025] Where Z nmrepresents the corresponding Zernike moment of order n and degree m, where both m and n are integers, with n≥0, n - |m| being even, and n≥|m|; x and y represent the digital image coordinate values; represents the complex conjugate of the Zernike polynomial, and ρ, θ are polar coordinates;

[0026] The relationship between the Zernike moments before and after rotation is as follows:

[0027] Z′ nm = Z nm e -imθ

[0028] Z′ nm is the Zernike moment of order n and degree m after rotation, and i is the imaginary unit;

[0029] The straight line within the circle centered at a certain pixel point (x0, y0) in the third-order ideal model of sub-pixel edge detection represents the ideal image edge; the third-order ideal model has three regions, which are pixels with gray values of h, h + k, and h + k1, where the gray value h represents the gray value of the background region, the gray value h + k1 represents the gray value of the transition region, and h + k represents the gray value of the target region; l1 represents the perpendicular distance from the origin to the gray level step position between the background region and the transition region, l2 represents the perpendicular distance from the origin to the gray level step position between the transition region and the target region, θ is the rotation angle of the Zernike moment, and the ideal image edge after rotation is perpendicular to the x-axis;

[0030] The solution is obtained as:

[0031]

[0032] Z′ 20 、Z′ 40 、Z′ 11 、Z′ 31 correspond to the Zernike moments of different orders after rotation respectively;

[0033] l2 takes the maximum value of the following solutions:

[0034]

[0035] Then the perpendicular distance L from the origin to the image edge can be obtained as equal to:

[0036]

[0037] Then the sub-pixel coordinates of the original image can be calculated as:

[0038]

[0039] where N is a constant, x′, y′ are the sub-pixel coordinates of the image, L is the vertical distance from the origin to the edge of the image, Im[Z 11 ]、Re[Z 11 ] represents the real and imaginary parts of the first-order Zernike moment of the digital image.

[0040] The system for implementing the self-calibration method for a large-format printed circuit board digital lithography system includes a lithography lens module, an alignment industrial camera, a color-changing film module, a precision motion platform, and a control module. The lithography lens module and the alignment industrial camera are fixed above the precision motion platform. The color-changing film module is fixed to the end of the precision motion platform. The color-changing film module is attached to transparent glass and has a light source for illumination to improve the accuracy of image recognition.

[0041] The control module includes a material number machine and a main control machine. The material number machine converts the input circuit board image document format file into a material number for management, divides the strips according to the lithography lens position relationship data of the lithography lens module, and sends the divided data information to each lithography lens for lithography; during self-calibration, a circular target pattern is photoetched on the color-changing film of the color-changing film module, and the image is collected by the alignment industrial camera. The position data of the alignment industrial camera in the alignment coordinate system is determined by the circle recognition algorithm, and converted into position data in the lithography coordinate system. The position relationship of the lithography lens is determined, and the position relationship of the lithography lens is compared with the previous reference data. When the allowable error range is exceeded, the strip division is re-performed, and the error compensation is performed to realize the self-calibration of the digital lithography system.

[0042] A computer device of the present invention comprises: a memory and a processor and a computer program stored in the memory. When the computer program is executed on the processor, the self-calibration method for a large-format printed circuit board digital lithography system is implemented.

[0043] Compared with the existing technology, the present invention has the following beneficial effects:

[0044] The self-calibration method for a digital photolithography system for large-format printed circuit boards proposed in the present invention does not require the production of a precise calibration plate. Instead, a reusable color-changing film module is used. The color of the film changes when exposed to a certain amount of light energy, and the color automatically disappears after 5-10 minutes of photolithography. This facilitates automatic calibration of the digital photolithography system without the need for human intervention. After the initial calibration reference data is stored, error compensation can be achieved within a large offset range to maintain the accuracy of the digital photolithography system.

[0045] The self-calibration of the present invention includes three calibration items, namely, the calibration of the CCD position relationship, the calibration of the position relationship of the digital lithography lens module, and the calibration of the offset between the alignment coordinate system and the lithography coordinate system, which almost cover all aspects where offset errors may occur in the digital lithography system, effectively realizing the acquisition and compensation of error data. At the same time, a modular design is adopted, which is convenient for adding other calibration items in the later stage, and further optimizing the self-calibration method for the digital lithography system of large-format printed circuit boards.

[0046] In the circular target recognition image algorithm of the present invention, sub-pixel edge detection based on improved Zernike moments is adopted. Different from the detection algorithm with pixel-level accuracy, it has more detailed edge information, higher accuracy, does not require additional hardware costs, and has appropriate detection efficiency while ensuring accuracy. Different from the traditional second-order ideal gray step model of Zernike moments, considering the gray values of image gradients, a third-order gray transition edge model of the image edge is used for derivation, which has higher accuracy. Brief Description of the Drawings

[0047] In order to more clearly show the technical solution of the present invention, the following can be further described with reference to the drawings in the specification at the end of this document.

[0048] Figure 1 It is a schematic diagram of the overall structure of an embodiment of a self-calibration method for a digital lithography system of large-format printed circuit boards;

[0049] Figure 2 It is a flowchart of an embodiment of a self-calibration method for a digital lithography system of large-format printed circuit boards according to the present invention;

[0050] Figure 3 It is a third-order Zernike moment gray transition edge model for an embodiment;

[0051] Figure 4 It is a front and rear rotation diagram of a third-order Zernike moment gray edge model for an embodiment;

[0052] In the figure: 1 - lithography lens module, 2 - left and right alignment industrial cameras, 3 - color-changing film module, 4 - precision motion platform. Detailed Description of the Preferred Embodiment

[0053] Next, with reference to the drawings in the embodiments of the present invention, a specific and complete description of the present invention will be given. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0054] As Figure 2 shown, an embodiment of a self-calibration method for a digital lithography system of large-format printed circuit boards includes the following steps:

[0055] Step 1: Calibrate to establish the coordinate relationship of the entire digital lithography system, that is, establish the conversion relationship between the lithography coordinate system and the alignment coordinate system. After the system runs stably and the error of manual self-calibration meets the requirements, establish a data benchmark, that is, obtain the position data of the lithography lens module, alignment industrial camera, etc.

[0056] The data benchmark is established after multiple runs of automatic calibration. When the camera's point capture position error meets the specified requirements and stable calibration can be achieved, the data benchmark established by manually running the automatic calibration is carried out, which includes the position data of the lithography lens group, alignment industrial camera, etc., and is used for subsequent data comparison for calibration compensation.

[0057] Preferably, the device that uses automatic calibration for the first time must first perform data backup to avoid data loss. After multiple runs of automatic calibration, if the camera's point capture position error meets the specified requirements and the device is stably calibrated within the allowable error range, manually run the automatic calibration to establish the benchmark as soon as possible to avoid the lithography target on the color-changing film module 3 from dissipating and blurring, which affects the calibration accuracy. After setting the automatic calibration parameters in the calibration items, obtain the position data of the lithography lens group, alignment industrial camera, etc., and use it for subsequent data comparison for calibration compensation.

[0058] Step 2: Perform target pattern lithography on the color-changing film module 3. Each lens lithographs a group of circles. The front and rear rows of the lithography lens module 1 are distinguished by double concentric circles of different sizes, and the appropriate front and rear row distances are adjusted to ensure that the patterns do not overlap.

[0059] As an embodiment, each group of lenses of the lithography lens module 1 lithographs three circles with a fixed spacing, and one is the center circle of the lens, which is convenient for calculating data such as lens magnification, angle, and spacing. If it is a double-row lens, the front and rear row lenses need to be lithographed at a certain distance in the y direction, and at the same time, circles of different sizes are used for distinction to avoid interference between the front and rear rows and facilitate distinction. It is necessary to ensure that the circular targets corresponding to each lithography lens are all on the color-changing film module, and the circular target patterns projected by each lithography lens are clear and there is no overlapping area between them. At the same time, ensure that there is no defocus situation in both the static projection of the lithography lens and the image recognized by the alignment industrial camera. Manually perform self-calibration tests, and judge whether to re-lithograph the circular target according to the color depth of the exposed pattern on the color-changing film module, so as to increase the calibration accuracy.

[0060] Preferably, the color-changing film module used in Step 2 will change color when it receives a certain amount of light energy and will automatically disappear after 5-10 minutes of lithography pattern, and can be reused. As Figure 1 shown, the color-changing film module 3 is fixed at the end of the precision motion platform 4. The color-changing film is attached to the transparent glass and has light source illumination to improve the accuracy of image recognition.

[0061] In Step 2, the height difference between the color-changing film module and the moving platform in the z-axis direction needs to be measured. If the height of the color-changing film module is greater than that of the moving platform, the lithography focal plane of the target pattern = the actual lithography focal plane - the height difference; if the height of the color-changing film module is less than that of the moving platform, the lithography focal plane of the target pattern = the actual lithography focal plane + the height difference. Ensure that the automatically calibrated focal plane and the actual lithography focal plane are the same focal plane.

[0062] Step 3: Use the circular target lithographed on the color-changing film module in Step 2 to calibrate the position relationship of the CCD, that is, to calibrate the position relationship of the two industrial cameras for alignment on the left and right. Use the two industrial cameras for alignment on the left and right 2 to identify the same target dot, and obtain the position data of this dot under the two industrial cameras for subsequent comparison.

[0063] The calibration of the CCD position relationship, that is, to determine the coordinate transformation relationship between the two industrial cameras for alignment on the left and right 2. Use the two industrial cameras for alignment on the left and right to capture the center circle of a group of circles corresponding to the same lithography lens, obtain the corresponding position data, and calculate the coordinate transformation relationship between the two.

[0064] Step 4: Use the circular target lithographed on the color-changing film module in Step 2 to calibrate the position relationship of the lithography lens module. Identify a group of circles corresponding to each lithography lens of the lithography lens module 1 through the industrial camera for alignment, obtain the position data of each dot for calculation, and obtain the magnification, angle, and spacing data of the lithography lens.

[0065] Step 5: Use the circular target lithographed on the color-changing film module in Step 2 to calibrate the offset between the alignment coordinate system and the lithography coordinate system, that is, to calibrate the origin position of the lithography coordinate system. Capture a group of circles lithographed by the first lithography lens in the lithography lens module 1 and obtain its position data.

[0066] Step 6: Compare the position data obtained through Steps 3, 4, and 5 with the position data benchmark established in Step 1. Determine whether to perform compensation based on the error size. Divide the lithography area by stripes, and distribute the compensated stripe data to each lithography lens of the lithography lens module 1 to complete the self-calibration.

[0067] Preferably, the recognition in Steps 3 to 5 is performed using a circular target recognition algorithm. The circular target recognition algorithm is a circular recognition algorithm based on sub-pixel edge detection of improved Zernike moments. The specific method includes the following steps:

[0068] (1) Image preprocessing, reducing the size of the image data, weakening the influence of image noise, and improving the processing efficiency of subsequent algorithms.

[0069] Preferably, the image preprocessing in the circular recognition algorithm specifically includes:

[0070] 1) Grayscale the image, convert the color image to a grayscale image, reduce the size of the image data, and improve the operation speed of subsequent algorithms;

[0071] 2) Extract the region of the picture captured by the industrial camera, that is, crop the edge part of the picture, reduce the data range processed by subsequent algorithms, and improve the algorithm efficiency;

[0072] 3) Perform image filtering on the image to further simplify the data, highlight the edge information of the image, and reduce noise interference.

[0073] (2) Calculate the image obtained after preprocessing, obtain the Zernike moment parameters, and perform sub-pixel edge detection and judgment, and store the obtained sub-pixel point set;

[0074] (3) Use the sub-pixel point set obtained by the improved Zernike moment sub-pixel edge detection to perform random Hough transform, randomly select in the sub-pixel point set, for each group of randomly selected sub-pixel edge points, generate a set of possible center and radius combinations, convert the center coordinates and radius parameters into polar coordinate forms, perform vote accumulation in the parameter space, and take the center and radius combination with the highest number of votes as the detected circle;

[0075] Preferably, the circle recognition algorithm uses the improved Zernike moment sub-pixel edge detection. Different from the traditional Zernike moment sub-pixel coordinate detection based on the ideal image second-order gray level step model, the third-order gray edge transition model is used for derivation. The definition formula of the nth-order and mth-degree Zernike moment of the two-dimensional edge continuous function f(x, y) is as follows:

[0076]

[0077] And under discrete conditions, the nth-order and mth-degree Zernike moment of the digital image f(x, y) within the unit circle can be defined as:

[0078]

[0079] Among them, Z nm represents the corresponding nth-order and mth-degree Zernike moment. In the formula, both m and n are integers, where n≥0, n - |m| is an even number, and n≥|m|; x and y represent the digital image coordinate values; represents the complex conjugate of the Zernike polynomial, and ρ, θ are the polar coordinate expressions.

[0080] The relationship formula of the Zernike moment before and after rotation is as follows:

[0081] Z′ nm =Z nm e -imθ

[0082] Z ′ nm is the rotated Zernike moment of order n and degree m, and i is the imaginary unit.

[0083] The ideal model of the third-order gray-level edge transition for Zernike sub-pixel edge detection is as Figure 3 shown. The straight line inside the circle centered at a certain pixel point (x0, y0) represents the ideal image edge; the ideal model of the third-order gray-level edge transition for Zernike sub-pixel edge detection includes three regions, which are pixels with gray levels of h, h + k, and h + k1, where the gray level h represents the gray level value of the background region, the gray level h + k1 represents the gray level of the transition region, and h + k represents the gray level value of the target region. l1 represents the vertical distance from the origin to the gray-level step position between the background region and the transition region, l2 represents the vertical distance from the origin to the gray-level step position between the transition region and the target region, θ is the rotation angle of the Zernike moment, and the ideal image edge after rotation is perpendicular to the x-axis. Since the modulus value of the Zernike moment has rotational invariance, the operation can be simplified by rotation, as Figure 4 shown.

[0084] The solution is obtained as:

[0085]

[0086] where Z′ 20 、Z′ 40 、Z′ 11 、Z′ 31 correspond to the Zernike moments of different orders after rotation respectively. l2 takes the maximum value of the following solutions:

[0087]

[0088] Then the vertical distance L from the origin to the image edge can be obtained as:

[0089]

[0090] Then the sub-pixel coordinates of the original image can be calculated as:

[0091]

[0092] where N is a positive number. As an example, N is taken as 7. L is the vertical distance from the origin to the image edge, x′, y′ are the sub-pixel coordinates of the image, Im[Z 11 、Re[Z 11 represent the real and imaginary parts of the Zernike moment of order 1 and degree 1 of the digital image.

[0093] (4) Obtain the circular coordinate position data and perform subsequent steps.

[0094] As Figure 1 shown, this embodiment further provides a system for implementing the self-calibration method of a large-format printed circuit board digital lithography system, including a lithography lens module 1, a precision motion platform 4, a color-changing film module 3, a alignment industrial camera 2, and a control module. The lithography lens module 1 and the alignment industrial camera 2 are located above the precision motion platform 4 and fixed at corresponding positions. During lithography and alignment operations, the precision motion platform will move under the relevant equipment. The color-changing film module 3 is fixed at the end of the precision motion platform 4. The color-changing film of the color-changing film module 3 is attached to a transparent glass and has a light source illumination to improve the accuracy of image recognition.

[0095] The control module consists of a material number machine and a main control machine. The material number machine manages the input Gerber (circuit board image document format) and other files by making material numbers. The main control machine is responsible for the control and scheduling of each component of the entire system, divides strips according to data such as the position relationship of the digital lithography lens, and sends the divided data information to each digital lithography lens for lithography. During self-calibration, a circular target pattern is lithographed on the color-changing film, the image is collected by the alignment industrial camera, the position data in the alignment coordinate system is determined through a circle recognition algorithm, converted into the position data in the lithography coordinate system, the position relationship of equipment such as the digital lithography lens is determined, compared with the previous reference data, and when the allowable error range is exceeded, strip division and error compensation are performed again to achieve the self-calibration of the digital lithography system.

[0096] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor limit the invention to the specific embodiments described. Obviously, many modifications and changes can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principle and practical application of the present invention, so that those skilled in the art can well understand and utilize the present invention.

Claims

1. A self-calibration method for a digital lithography system of a large-format printed circuit board, characterized in that, It includes the following steps: Step 1: Calibrate to establish the coordinate relationship of the entire digital lithography system, that is, establish the conversion relationship between the lithography coordinate system and the alignment coordinate system. After the system runs stably and the error of manual self-calibration meets the requirements, establish a data benchmark. Step 2: Perform target pattern lithography on the color-changing film module. Each lens lithographs a group of circles. The front and rear rows are distinguished by double concentric circles of different sizes, and the distance between the front and rear rows is adjusted to ensure that the patterns do not overlap. Step 3: Use the circular target lithographed on the color-changing film module in Step 2 to calibrate the position relationship of the CCD, that is, calibrate the position relationship of the two alignment industrial cameras on the left and right. Use the two cameras to identify the same target dot, and obtain the position data of the dot under the two alignment industrial cameras for subsequent comparison. Step 4: Use the circular target lithographed on the color-changing film module in Step 2 to calibrate the position relationship of the lithography lens module. Identify a group of circles corresponding to each lithography lens through the alignment industrial camera, calculate the position data of each dot, and obtain the magnification, angle, and spacing data of the lithography lens. Step 5: Use the circular target lithographed on the color-changing film module in Step 2 to calibrate the offset between the alignment coordinate system and the lithography coordinate system, that is, calibrate the origin position of the lithography coordinate system. Identify a group of circles lithographed by the first lithography lens and obtain its position data. Step 6: Compare the position data obtained in Steps 3 to 5 with the position data benchmark established in Step 1, perform error compensation, divide the lithography area by stripes, and assign the compensated stripe data to the lithography lens to complete self-calibration.

2. The self-calibration method for a large-format printed circuit board digital lithography system according to claim 1, wherein In Step 1, the data benchmark is established after running the automatic calibration multiple times. When the camera grabbing point position error meets the specified requirements and can be stably calibrated, the data benchmark is established by manually running the automatic calibration. The data benchmark includes the position data of the lithography lens group and the alignment industrial camera, which is used for subsequent data comparison for calibration compensation.

3. The self-calibration method for a large-format printed circuit board digital lithography system according to claim 1, characterized in that In Step 2, when performing target pattern lithography, each lens lithographs three circles with a fixed spacing value. One of them is the center circle of the lens, which is convenient for calculating the magnification, angle, and spacing data of the lens. If it is a double-row lens, the front and rear row lenses need to be lithographed at a certain distance in the y direction to avoid interference between the front and rear rows, and at the same time, circles of different sizes are used for distinction.

4. The self-calibration method for a large-format printed circuit board digital lithography system according to claim 1, wherein, The color-changing film module will gradually recover after being lithography-changed and can be reused; the color-changing film module is fixed at the end of the precision motion platform. The color-changing film is attached to the transparent glass and has independent light source illumination.

5. The self-calibration method for a large-format printed circuit board digital lithography system according to claim 1, characterized in that, In Step 2, it is necessary to measure the height difference between the color-changing film module and the precision motion platform in the z-axis direction. If the height of the color-changing film module is greater than the height of the motion platform, the target pattern lithography focal plane = the actual lithography focal plane - the height difference; if the height of the color-changing film module is less than the height of the motion platform, the target pattern lithography focal plane = the actual lithography focal plane + the height difference, to ensure that the automatic calibration focal plane and the actual lithography focal plane are the same focal plane.

6. The self-calibration method for a large-format printed circuit board digital lithography system according to claim 1, characterized in that, In Step 2, the light source band for target pattern lithography on the color-changing film module is the 405 band, and the lithography energy and time are controlled.

7. A self-calibration method for a large-format printed circuit board digital lithography system according to claim 1, characterized in that, Step 3: The left and right alignment industrial cameras identify the central circle of a set of circles corresponding to the same lithography lens, and the identified lithography lens is the one located at the very center, ensuring that both the left and right alignment industrial cameras have travel.

8. The self-calibration method for a large-format printed circuit board digital lithography system according to claim 1, characterized in that, For the identification in Steps 3 to 5, the circular target recognition algorithm is used for identification. The circular target recognition algorithm is a circle recognition algorithm based on sub-pixel edge detection of improved Zernike moments. The definition formula of the nth-order and mth-degree Zernike moments of the two-dimensional edge continuous function f(x, y) is as follows: Under discrete conditions, the nth-order and mth-degree Zernike moments of the digital image f(x, y) within the unit circle can be defined as: where Z nm represents the corresponding Zernike moment of order n and degree m, where both m and n are integers, with n ≥ 0, n - |m| being even, and n ≥ |m|; x and y represent the digital image coordinate values; represents the complex conjugate of the Zernike polynomial, and ρ, θ are polar coordinates; The relational formula of Zernike moments before and after rotation is as follows: Z ′ nm = Z nm e -imθ Z ′ nm is the rotated Zernike moment of order n and degree m, and i is the imaginary unit; The third-order ideal model of sub-pixel edge detection uses the straight line within a circle centered on a certain pixel point (x0, y0) to represent the ideal image edge; the third-order ideal model has three regions, namely pixels with gray values of h, h + k, and h + k1, where the gray value h represents the gray value of the background region, the gray value h + k1 represents the gray value of the transition region, and h + k represents the gray value of the target region; l1 represents the vertical distance from the origin to the gray level step position between the background region and the transition region, l2 represents the vertical distance from the origin to the gray level step position between the transition region and the target region, θ is the rotation angle of the Zernike moment, and the ideal image edge after rotation is perpendicular to the x-axis; The solution is obtained as: Z2 ′ 0, Z ′ 40 , Z1 ′ 1, Z3 ′ 1 correspond to the Zernike moments of different orders after rotation respectively; l2 takes the maximum value of the following solutions: Then the vertical distance L from the origin to the image edge can be obtained as: Then the sub-pixel point coordinates of the original image can be calculated as: where N is a constant, x ′ , y ′ are the sub-pixel coordinates of the image, L is the perpendicular distance from the origin to the image edge, Im[Z 11 , Re[Z 11 represent the real and imaginary parts of the first-order and first-degree Zernike moments of the digital image.

9. A system for implementing the self-calibration method of the large-format printed circuit board digital lithography system according to claim 1, characterized in that: It includes a lithography lens module (1), an alignment industrial camera (2), a color-changing film module (3), a precision motion platform (4), and a control module; the lithography lens module (1) and the alignment industrial camera (2) are fixed above the precision motion platform (4); the color-changing film module (3) is fixed at the end of the precision motion platform (4), and the color-changing film of the color-changing film module (3) is attached to the transparent glass and has light source illumination to improve the accuracy of image recognition; The control module includes a material number machine and a main control machine. The material number machine makes the input circuit board image document format file into a material number for management, divides it into strips according to the position relationship data of the lithography lenses of the lithography lens module (1), and sends the divided data information to each lithography lens for lithography; during self-calibration, a circular target pattern is lithographed on the color-changing film of the color-changing film module (3), the image is collected by the alignment industrial camera (2), the position data of the alignment industrial camera (2) in the alignment coordinate system is determined through the circle recognition algorithm, converted into the position data in the lithography coordinate system, the position relationship of the lithography lens is determined, and the position relationship of the lithography lens is compared with the previous reference data. When it exceeds the allowable error range, re-strip division and error compensation are performed to achieve self-calibration of the digital lithography system.

10. A computer device, characterized in that, It includes: a memory, a processor, and a computer program stored on the memory. When the computer program is executed on the processor, it implements the self-calibration method for the large-format printed circuit board digital lithography system as described in any one of claims 1 to 7.

Images