Glass substrate positioning method for moving exposure mechanism in edge exposure equipment

Abstract

This invention belongs to the field of photolithography technology and equipment, specifically relating to a glass substrate positioning method for a moving exposure mechanism in an edge exposure device. The method includes: accurately identifying the true edge points on the upper surface of the glass substrate by calculating the asymmetric index of the image grayscale gradient and local standard deviation; establishing an elastic physical model to calculate the vertical sinking of the edge points due to suspension, and correcting the observed coordinates in conjunction with perspective geometry principles to restore their true physical position in an ideal plane; performing weighted fitting on the set of true coordinate points to accurately calculate the rotation and translation deviations of the glass substrate, thereby achieving high-precision and high-robust alignment control.

Description

Technical Field

[0001] This invention relates to the field of photolithography processes and equipment. More specifically, this invention relates to a glass substrate positioning method for a moving exposure mechanism in an edge exposure apparatus. Background Technology

[0002] In the photolithography process of flat panel displays and semiconductor manufacturing, after photoresist is coated onto the glass substrate, a relatively thick layer of photoresist often accumulates at its edges. This layer of photoresist is easily peeled off and forms particles during subsequent etching or cleaning processes, leading to circuit or pixel defects on the surface of the glass substrate and severely affecting product yield. To ensure product quality and yield, edge exposure equipment must be used to precisely expose and remove the photoresist at the edges of the glass substrate. This process requires the moving exposure mechanism to align with the actual physical edge of the glass substrate with unprecedented precision.

[0003] Existing edge exposure equipment mainly relies on machine vision for positioning and calibration. The basic process is usually as follows: the camera acquires local images of the edge of the glass substrate; the edge contour points are extracted by the edge detection algorithm; and the translational and rotational deviations of the glass substrate edge relative to the exposure coordinate system are calculated using geometric fitting technology, so as to adjust the movement trajectory of the exposure head.

[0004] However, in real-world industrial environments, especially when dealing with large-size, ultra-thin glass substrates and complex operating conditions, this traditional positioning method based on two-dimensional geometric assumptions reveals significant technical shortcomings. Specifically, the physical edges of the glass substrate are ground, resulting in a specific chamfered structure. This generates complex multiple reflections and specular highlights in camera imaging, forming multiple parallel false edges that interfere with edge recognition. Traditional edge detection operators struggle to distinguish these false edges caused by chamfer reflections from the true edges of the glass substrate's upper surface. Furthermore, large, ultra-thin glass substrates, when suspended at the edges or with uneven adhesion, are prone to minute but not negligible elastic deflection under gravity. This z-axis deformation in three-dimensional space, under the camera's perspective projection, causes nonlinear distortion of the projected coordinates of the glass substrate edges in the image. Existing positioning algorithms typically treat the glass substrate as a rigid plane, performing only simple two-dimensional affine transformation corrections, ignoring the geometric errors introduced by depth variations. Such errors are unacceptable in edge exposure equipment that demands micron-level exposure accuracy, leading to a systematic deviation between the positioning results and the true physical position of the glass substrate, ultimately affecting the uniformity of exposure. Summary of the Invention

[0005] To address the technical problems of low positioning accuracy and poor robustness in existing technologies for positioning large-size glass substrates, which suffer from neglecting projection distortion caused by non-rigid gravitational deflection and failing to effectively eliminate interference from complex light fields at chamfered corners, this invention provides a glass substrate positioning method for a moving exposure mechanism in an edge exposure device. The method includes: acquiring a grayscale image of the edge region of the glass substrate; obtaining the asymmetry index of a pixel based on the grayscale gradient magnitude and the standard deviation of pixel grayscale values ​​within a local window; identifying local maxima of the asymmetry index to obtain candidate edge points on the upper surface of the glass substrate; obtaining the horizontal distance from the candidate edge point to the edge of the vacuum adsorption area on the plane of the glass substrate based on the coordinates of the edge of the vacuum adsorption area and the coordinates of the candidate edge point; obtaining the vertical sinking amount of the candidate edge point based on the material properties of the glass substrate and the horizontal distance; obtaining the true coordinates of the candidate edge point based on the vertical sinking amount and the observed coordinates; performing a straight line fit on the set of points with the true coordinates to obtain the straight line equation of the glass substrate edge; calculating the rotational and translational deviations of the glass substrate based on the straight line equation; and sending the deviation amount to the motion controller of the moving exposure mechanism.

[0006] This invention acquires grayscale images of the edge region of a glass substrate and calculates an asymmetry index based on the grayscale gradient magnitude and the standard deviation of pixel grayscale values ​​within a local window to identify candidate edge points. It then calculates the vertical drop by combining the material properties of the glass substrate and the horizontal distance from the candidate edge point to the edge of the vacuum adsorption area. The vertical drop is then used to correct the observed coordinates to obtain the true coordinates and fit the edge line equation. This effectively overcomes the false edge interference caused by chamfer reflection and the impact of elastic deflection caused by gravity on the visual positioning accuracy of large-size glass substrates. It ensures that the moving exposure mechanism can perform accurate path following and exposure operations based on the true physical edge after eliminating visual deviations.

[0007] Preferably, the asymmetric exponent satisfies the expression: In the formula, The asymmetry index of a pixel; Represents pixels The grayscale gradient magnitude at that location; It represents the standard deviation of pixel grayscale values ​​within a local window along the positive gradient direction; It represents the standard deviation of pixel grayscale values ​​within a local window along the negative gradient direction; It is a normal number.

[0008] This invention calculates the asymmetry index by multiplying the grayscale gradient magnitude at a pixel by a specific factor that includes the standard deviation of pixel grayscale values ​​within a local window along both the positive and negative gradient directions. Based on the asymmetric light intensity characteristics of a real edge where one side is a smooth surface and the other side is a chamfered reflective area, this method suppresses the false edge response values ​​where the light intensity fluctuations on both sides of the chamfer are large, thus helping edge exposure equipment to accurately distinguish between the real edge on the upper surface of the glass substrate and the multiple false edges generated by the chamfered structure in complex light field environments.

[0009] Preferably, the vertical subsidence satisfies the expression: In the formula, Indicates the vertical sinking amount; This indicates the maximum amount of sinking of the glass substrate at its maximum suspension length. It is the constant of pi; This indicates the horizontal distance of the current candidate edge point from the edge of the vacuum adsorption region; This indicates the maximum span length of the glass substrate design.

[0010] This invention constructs a cosine function model that includes the maximum sinking amount, the constant of pi, the horizontal distance from the current candidate edge point to the edge of the vacuum adsorption zone, and the maximum span length of the glass substrate design to calculate the vertical sinking amount. This model simulates the nonlinear deformation curve of the glass substrate under gravity, similar to a cantilever beam, when the edge is suspended. It provides accurate depth information that conforms to physical facts for subsequent elimination of geometric errors introduced by depth changes, and solves the systematic deviation problem of positioning caused by treating large-size ultra-thin glass substrates as rigid planes.

[0011] Preferably, the actual coordinates satisfy the expression: In the formula, Represents the actual coordinates; Indicates the observed coordinates; Indicates the vertical sinking amount; This represents the tangent of the angle between the camera's line of sight and the normal plane of the glass substrate edge. The angle of incidence for the camera; This represents the vertical distance from the camera's optical center to the reference plane of the glass substrate.

[0012] This invention calculates the true coordinates based on the observation coordinates, vertical sag, the tangent of the angle between the camera's line of sight and the normal plane of the glass substrate edge, and the vertical distance from the camera's optical center to the glass substrate's reference plane. This method uses the principle of geometric optics to decompose the visual deviation caused by sag into the lateral translation caused by oblique observation and the scale scaling error caused by the increase in object distance, and compensates for them respectively. This accurately restores the two-dimensional image observation position affected by gravity sag to the three-dimensional physical true position of the glass substrate in an ideal plane state.

[0013] Preferably, the method for obtaining the maximum subsidence includes: using a high-precision laser displacement sensor to measure the actual maximum subsidence of the glass substrate under the maximum span length condition as the maximum subsidence.

[0014] Preferably, the method for obtaining the horizontal distance includes: determining the pixel coordinates of the edge of the vacuum adsorption area in the grayscale image coordinate system and the pixel equivalent of the camera; calculating the absolute value of the difference between the pixel coordinates of the candidate edge point and the pixel coordinates of the edge of the vacuum adsorption area, and multiplying the absolute value by the pixel equivalent as the horizontal distance from the candidate edge point to the edge of the vacuum adsorption area on the glass substrate plane.

[0015] Preferably, the method for obtaining the vertical distance from the camera optical center to the glass substrate reference plane includes: measuring the vertical distance from the camera optical center to the glass substrate reference plane using a high-precision ranging device.

[0016] Preferably, the method for obtaining the tangent of the angle between the camera's line of sight and the normal plane of the glass substrate edge includes: determining the camera's focal length and principal point coordinates; and calculating the tangent of the incident angle based on perspective geometry, wherein the tangent is equal to the product of the absolute value of the difference between the current pixel's image coordinates and the principal point's coordinates in that direction and the pixel size, divided by the effective focal length.

[0017] Preferably, the step of fitting a straight line to the set of points with true coordinates includes: setting weights based on the asymmetric exponent of the candidate edge points; and performing least squares fitting using the weights.

[0018] This invention sets weights based on the asymmetry index of candidate edge points and performs linear fitting on the point set of real coordinates. This allows high-confidence edge points with stronger light field asymmetry characteristics to occupy a larger proportion when fitting the linear equation, thereby reducing the influence of noise points or potential anomalies when calculating the rotational and translational deviations of the glass substrate. This improves the positioning robustness and fitting accuracy of the moving exposure mechanism for edge lines.

[0019] Preferably, the calculation of the rotational deviation and translational deviation of the glass substrate includes: calculating the arctangent of the slope of the linear equation as the rotational deviation of the glass substrate; and calculating the difference between the intercept of the linear equation and the preset standard exposure position as the translational deviation of the glass substrate.

[0020] The beneficial effects of this invention are as follows:

[0021] This invention utilizes an asymmetric exponent that includes the grayscale gradient magnitude and the standard deviation of pixel grayscale values ​​within a local window to identify candidate edge points. By quantifying the difference in light field distribution between the smooth side of the real edge and the violent fluctuation of light intensity on the chamfered side, it effectively eliminates multiple specular reflections and false edge interference caused by the chamfered grinding structure, ensuring that the edge exposure equipment can accurately lock the physical edge of the upper surface of the glass substrate in complex light field environments.

[0022] This invention constructs an elastic sinking model based on the material properties of the glass substrate and the horizontal distance from the candidate edge point to the edge of the vacuum adsorption area to obtain the vertical sinking amount. This method incorporates the gravitational deflection of the large-size ultrathin glass substrate in the edge-suspended state into the positioning calculation, which solves the systematic positioning error introduced by the prior art that treats the glass substrate as a rigid plane and ignores the depth change. It provides the moving exposure mechanism with glass substrate edge depth information that conforms to the physical reality.

[0023] This invention combines the vertical sag, observation coordinates, and the tangent of the angle between the camera's line of sight and the normal plane of the glass substrate edge to calculate the true coordinates. It uses the principle of geometric optics to compensate for the lateral translation deviation and scaling error caused by the sag of the glass substrate edge, and restores the two-dimensional image data affected by perspective projection distortion to the true physical position on the ideal plane. This guides the moving exposure mechanism to perform uniform and accurate exposure operations based on a high-precision edge trajectory. Attached Figure Description

[0024] Figure 1 This is a flowchart illustrating the glass substrate positioning method for the moving exposure mechanism of the edge exposure equipment in this invention;

[0025] Figure 2 This is a schematic diagram of a glass substrate;

[0026] Figure 3 This is a thermogram schematically illustrating the asymmetry index of a glass substrate;

[0027] Figure 4 This is a schematic diagram illustrating candidate edge points of a glass substrate;

[0028] Figure 5 This is a schematic diagram showing the straight lines at the edge of a glass substrate. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0031] This invention discloses a glass substrate positioning method for a moving exposure mechanism in an edge exposure apparatus, referring to... Figure 1 This includes steps S1 to S4:

[0032] S1. Obtain the grayscale image of the edge region of the glass substrate, and obtain the asymmetry index of the pixel based on the grayscale gradient magnitude and the standard deviation of the pixel grayscale value within the local window; identify the local maxima of the asymmetry index to obtain candidate edge points on the upper surface of the glass substrate.

[0033] It should be noted that in the edge exposure process, the physical edges of the glass substrate undergo chamfering and grinding, forming a complex polyhedral structure. Under illumination, the chamfered area produces multiple specular reflections and forms multiple parallel pseudo-edges, greatly interfering with positioning accuracy. This invention utilizes the essential difference in light field distribution on both sides of the real edge for identification. Because the upper surface of the glass substrate has uniform light reflection characteristics, while the light intensity fluctuates drastically in the chamfered area, this asymmetry in light intensity—stable on one side and drastic on the other—is the core physical characteristic for distinguishing between real and pseudo-edges.

[0034] Specifically, a grayscale image of the glass substrate edge region is acquired, and a one-dimensional line segment-shaped local window perpendicular to the edge direction is defined, dividing the line segment into two equal-sized half-windows. The empirical range for the half-window width is between 5 and 10 pixels. When the detection scene has low resolution, low magnification, or a wide edge light intensity transition area, a larger half-window width should be selected to ensure complete coverage of the edge transition area. When the scene has high resolution, high magnification, or real-time performance is required, a smaller half-window width should be selected to accurately capture local features and reduce computation. If the half-window width exceeds this range, effective edge information may be missed due to an excessively small window, or too much background noise may be introduced due to an excessively large window, compromising locality constraints. In this embodiment, 7 pixels are selected as the half-window width; implementers can choose the appropriate half-window width based on actual conditions.

[0035] Furthermore, for any pixel in the grayscale image, the asymmetry index of that pixel is calculated, and the asymmetry index satisfies the expression:

[0036]

[0037] In the formula, This indicates that the coordinates in the grayscale image are The asymmetry index of the pixels; Represents pixels The grayscale gradient magnitude at that location; It represents the standard deviation of pixel grayscale values ​​within a local window along the positive gradient direction, i.e., pointing towards the chamfered side of the outer edge of the glass substrate; It represents the standard deviation of pixel grayscale values ​​within a local window along the negative gradient direction, i.e., pointing towards the inner plane of the glass substrate. It is a very small positive number, set as .

[0038] In the formula, when the pixel is located at the edge of the actual upper surface, one side of it is a smooth glass substrate surface, resulting in... Extremely small, the other side has a complex chamfered reflective area leading to Larger Approaching 1; when a pixel is located on a pseudo-edge inside a chamfer, the light intensity fluctuations on both sides are large and the values ​​are close. The response value approaches 0, thus significantly suppressing the response value of pseudo-edges.

[0039] Furthermore, the local window moves along the edge, and for the pixel with the largest asymmetry index within all local windows, an adaptive threshold algorithm is used for further filtering, with pixels greater than the adaptive threshold being selected as candidate edge points.

[0040] For example, Figure 2 This is a schematic diagram of a glass substrate.

[0041] For example, Figure 3 It is a thermogram of the asymmetry index of the glass substrate.

[0042] For example, Figure 4 This is a schematic diagram of candidate edge points on a glass substrate.

[0043] S2. Based on the coordinates of the edge of the vacuum adsorption zone and the coordinates of the candidate edge point, obtain the horizontal distance from the candidate edge point to the edge of the vacuum adsorption zone on the plane of the glass substrate; based on the material properties of the glass substrate and the horizontal distance, obtain the vertical sinking amount of the candidate edge point.

[0044] It should be noted that large, ultra-thin glass substrates, when suspended at their edges, will elastically sag downwards due to their own gravity. Existing positioning methods ignore this physical sag, leading to a discrepancy between the visually observed position in the image and the actual physical position. This invention treats the edge of the glass substrate as a cantilever beam structure and calculates its actual vertical drop relative to a reference plane under gravity, providing accurate depth information for subsequent elimination of visual bias.

[0045] Specifically, for each extracted candidate edge point, the vertical sinking amount of the candidate edge point is calculated, and the vertical sinking amount satisfies the expression:

[0046]

[0047] In the formula, The horizontal distance from the boundary of the vacuum adsorption region is... The vertical drop of the candidate edge point relative to the reference plane of the glass substrate; This indicates the maximum amount of sinking of the glass substrate at its maximum suspension length. It is the constant of pi; This indicates the horizontal distance of the current candidate edge point from the edge of the vacuum adsorption region; This indicates the maximum span length of the glass substrate design.

[0048] In the formula, trigonometric functions are used to approximate the deformation curve of the cantilever beam under uniformly distributed load; when the horizontal distance When the value is 0, the vertical sinking is 0, which meets the physical edge condition of no deformation at the boundary of the vacuum adsorption zone; as the horizontal distance increases... With the increase of [missing information], the vertical sinking exhibits a non-linear growth, reflecting the physical phenomenon that the sinking becomes more severe the farther away from the boundary of the vacuum adsorption zone; when the horizontal distance [missing information]... Approaching the maximum span length At that time, the vertical subsidence approaches the maximum subsidence. .

[0049] More specifically, before system deployment, a high-precision laser displacement sensor is used to measure the maximum span length of a glass substrate with known material and thickness. The actual maximum sinking amount under the given conditions is used as the maximum sinking amount for this glass model. Using the pre-calibrated camera parameters, the pixel coordinates of the vacuum adsorption area edge in the grayscale image coordinate system and the pixel equivalent of the camera are determined. The absolute value of the difference between the pixel coordinates of the candidate edge point and the pixel coordinates of the vacuum adsorption area edge is calculated, and this absolute value is multiplied by the pixel equivalent to obtain the horizontal distance from the candidate edge point to the edge of the vacuum adsorption area on the glass substrate plane.

[0050] S3. Obtain the true coordinates of the candidate edge points based on the vertical subsidence and observation coordinates of the candidate edge points.

[0051] It should be noted that the grayscale image captured by the camera is a perspective projection of a three-dimensional object onto a two-dimensional plane. Due to the non-negligible vertical drop of the candidate edge points, this drop will change the imaging distance and viewing angle, resulting in a visual deviation caused by drooping in the image. The purpose of this invention is to use the principles of geometric optics to eliminate this visual deviation caused by drooping and restore the observed coordinates on the image to the true coordinates of the candidate edge points in an ideal plane.

[0052] Specifically, the true coordinates of the candidate edge points are determined, and the true coordinates satisfy the expression:

[0053]

[0054] In the formula, This represents the true coordinates of the candidate edge points to be solved on the reference plane of the glass substrate; This represents the observation coordinates obtained by transforming the image pixel coordinates using a standard pinhole camera model; This indicates the vertical subsidence of the candidate edge point; This represents the tangent of the angle between the camera's line of sight and the normal plane of the glass substrate edge. The angle of incidence for the camera; This represents the vertical distance from the camera's optical center to the reference plane of the glass substrate.

[0055] In the formula, the visual deviation caused by sag is decomposed into two independent geometric components for compensation. The first term uses trigonometric geometry to compensate for the lateral translation of the imaging point on the imaging plane caused by the sinking of the glass substrate edge and the oblique observation of the camera. The second term uses the principle of similar triangles to compensate for the scaling error caused by the increase in the actual distance between the object and the camera due to the sinking of the edge, which leads to a slight change in the magnification of the image. Through this equation, the observed coordinates affected by sag are accurately corrected to the true coordinates.

[0056] It should be further explained that: the system working distance is obtained through geometric system calibration, that is, by measuring the vertical distance from the camera's optical center to the glass substrate reference plane using a high-precision ranging device, and using this distance as a system constant. The incident angle is obtained in real time through camera intrinsic parameter calibration; the camera's focal length and principal point coordinates are determined through the offline Zhang Zhengyou calibration method, and the tangent value of the angle between the camera's line of sight and the normal plane of the glass substrate edge is calculated in real time based on perspective geometry. This tangent value is equal to the product of the absolute value of the difference between the current pixel's image coordinates and the principal point's coordinates in that direction and the pixel size, divided by the effective focal length.

[0057] S4. Perform linear fitting on the point set of real coordinates to obtain the linear equation of the glass substrate edge; calculate the rotational and translational deviations of the glass substrate based on the linear equations, and send the deviation amount to the motion controller of the moving exposure mechanism.

[0058] It should be noted that the present invention has obtained a set of high-precision coordinate points that both eliminate chamfer reflection interference and compensate for the drooping deformation of the glass substrate. These points represent the physical edge position of the glass substrate in an ideal planar state. The present invention uses these high-precision points to calculate the pose of the glass substrate to guide the moving exposure mechanism to perform precise path following.

[0059] Specifically, for each edge of the glass substrate, the weighted least squares method is used to fit a straight line to the set of points with the corrected real coordinates to obtain the straight line equation of the glass substrate edge; the rotational deviation of the glass substrate is calculated based on the slope of the fitted straight line, and the translational deviation is calculated based on the intercept of the straight line and the current position of the exposure head; the calculated rotational and translational deviations are sent to the motion controller of the moving exposure mechanism to adjust its motion trajectory so that it is precisely aligned with the real physical edge of the glass substrate for exposure.

[0060] This completes the glass substrate positioning method for the moving exposure mechanism of the edge exposure equipment.

[0061] For example, Figure 5 This is a schematic diagram of the straight line at the edge of the glass substrate. The selected edge is the right edge of the glass substrate. The red line is the straight line fitted to the edge of the glass substrate, and the blue line is the set of points with corrected real coordinates.

Claims

1. A method for positioning a glass substrate in a moving exposure mechanism of an edge exposure apparatus, characterized in that, include: A grayscale image of the edge region of the glass substrate is acquired. The asymmetry index of the pixel is obtained based on the grayscale gradient magnitude and the standard deviation of the pixel grayscale value within the local window. The local maxima of the asymmetry index are identified to obtain candidate edge points on the upper surface of the glass substrate. The horizontal distance from the edge of the vacuum adsorption region to the edge of the vacuum adsorption region on the glass substrate plane is obtained based on the coordinates of the candidate edge point and the coordinates of the candidate edge point; the vertical sinking amount of the candidate edge point is obtained based on the material properties of the glass substrate and the horizontal distance. The true coordinates of the candidate edge points are obtained based on the vertical subsidence and observed coordinates of the candidate edge points. By fitting a straight line to the set of points with the true coordinates, the straight line equation of the edge of the glass substrate is obtained. The rotational and translational deviations of the glass substrate are calculated based on the linear equation, and the deviations are sent to the motion controller of the moving exposure mechanism.

2. The glass substrate positioning method for a moving exposure mechanism in an edge exposure device according to claim 1, characterized in that, The asymmetric exponent satisfies the expression: ； In the formula, The asymmetry index of a pixel; Represents pixels The grayscale gradient magnitude at that location; It represents the standard deviation of pixel grayscale values ​​within a local window along the positive gradient direction; It represents the standard deviation of pixel grayscale values ​​within a local window along the negative gradient direction; It is a normal number.

3. The glass substrate positioning method for a moving exposure mechanism in an edge exposure device according to claim 1, characterized in that, The vertical subsidence satisfies the expression: ； In the formula, Indicates the vertical sinking amount; This indicates the maximum amount of sinking of the glass substrate at its maximum suspension length. It is the constant of pi; This indicates the horizontal distance of the current candidate edge point from the edge of the vacuum adsorption region; This indicates the maximum span length of the glass substrate design.

4. The glass substrate positioning method for a moving exposure mechanism in an edge exposure device according to claim 1, characterized in that, The actual coordinates satisfy the expression: ； In the formula, Represents the actual coordinates; Indicates the observed coordinates; Indicates the vertical sinking amount; This represents the tangent of the angle between the camera's line of sight and the normal plane of the glass substrate edge. The angle of incidence for the camera; This represents the vertical distance from the camera's optical center to the reference plane of the glass substrate.

5. The glass substrate positioning method for a moving exposure mechanism in an edge exposure device according to claim 3, characterized in that, The methods for obtaining the maximum subsidence include: The maximum actual subsidence of the glass substrate under the maximum span length condition is measured using a high-precision laser displacement sensor and taken as the maximum subsidence.

6. The glass substrate positioning method for a moving exposure mechanism in an edge exposure device according to claim 3, characterized in that, The method for obtaining the horizontal distance includes: Determine the pixel coordinates of the vacuum adsorption area edge in the grayscale image coordinate system and the pixel equivalent of the camera; calculate the absolute value of the difference between the pixel coordinates of the candidate edge point and the pixel coordinates of the vacuum adsorption area edge, and multiply the absolute value by the pixel equivalent to obtain the horizontal distance from the candidate edge point to the vacuum adsorption area edge on the glass substrate plane.

7. The glass substrate positioning method for a moving exposure mechanism in an edge exposure device according to claim 4, characterized in that, The method for obtaining the vertical distance from the camera's optical center to the reference plane of the glass substrate includes: The vertical distance from the optical center of the camera to the reference plane of the glass substrate is measured using a high-precision ranging device.

8. The glass substrate positioning method for a moving exposure mechanism in an edge exposure device according to claim 4, characterized in that, The method for obtaining the tangent of the angle between the camera's line of sight and the normal plane of the glass substrate edge includes: Determine the camera's focal length and principal point coordinates; based on perspective geometry, calculate the tangent of the incident angle, which is equal to the product of the absolute value of the difference between the current pixel's image coordinates and the principal point's coordinates in that direction and the pixel size, divided by the effective focal length.

9. The glass substrate positioning method for a moving exposure mechanism in an edge exposure device according to claim 1, characterized in that, The process of fitting a straight line to the point set of real coordinates includes: Weights are set based on the asymmetric exponent of the candidate edge points; least squares fitting is then performed using these weights.

10. The glass substrate positioning method for a moving exposure mechanism in an edge exposure apparatus according to claim 1, characterized in that, The calculation of the rotational and translational deviations of the glass substrate includes: The arctangent of the slope of the linear equation is calculated as the rotational deviation of the glass substrate; the difference between the intercept of the linear equation and the preset standard exposure position is calculated as the translational deviation of the glass substrate.

Images