A method for detecting defects on the front and back surfaces of ceramic substrates without flipping based on multi-roller transmission

Abstract

The invention discloses a front and back defect detection method for a ceramic substrate without flipping based on multi-roller transmission, which relates to the technical field of ceramic substrate surface defect detection, and comprises the following steps: obtaining the spacing and diameter of rollers used for transmission; selecting and setting a detection position from the gaps between the rollers; the front and back defect detection method for a ceramic substrate without flipping based on multi-roller transmission analyzes the change in the surface height of the ceramic substrate and the up and down fluctuation of the ceramic substrate caused by the contact or separation of the protrusions or depressions at the defective parts of the ceramic substrate with the rollers when the ceramic substrate moves on the multi-rollers by a laser interferometer, so as to determine the surface defects of the ceramic substrate when it is transported by the multi-rollers, so as to eliminate the influence of the up and down fluctuation of the ceramic substrate caused by the contact or separation of the protrusions or depressions at the defective parts of the ceramic substrate with the rollers during the analysis, so that the detection is suitable for multi-roller transmission.

Description

Technical Field

[0001] The invention relates to the technical field of ceramic substrate surface defect detection, and in particular to a method for detecting front and back surface defects of a ceramic substrate without flipping based on multi-roller transmission. Background Art

[0002] Ceramic substrate refers to a special process board in which copper foil is directly bonded to the surface (single or double sides) of alumina (Al2O3) or aluminum nitride (AlN) ceramic substrate at high temperature. The ultra-thin composite substrate has excellent electrical insulation performance, high thermal conductivity, excellent soft solderability and high adhesion strength. It can be etched with various patterns like PCB boards and has a large current carrying capacity. In order to ensure the performance of the ceramic substrate, its surface needs to be smooth without warping, bending, micro cracks, etc.

[0003] The prior art with publication number CN115931898A discloses a method, device and storage medium for visual detection of surface defects of ceramic substrates, wherein the method for visual detection of surface defects of ceramic substrates comprises the following steps: driving a ceramic substrate to be detected to move at a uniform speed by a conveying device, wherein a through hole is provided below the ceramic substrate; determining a stroboscopic timing of a first stroboscopic light source and a second stroboscopic light source connected by a light source controller based on a line frequency of a linear array camera provided above the through hole, wherein the first stroboscopic light source is provided below the through hole, and the second stroboscopic light source is provided obliquely above the through hole; collecting an image of the ceramic substrate by using a linear array camera according to the stroboscopic timing, and obtaining a first image and a second image of the ceramic substrate illuminated by the first stroboscopic light source and the second stroboscopic light source, respectively; selecting the corresponding first image and / or second image for analysis and processing according to the image surface features of different surface defects, and obtaining a detection result of the surface defects.

[0004] However, this detection method is greatly affected by the performance and parameters of the camera, as well as the glossiness of the ceramic substrate surface, and the detection result also depends on image processing. The detection process requires processing and analysis of a large number of images, which increases the workload and difficulty of processing, and reduces the efficiency and accuracy of processing to a certain extent. Summary of the invention

[0005] The object of the present invention is to provide a method for detecting front and back defects of a ceramic substrate without flipping based on multi-roller transmission, so as to solve the above-mentioned deficiencies in the prior art.

[0006] In order to achieve the above object, the present invention provides the following technical solution: a method for detecting front and back defects of a ceramic substrate without flipping based on multi-roller transmission, comprising the following steps:

[0007] Get the spacing and diameter of the rollers used for transmission;

[0008] Select from the gaps of the rollers to set the detection position;

[0009] Laser interferometers are arranged at the detection position and above the detection position, respectively detecting the distance change between the laser interferometer and the surface of the ceramic substrate transmitted from the roller, and obtaining distance change data; the laser emitted by the laser interferometer is perpendicular to the transportation direction of the transportation mechanism composed of the rollers; the laser emission direction of the laser interferometer at the detection position is upward, and the laser interferometer above the detection position and the laser interferometer at the detection position are preferably at a certain distance in the horizontal direction to prevent the lasers of the two laser interferometers from interfering with each other; at the same time, when selecting the laser interferometer, it is necessary to select a laser light path whose width is greater than the width of the ceramic substrate to ensure that the surface of the ceramic substrate can be measured when passing through the laser interferometer;

[0010] Obtain the horizontal distance from the laser interferometer to the roller axis to obtain the detection distance;

[0011] Build defect recognition models and defect height models for each defect type;

[0012] The ceramic substrate is transported by rollers and passes through a detection position to obtain corresponding distance change data, and a distance change curve is obtained through continuous distance change data;

[0013] Inputting the distance variation curve into a defect recognition model for recognition, and outputting a defect label corresponding to the defect type;

[0014] Based on the influence of defects on the moving trajectory of the ceramic substrate, the corresponding defect height model is selected, the distance change data is input for calculation, and the defect height is output. The shape and size of the defect can be expressed by the defect height at each defect location;

[0015] Constructing a standard three-dimensional model of the ceramic substrate, wherein the three-dimensional model of the ceramic substrate can be drawn and constructed using three-dimensional software based on standard parameters of the ceramic substrate, or downloaded and imported from other databases or websites;

[0016] Based on the defect height, the corresponding defect is drawn on the standard three-dimensional model to obtain the actual three-dimensional model corresponding to the ceramic substrate, and the drawn defect is marked on the actual three-dimensional model, which can facilitate users to intuitively observe the position and shape of the defect, while preventing the defect from being too small for users to find the defect position in the actual three-dimensional model.

[0017] Furthermore, the construction of the defect recognition model specifically includes the following steps:

[0018] Selecting a set number of ceramic substrates with various defect types as test substrates, respectively, and obtaining defect parameters of the test substrates, wherein the defect parameters are composed of defect heights at each defect location, and the defect height is a height difference between the defect and the surface of the ceramic substrate;

[0019] Setting defect labels corresponding to defect types and no defects one by one, and establishing an association relationship between the test substrates with the same defect types and the defect labels, for example, the defect labels include crack defect, convex defect, concave defect, missing defect, and no defect (normal);

[0020] The roller is made to uniformly transport the test substrate through the laser interferometer to obtain corresponding distance change data. When the test substrate passes through the laser interferometer, after the laser emitted by the laser interferometer contacts the surface of the test substrate, a part of the laser will be reflected back to the laser interferometer and received by the laser interferometer to obtain a corresponding interference image, and the relative movement distance between the test substrate and the contact position of the laser, i.e., the distance change data, is analyzed based on the interference image;

[0021] Based on the continuous distance change data of each test substrate, a corresponding distance change curve is obtained. If the surface of the test substrate is flat and flawless, the relative movement distance between the surface of the test substrate and the laser contact position is zero when the test substrate passes through the laser interferometer, and the corresponding distance change curve is also a horizontal straight line. If there are defects on the surface of the test substrate, the fluctuation characteristics and amplitude of the corresponding distance change curve will be different according to the type, shape and size of the defects. For example, if there is a crack somewhere on the surface of the test substrate, that is, the crack will be concave relative to the surface of the test substrate, when the crack passes through the test laser of the laser interferometer, the relative movement distance obtained by the test moves to a farther distance (which can be represented by a positive number); if there is a bulge caused by warping or bending on the surface of the test substrate, when the bulge passes through the test laser of the laser interferometer, the relative movement distance obtained by the test moves to a closer distance (which can be represented by a negative number);

[0022] Based on the distance change test curves and defect labels of each test substrate, a deep neural network model is trained to obtain a defect recognition model, which is used to output corresponding defect labels based on the input distance change curve.

[0023] Further, the distance change data includes first distance change data and second distance change data, the first distance change data is the distance change data obtained by the laser interferometer at the detection position, and the second distance change data is the distance change data obtained by the laser interferometer above the detection position;

[0024] The distance change curve includes a first distance change curve and a second distance change curve, wherein the first distance change curve is obtained from continuous first distance change data, and the second distance change curve is obtained from continuous second distance change data;

[0025] The defect label output after the first distance change curve is input into the defect recognition model is the first defect label, and the defect label output after the second distance change defect is input into the defect recognition model is the second defect label.

[0026] Furthermore, building a defect height model for each defect type specifically includes the following steps:

[0027] Construct different defect height models according to whether the defect will affect the moving trajectory of the ceramic substrate on the roller;

[0028] If the moving trajectory of the ceramic substrate is not affected, the defect height at each defect location is directly obtained through the distance change curve, that is, the defect height is equal to the distance change data corresponding to each point of the distance change curve, that is, the output of the defect height model is equal to the corresponding distance change data at this time;

[0029] Because if there are protrusions on the lower surface of the test substrate, or depressions caused by warping or bending, when the protrusion or depression of the test substrate passes through the roller, one side of the test substrate will be relatively lifted or dropped, resulting in a large fluctuation in the distance change curve. It is impossible to analyze the defect parameters only through the distance change curve. It is necessary to combine the spacing and diameter of the rollers and the detection distance for analysis to obtain specific parameters. Therefore, for the moving trajectory of the ceramic substrate, a regression analysis is performed based on the spacing and diameter of the rollers, the detection distance, the defect parameters of the test substrate and the distance change data to obtain a defect height calculation formula, which is used to obtain the defect height corresponding to each distance change data based on the known spacing and diameter of the rollers, the detection distance, and the input distance change data. That is, the defect height model at this time is the defect height calculation formula.

[0030] Furthermore, based on the influence of the defect on the moving trajectory of the ceramic substrate, the corresponding defect height calculation formula is selected, the distance change data is input for calculation, and the defect height is output, which specifically includes the following steps:

[0031] Based on the distance change data, determine whether there are defects that affect the movement trajectory of the ceramic substrate;

[0032] If not, directly obtain the defect height at each defect location based on the distance change curve, specifically obtain the defect height on the lower surface of the ceramic substrate based on the first distance change data, and obtain the defect height on the upper surface of the ceramic substrate based on the second distance change data;

[0033] If so, determining whether there is a defect on the upper surface of the ceramic substrate based on the second defect label;

[0034] Because when there is no defect on the upper surface of the ceramic substrate, the fluctuation of the second distance change data reflects the movement trajectory of the ceramic substrate affected by the contact or separation between the lower surface defect and the roller, and the directions of the first distance change data and the second distance change data detected by the two laser interferometers are opposite. Therefore, when there is no defect on the upper surface of the ceramic substrate, the first distance change curve and the second distance change curve are added after the phase is consistent, so that part of the data reflecting the movement trajectory of the ceramic substrate in the first distance change curve can be offset, and only part of the data reflecting the defect remains. Therefore, if there is no defect on the upper surface of the ceramic substrate, the second defect label is output, and the first distance change curve and the second distance change curve are adjusted to be consistent in phase and then added to obtain the defect height curve of the lower surface of the substrate, wherein, because there may be a certain horizontal distance between the detection position and the laser interferometer above it, the distance change curve obtained during the detection has a phase difference in which the horizontal coordinate is time and the vertical coordinate is the distance change data, and the phase difference between the first distance change curve and the second distance change curve can be made zero by horizontal translation;

[0035] Based on the defect height curve of the lower surface of the substrate, the defect height of the corresponding position of the defect on the lower surface of the ceramic substrate is directly obtained.

[0036] Furthermore, based on the influence of the defect on the moving trajectory of the ceramic substrate, the corresponding defect height calculation formula is selected, the distance change data is input for calculation, and the defect height is output, and the following steps are also included:

[0037] If there is a defect on the upper surface of the ceramic substrate, the first distance change data is input into the defect height calculation formula to obtain the defect height of the lower surface of the ceramic substrate;

[0038] Subtract the corresponding defect height from the distance change data between the first distance change curve and the portion corresponding to the defect to obtain a substrate trajectory curve;

[0039] The second distance variation curve is adjusted to be consistent with the substrate trajectory curve in phase, and then added to obtain a substrate upper surface defect height curve;

[0040] Based on the defect height curve on the substrate surface, the defect height at the corresponding position of the defect on the ceramic substrate surface is directly obtained.

[0041] Furthermore, the method also includes setting up multiple detection positions, and the laser interferometer at each detection position and above the detection position emits lasers at different angles with the moving direction of the ceramic substrate, so as to facilitate the use of the laser interferometer to detect the ceramic substrate from different angles and improve the detection accuracy.

[0042] 1. Compared with the prior art, the method for detecting defects on the front and back surfaces of ceramic substrates without flipping based on multi-roller transmission provided by the present invention performs defect detection on the front and back surfaces of the ceramic substrate respectively by arranging laser interferometers in the gap between the rollers and above the gaps, without flipping the ceramic substrate, thus simplifying the detection steps and improving the detection efficiency.

[0043] 2. Compared with the prior art, the front and back defect detection method of ceramic substrates without flipping based on multi-roller transmission provided by the present invention analyzes the change in the surface height of the ceramic substrate and the up and down fluctuation of the ceramic substrate caused by the contact or separation of the protrusions or depressions at the defects with the rollers when the ceramic substrate moves on the multi-rollers through a laser interferometer to determine the surface defects of the ceramic substrate when it is transported by the multi-rollers, so as to eliminate the influence of the up and down fluctuation of the ceramic substrate caused by the contact or separation of the protrusions or depressions at the defects with the rollers during the analysis, so that the detection is suitable for multi-roller transmission. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0045] Figure 1 A method step diagram provided for an embodiment of the present invention;

[0046] Figure 2 A schematic diagram showing the influence of defects on the moving track of a ceramic substrate provided in an embodiment of the present invention.

[0047] Description of reference numerals:

[0048] 1. Roller; 2. Ceramic substrate; 3. Protrusion defect. DETAILED DESCRIPTION

[0049] In order to enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0050] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0051] In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined. In addition, the terms "installed", "connected" and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal connection of two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0052] Example embodiments will be described more fully below with reference to the accompanying drawings, but the example embodiments may be embodied in different forms and should not be construed as limited to the embodiments set forth herein. On the contrary, the purpose of providing these embodiments is to make the present disclosure thorough and complete and to enable those skilled in the art to fully understand the scope of the present disclosure.

[0053] In the absence of conflict, the various embodiments of the present disclosure and the various features therein may be combined with each other.

[0054] As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0055] The terms used herein are only used to describe specific embodiments and are not intended to limit the present disclosure. As used herein, the singular forms "a", "an" and "the" are also intended to include the plural forms, unless the context clearly indicates otherwise. It will also be understood that when the terms "comprising" and / or "made of" are used in this specification, the presence of the features, wholes, steps, operations, elements and / or components is specified, but the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or groups thereof is not excluded.

[0056] The embodiments described herein may be described with reference to plan views and / or cross-sectional views by means of idealized schematic diagrams of the present disclosure. Therefore, the example illustrations may be modified according to manufacturing techniques and / or tolerances. Therefore, the embodiments are not limited to the embodiments shown in the accompanying drawings, but include modifications of the configurations formed based on the manufacturing process. Therefore, the regions illustrated in the accompanying drawings have schematic properties, and the shapes of the regions shown in the figures illustrate the specific shapes of the regions of the elements, but are not intended to be limiting.

[0057] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art and the present disclosure, and will not be interpreted as having an idealized or overly formal meaning unless explicitly defined as such herein.

[0058] See also Figure 1-Figure 2 , a method for detecting front and back defects of a ceramic substrate without flipping based on multi-roller transmission, comprising the following steps:

[0059] S1. Obtain the spacing and diameter of the rollers used for transmission.

[0060] S2. Select from the gap between the rollers and set the detection position. When setting the detection position, multiple detection positions can be set. The laser emitted by each detection position and the laser interferometer above the detection position has a different angle with the moving direction of the ceramic substrate, which is convenient for using the laser interferometer to detect the ceramic substrate from different angles, thereby improving the detection accuracy.

[0061] S3. Laser interferometers are arranged at the detection position and above the detection position to respectively detect the distance change between the laser interferometer and the surface of the ceramic substrate transmitted from the roller to obtain distance change data; the laser emitted by the laser interferometer is perpendicular to the transportation direction of the transportation mechanism composed of the rollers; the laser emission direction of the laser interferometer at the detection position is upward, and the laser interferometer above the detection position and the laser interferometer at the detection position are preferably at a certain distance in the horizontal direction to prevent the lasers of the two laser interferometers from interfering with each other; at the same time, when selecting the laser interferometer, it is necessary to select a laser light path whose width is greater than the width of the ceramic substrate to ensure that the surface of the ceramic substrate can be measured when passing through the laser interferometer;

[0062] The distance change data includes first distance change data and second distance change data, the first distance change data is the distance change data obtained by the laser interferometer at the detection position, and the second distance change data is the distance change data obtained by the laser interferometer above the detection position;

[0063] The distance change curve includes a first distance change curve and a second distance change curve, the first distance change curve is obtained by continuous first distance change data, and the second distance change curve is obtained by continuous second distance change data;

[0064] The defect label output after the first distance change curve is input into the defect recognition model is the first defect label, and the defect label output after the second distance change defect is input into the defect recognition model is the second defect label.

[0065] S4. Obtain the horizontal distance from the laser interferometer to the roller axis to obtain the detection distance.

[0066] S5. Build a defect recognition model and a defect height model for each defect type;

[0067] Among them, building a defect recognition model specifically includes the following steps:

[0068] S5.1.1. Select a set number of ceramic substrates with various defect types as test substrates, and obtain defect parameters of the test substrates, where the defect parameters are composed of defect heights at each defect location, and the defect height is the height difference between the defect and the surface of the ceramic substrate;

[0069] S5.1.2. Defect labels corresponding to defect types and no defects are set, and the test substrates with the same defect types are associated with the defect labels. For example, the defect labels include crack defect, convex defect, concave defect, missing defect, and no defect (normal).

[0070] S5.1.3. The roller is made to uniformly transport the test substrate through the laser interferometer to obtain corresponding distance change data. When the test substrate passes through the laser interferometer, after the laser emitted by the laser interferometer contacts the surface of the test substrate, a part of the laser will be reflected back to the laser interferometer and received by the laser interferometer to obtain the corresponding interference image. Based on the interference image, the relative movement distance between the test substrate and the contact position of the laser, i.e., the distance change data, is analyzed;

[0071] S5.1.4. Based on the continuous distance change data of each test substrate, a corresponding distance change curve is obtained. If the surface of the test substrate is flat and flawless, the relative movement distance between the surface of the test substrate and the laser contact position is zero when the test substrate passes through the laser interferometer, and the corresponding distance change curve is also a horizontal straight line. If there are defects on the surface of the test substrate, the fluctuation characteristics and amplitude of the corresponding distance change curve will be different depending on the type, shape and size of the defects. For example, if there is a crack somewhere on the surface of the test substrate, that is, the crack will be concave relative to the surface of the test substrate, when the crack passes through the test laser of the laser interferometer, the relative movement distance obtained by the test moves to a farther distance (which can be represented by a positive number); if there is a bulge caused by warping or bending on the surface of the test substrate, when the bulge passes through the test laser of the laser interferometer, the relative movement distance obtained by the test moves to a closer distance (which can be represented by a negative number);

[0072] S5.1.5. Based on the distance change test curve and defect label of each test substrate, a deep neural network model is trained to obtain a defect recognition model, which is used to output the corresponding defect label based on the input distance change curve.

[0073] Building a defect height model for each defect type includes the following steps:

[0074] S5.2.1. Construct different defect height models according to whether the defect will affect the moving trajectory of the ceramic substrate on the roller;

[0075] S5.2.2. If there is no influence on the moving trajectory of the ceramic substrate, the defect height at each defect location is directly obtained through the distance change curve, that is, the defect height is equal to the distance change data corresponding to each point of the distance change curve, that is, the output of the defect height model is equal to the corresponding distance change data;

[0076] S5.2.3. If there are protrusions on the lower surface of the test substrate, or depressions caused by warping or bending, when the protrusion or depression of the test substrate passes through the roller, one side of the test substrate will be relatively lifted or dropped, resulting in a large fluctuation in the distance change curve. It is impossible to analyze the defect parameters only through the distance change curve. It is necessary to analyze the spacing and diameter of the rollers and the detection distance to obtain specific parameters. Therefore, for the moving trajectory of the ceramic substrate, a regression analysis is performed based on the spacing and diameter of the rollers, the detection distance, the defect parameters of the test substrate and the distance change data to obtain a defect height calculation formula, which is used to obtain the defect height corresponding to each distance change data based on the known spacing and diameter of the rollers, the detection distance, and the input distance change data. That is, the defect height model at this time is the defect height calculation formula.

[0077] S6, transporting the ceramic substrate by rollers and passing through the detection position to obtain corresponding distance change data, and obtaining a distance change curve through continuous distance change data;

[0078] S7, inputting the distance change curve into the defect recognition model for recognition, and outputting a defect label corresponding to the defect type;

[0079] S8. Based on the influence of the defect on the moving trajectory of the ceramic substrate, a corresponding defect height model is selected, the distance change data is input for calculation, and the defect height is output. The shape and size of the defect can be expressed by the defect height at each defect location, which specifically includes the following steps:

[0080] S8.1. Determine whether there is a defect based on the distance change data, which affects the moving trajectory of the ceramic substrate;

[0081] S8.2, if not, directly obtain the defect height at each defect location based on the distance change curve, specifically obtain the defect height on the lower surface of the ceramic substrate based on the first distance change data, and obtain the defect height on the upper surface of the ceramic substrate based on the second distance change data;

[0082] S8.3. If yes, based on the second defect label, determine whether there is a defect on the upper surface of the ceramic substrate;

[0083] S8.4. Because when there is no defect on the upper surface of the ceramic substrate, the fluctuation of the second distance change data reflects the movement trajectory of the ceramic substrate affected by the contact or separation between the lower surface defect and the roller, and the directions of the first distance change data and the second distance change data detected by the two laser interferometers are opposite. Therefore, when there is no defect on the upper surface of the ceramic substrate, the first distance change curve and the second distance change curve are added after being phase-matched, so that part of the data reflecting the movement trajectory of the ceramic substrate in the first distance change curve can be offset, and only part of the data reflecting the defect remains. Therefore, if there is no defect on the upper surface of the ceramic substrate, the second defect label is output, and the first distance change curve and the second distance change curve are adjusted to be consistent in phase and then added to obtain the defect height curve of the lower surface of the substrate, wherein, because there may be a certain horizontal distance between the detection position and the laser interferometer above it, the distance change curve obtained during the detection has a phase difference in which the horizontal coordinate is time and the vertical coordinate is the distance change data, and the phase difference between the first distance change curve and the second distance change curve can be made zero by horizontal translation;

[0084] S8.5, based on the defect height curve of the lower surface of the substrate, directly obtain the defect height of the corresponding position of the defect on the lower surface of the ceramic substrate;

[0085] S8.6. If there is a defect on the upper surface of the ceramic substrate, input the first distance change data into a defect height calculation formula to obtain the defect height on the lower surface of the ceramic substrate;

[0086] S8.7, subtracting the corresponding defect height from the distance change data between the first distance change curve and the corresponding part of the defect to obtain a substrate trajectory curve;

[0087] S8.8, adjusting the phases of the second distance variation curve and the substrate trajectory curve to be consistent, and then adding them together to obtain a surface defect height curve of the substrate;

[0088] S8.9. Based on the surface defect height curve of the substrate, the defect height at the corresponding position of the surface defect on the ceramic substrate is directly obtained.

[0089] S9, constructing a standard three-dimensional model of a ceramic substrate, wherein the three-dimensional model of the ceramic substrate can be drawn and constructed using three-dimensional software based on standard parameters of the ceramic substrate, or downloaded and imported from other databases or websites;

[0090] S10. Based on the defect height, the corresponding defect is drawn on the standard three-dimensional model to obtain the actual three-dimensional model corresponding to the ceramic substrate, and the drawn defect is marked on the actual three-dimensional model, so that the user can intuitively observe the position and shape of the defect, and at the same time prevent the defect from being too small for the user to find the defect position in the actual three-dimensional model.

[0091] The above only describes some exemplary embodiments of the present invention by way of illustration. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the present invention.

Claims

1. A method for detecting defects on the front and back surfaces of ceramic substrates without flipping based on multi-roller transmission, characterized in that: The following steps are involved: Get the spacing and diameter of the rollers used for transmission; Select from the gaps of the rollers to set the detection position; A laser interferometer is arranged at the detection position and above the detection position, for respectively detecting the distance change between the laser interferometer and the surface of the ceramic substrate transmitted from the roller, and obtaining distance change data; Obtain the horizontal distance from the laser interferometer to the roller axis to obtain the detection distance; Building a defect recognition model and defect height models for each defect type, wherein the defect types include crack defects, convex defects, concave defects, missing defects and no defects; The ceramic substrate is transported by rollers and passes through a detection position to obtain corresponding distance change data, and a distance change curve is obtained through continuous distance change data; Inputting the distance variation curve into a defect recognition model for recognition, and outputting a defect label corresponding to the defect type; Based on the influence of defects on the moving trajectory of the ceramic substrate, the corresponding defect height model is selected, the distance change data is input for calculation, and the defect height is output; constructing a standard three-dimensional model of the ceramic substrate; Based on the defect height, a corresponding defect is drawn on the standard three-dimensional model to obtain an actual three-dimensional model corresponding to the ceramic substrate, and the drawn defect is marked on the actual three-dimensional model; The construction of the defect recognition model specifically includes the following steps: Selecting a set number of ceramic substrates with various defect types as test substrates, respectively, and obtaining defect parameters of the test substrates, wherein the defect parameters are composed of defect heights at each defect location, and the defect height is a height difference between the defect and the surface of the ceramic substrate; Setting defect labels corresponding to defect types one by one, and establishing an association relationship between the test substrates with the same defect types and the defect labels; Allow the roller to uniformly transport the test substrate through the laser interferometer to obtain corresponding distance change data; Based on the continuous distance change data of each test substrate, a corresponding distance change curve is obtained; Based on the distance change test curve and defect label of each test substrate, a deep neural network model is trained to obtain a defect recognition model, which is used to output the corresponding defect label based on the input distance change curve; The distance change data includes first distance change data and second distance change data, the first distance change data is distance change data obtained by a laser interferometer at a detection position, and the second distance change data is distance change data obtained by a laser interferometer above the detection position; The distance change curve includes a first distance change curve and a second distance change curve, wherein the first distance change curve is obtained from continuous first distance change data, and the second distance change curve is obtained from continuous second distance change data; The defect label output after the first distance change curve is input into the defect recognition model is the first defect label, and the defect label output after the second distance change defect is input into the defect recognition model is the second defect label; Building a defect height model for each defect type specifically includes the following steps: Construct different defect height models according to whether the defect will affect the moving trajectory of the ceramic substrate on the roller; The moving trajectory of the ceramic substrate is not affected, and the defect height at each defect location is directly obtained through the distance change curve; For the moving trajectory of the ceramic substrate, regression analysis is performed based on the spacing and diameter of the rollers, the detection distance, the defect parameters of the test substrate and the distance change data to obtain the defect height calculation formula, which is used to obtain the defect height corresponding to each distance change data based on the known spacing and diameter of the rollers, the detection distance, and the input distance change data; Based on the influence of the defect on the moving trajectory of the ceramic substrate, the corresponding defect height calculation formula is selected, the distance change data is input for calculation, and the defect height is output, which specifically includes the following steps: Based on the distance change data, determine whether there are defects that affect the movement trajectory of the ceramic substrate; If not, the defect height at each defect location is directly obtained based on the distance variation curve; If so, determining whether there is a defect on the upper surface of the ceramic substrate based on the second defect label; If there is no defect on the upper surface of the ceramic substrate, a second defect label is output, and the first distance change curve and the second distance change curve are adjusted in phase and then added to obtain a defect height curve of the lower surface of the substrate; Based on the defect height curve of the lower surface of the substrate, the defect height of the corresponding position of the defect on the lower surface of the ceramic substrate is directly obtained; The method further includes the following steps: selecting a corresponding defect height calculation formula based on the influence of the defect on the moving trajectory of the ceramic substrate, inputting distance change data for calculation, and outputting the defect height: If there is a defect on the upper surface of the ceramic substrate, the first distance change data is input into the defect height calculation formula to obtain the defect height of the lower surface of the ceramic substrate; Subtract the corresponding defect height from the distance change data between the first distance change curve and the portion corresponding to the defect to obtain a substrate trajectory curve; The second distance variation curve is adjusted to be consistent with the substrate trajectory curve in phase, and then added to obtain a substrate upper surface defect height curve; Based on the defect height curve on the substrate surface, the defect height at the corresponding position of the defect on the ceramic substrate surface is directly obtained.

2. The method for detecting front and back defects of ceramic substrates without flipping based on multi-roller transmission according to claim 1 is characterized in that: The method further comprises setting a plurality of detection positions, wherein the laser emitted by each detection position and the laser interferometer above the detection position has a different angle with the moving direction of the ceramic substrate.

Images