High-precision measuring device and method suitable for PECVD (Plasma Enhanced Chemical Vapor Deposition) metal substrate position variation

Abstract

The invention discloses a high-precision measuring device and method suitable for PECVD (Plasma Enhanced Chemical Vapor Deposition) metal substrate position variation. The high-precision measuring device comprises a shell, a development board, a power supply module, a light source module, a linear polarizer, a guide rail, a double-telecentric lens and an industrial camera, the industrial camera and the light source module are arranged in parallel, face the side face of the machine shell and are installed on the guide rail. The double telecentric lens is mounted on the industrial camera; the linear polarizer covers the front of the double telecentric lens and the light source module; the power supply module and the development board are respectively arranged at the rear sides of the light source module and the industrial camera; the guide rail is composed of two vertical rods and a horizontal rod, the horizontal rod and the vertical rods are fixed through screws, and the height of the horizontal rod is adjusted by fixing screw holes with different heights. The industrial camera, the light source module and the development board are connected with the power supply module through power lines; the industrial camera is connected with the development board through a data line. The method overcomes the defects that an existing measurement method based on computer imaging is high in algorithm complexity and large in computing resource demand.

Description

Technical Field

[0001] The present invention belongs to the technical field of computer vision, and in particular relates to a high-precision measurement device and method suitable for the position variation of a PECVD metal substrate. Background Art

[0002] PECVD stands for Plasma Enhanced Chemical Vapor Deposition, which is a highly regarded thin film preparation technology in the field of materials science. This technology uses the high activity and high energy state of plasma to provide a driving force for chemical reactions, and can accurately and efficiently deposit thin films with excellent properties on substrates.

[0003] During the PECVD process, the precise position of the substrate is of great significance to ensure uniform deposition of the film and strong adhesion between the film and the substrate. Once the substrate is offset, it will not only destroy the uniformity of the film, resulting in uneven distribution of film thickness and composition on the substrate surface, affecting its overall performance, but also weaken the adhesion between the film and the substrate, making the film easy to fall off or crack, thereby affecting the stability and reliability of the film.

[0004] There are three main methods for detecting PECVD substrate offset: laser interferometry, capacitance sensor, and computer vision. The laser interferometry method determines the offset by measuring the change in interference fringes generated after the laser is reflected on the substrate surface. Since it can only be installed inside the PECVD equipment, it is rarely used in actual production. The capacitance sensor method detects the change in the distance between the substrate and the sensor based on the change in capacitance to determine the offset, but it has high requirements for the sensitivity and stability of the sensor and will be affected by electromagnetic interference. The method based on computer vision measures from the outside of the equipment through the side observation window of the PECVD equipment, which does not affect the normal production and use of the equipment, and has the advantages of low equipment cost, high sensitivity, and no electromagnetic interference, and has good development prospects. However, this detection method based on computer vision is still in the preliminary research stage. The existing method has the disadvantages of high algorithm complexity, large computing resource requirements, and is easily affected by ambient light, metal substrates, and glare on the side observation window of the PECVD equipment. Summary of the invention

[0005] In order to overcome the shortcomings of the prior art, the present invention provides a high-precision measurement device and method for the position change of a PECVD metal substrate, including a housing, a development board, a power supply module, a light source module, a linear polarizer, a guide rail, a bi-telecentric lens and an industrial camera; the industrial camera is placed parallel to the light source module, facing the side of the housing, and is installed on the guide rail; the bi-telecentric lens is installed on the industrial camera; the linear polarizer covers the bi-telecentric lens and the light source module; the power supply module and the development board are placed on the back side of the light source module and the industrial camera respectively; the guide rail consists of two vertical rods and a horizontal rod, the horizontal rod and the vertical rod are fixed by screws, and the height of the horizontal rod is adjusted by fixing screw holes of different heights; the industrial camera, the light source module, and the development board are connected to the power supply module through a power cord; the industrial camera is connected to the development board through a data cable. The present invention overcomes the shortcomings of the existing measurement method based on computer imaging, which has high algorithm complexity, large computing resource requirements, and is easily affected by ambient light, metal substrates, and glare on the surface of the side observation window of the PECVD equipment.

[0006] The technical solution adopted by the present invention to solve the technical problem is as follows:

[0007] A high-precision measurement device for the position variation of a PECVD metal substrate, comprising a housing, a development board, a power supply module, a light source module, a linear polarizer, a guide rail, a double telecentric lens and an industrial camera;

[0008] The industrial camera is placed parallel to the light source module, facing the side of the housing, and installed on the guide rail;

[0009] The bi-telecentric lens is installed on the industrial camera;

[0010] The linear polarizer covers the bi-telecentric lens and the light source module;

[0011] The power supply module and the development board are placed on the rear side of the light source module and the industrial camera respectively;

[0012] The guide rail is composed of two vertical rods and a horizontal rod, the horizontal rod and the vertical rod are fixed by screws, and the height of the horizontal rod is adjusted by fixing screw holes at different heights;

[0013] The industrial camera, light source module and development board are connected to the power supply module via a power cord;

[0014] The industrial camera is connected to the development board via a data cable.

[0015] Preferably, the industrial camera is connected to the bi-telecentric lens via a screw hole.

[0016] Preferably, the guide rail is fixed to the bottom of the casing, and the industrial camera and light source module are fixed to the guide rail by screws.

[0017] Preferably, there are multiple screw holes on the guide rail, and the industrial camera and the light source module are fixed by screws according to the size of the detection window on the side of the PECVD equipment to ensure that the camera field of view and the light source module can cover the detection window.

[0018] A high-precision measurement method for the position variation of a PECVD metal substrate comprises the following steps:

[0019] Step 1: Install the bi-telecentric lens according to the size of the side inspection window of the PECVD equipment, adjust the industrial camera and light source module to the appropriate position, and fix them with screws; fix the linear polarizer in front of the light source module, and rotate the angle of the linear polarizer in front of the bi-telecentric lens to ensure that the captured image is not interfered by ambient light, metal substrate and glare on the surface of the side observation window of the PECVD equipment;

[0020] Step 2: Take the image A0 of the initial frame, define a detection area in the image, cut out the image in the area and record it as B0;

[0021] Step 3: Perform Gaussian filtering on image B0, binarize the filtered image, and generate a minimum rectangular detection frame containing all pixel points with a pixel value of 255 in the binary image; extract the ordinate values ​​of the upper and lower boundaries of the rectangular detection frame, and calculate the ordinate of the geometric center of the rectangular frame, and calculate the corresponding value of the size of a single pixel point and the real world according to the camera pixel size and the magnification of the bi-telecentric lens;

[0022] Step 4: Establish a mathematical model in the image coordinate system where y up ,y low Represents the vertical coordinate value of the upper and lower boundaries of the rectangular detection box, W B0 represents the width of the rectangular area, D represents the thickness of the metal substrate, and θ represents the inclination angle of the metal substrate relative to the horizontal plane; the inclination angle of the metal substrate relative to the horizontal plane is obtained by the model calculation;

[0023] Step 5: Use the same method as the initial frame image to capture the i-th frame image A i , and extract the detection area B i ; For image B i Gaussian filtering and binarization are performed to generate a minimum rectangular detection frame containing all pixel points with a pixel value of 255 in the binarized image, the ordinate values ​​of the upper and lower boundaries of the rectangular detection frame are extracted, and the ordinate of the geometric center of the rectangular frame is calculated, and the inclination angle of the metal substrate relative to the horizontal plane is obtained using a mathematical model;

[0024] Step 6: Calculate the pixel change of the i-th frame relative to the initial frame, convert the pixel points into real distances using the size corresponding values, and finally obtain the displacement change; calculate the angle change of the metal substrate of the i-th frame relative to the horizontal plane, and perform a difference operation with the angle change of the initial frame relative to the horizontal plane to obtain the change in inclination angle.

[0025] Furthermore, the step 2 is specifically as follows: capture the image A0 of the initial frame, define a detection area in the image, the upper left corner vertex coordinates of the area are (x1, y1), the lower right corner vertex coordinates are (x2, y2), and the image in the area is cut out and recorded as B0.

[0026] Furthermore, the step 3 is specifically as follows:

[0027] Step 3-1: Perform Gaussian filtering on image B0, binarize the filtered image using Otsu's method, and generate a rectangle in the binarized image. The rectangle is defined as the minimum rectangular detection frame that contains all pixels with a pixel value of 255;

[0028] Step 3-2: Extract the ordinate value y of the upper boundary of the rectangular detection box up0 , the ordinate value y of the lower boundary low0 , calculate the ordinate of the geometric center of the rectangle as According to the camera pixel size u and the bi-telecentric lens magnification A, the corresponding value α of the size of a single pixel and the real world is calculated.

[0029] Furthermore, the step 4 is specifically as follows: establishing a mathematical model where y up ,y low Represents the vertical coordinate value of the upper and lower boundaries of the rectangular detection box, W B0 represents the width of the rectangular area, D represents the thickness of the metal substrate, and θ represents the inclination angle of the metal substrate relative to the horizontal plane; the variable y low ,y up , W B0 , D are substituted into the mathematical model to obtain the inclination angle θ0 of the metal substrate relative to the horizontal plane.

[0030] Furthermore, the step 5 is specifically as follows:

[0031] Step 5-1: Use the same method as the initial frame image to capture the i-th frame image A i , and extract the detection area B i ;

[0032] Step 5-2: Image B i Perform Gaussian filtering and binarization to generate a minimum rectangular detection box containing all pixels with a pixel value of 255 in the binary image;

[0033] Step 5-3: Extract the ordinate value y of the upper boundary of the rectangular detection box upi , the ordinate value y of the lower boundary lowi , then the ordinate of the geometric center of the rectangle is Using mathematical models Calculate the inclination angle θ of the metal substrate relative to the horizontal plane i .

[0034] Furthermore, the step 6 is specifically as follows:

[0035] Step 6-1: Calculate the pixel change △y of the vertical coordinate of the i-th frame relative to the initial frame pixels =y Pi -y P0 , use the size corresponding value α to convert the pixel point into the real distance, and finally get the displacement change △y real =α·△y pixels ;

[0036] Step 6-2: The angle change θ of the metal substrate of the i-th frame relative to the horizontal plane i , and perform a difference operation with the angle change θ0 of the initial frame relative to the horizontal plane to obtain the change in inclination △θ.

[0037] The beneficial effects of the present invention are as follows:

[0038] The method of the present invention can be applied to the measurement system of the PECVD substrate offset, has a certain robustness to ambient light and target surface glare, can greatly improve the accuracy of metal substrate offset detection, overcomes the problem that the laser interferometer method and the capacitive sensor method affect the normal production of the equipment and are susceptible to electromagnetic interference, and overcomes the problem that the computer vision method has large computing resource requirements and is susceptible to the influence of ambient light, the metal substrate and the glare of the side observation window of the PECVD equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic diagram of the measuring device of the present invention;

[0040] Figure 2 The initial frame and the captured image thereof according to the embodiment of the present invention;

[0041] Figure 3 An embodiment of the present invention generates an image of a detection frame on a binary image of an initial frame;

[0042] Figure 4 A geometrical schematic diagram of a mathematical model of an embodiment of the present invention;

[0043] Figure 5 The i-th frame and its captured image in the embodiment of the present invention;

[0044] Figure 6Generate an image of a detection frame on the binary image of the i-th frame in the embodiment of the present invention. ;

[0045] In the picture: 1 housing, 2 development board, 3 power supply module, 4 light source module, 5 linear polarizer, 6 guide rail, 7 bi-telecentric lens, 8 industrial camera. DETAILED DESCRIPTION

[0046] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0047] The purpose of the present invention is to provide a high-precision measurement device and method suitable for the position change of a PECVD metal substrate, so as to overcome the shortcomings of the existing measurement method based on computer imaging, such as high algorithm complexity, large computing resource requirements, and susceptibility to the influence of ambient light, glare on the surface of the metal substrate and the side observation window of the PECVD equipment.

[0048] The present invention adopts the following technical solutions:

[0049] A high-precision measurement method and device suitable for the position variation of a PECVD metal substrate, comprising a housing, a development board, a power supply module, a light source module, a linear polarizer, a guide rail, a double telecentric lens, and an industrial camera.

[0050] A high-precision measurement method for the position variation of a PECVD metal substrate comprises the following steps:

[0051] S1. Install a suitable bi-telecentric lens according to the size of the side detection window of the PECVD equipment, adjust the industrial camera and light source module to the appropriate position, and fix them with screws; fix the linear polarizer in front of the light source module, and rotate the linear polarizer in front of the bi-telecentric lens to a suitable angle to ensure that the captured image is not disturbed by ambient light, metal substrate and glare on the surface of the side observation window of the PECVD equipment;

[0052] S2. Take an image A0 of the initial frame, define a detection area in the image, cut out the image in the area and record it as B0.

[0053] Take the image A0 of the initial frame, define a detection area in the image, the coordinates of the upper left corner vertex of the area are (x1, y1), the coordinates of the lower right corner vertex are (x2, y2), and the image in the area is cut out and recorded as B0.

[0054] S3, perform Gaussian filtering on the image B0, binarize the filtered image, and generate a minimum rectangular detection frame in the binarized image that contains all pixel points with a pixel value of 255. Extract the ordinate values ​​of the upper and lower boundaries of the rectangular detection frame, calculate the ordinate of the geometric center of the rectangular frame, and calculate the corresponding value of the size of a single pixel point and the real world according to the camera pixel size and the bi-telecentric lens magnification.

[0055] Perform Gaussian filtering on image B0, binarize the filtered image using Otsu's method, and generate a rectangle in the binarized image. The rectangle is defined as the minimum rectangular detection box containing all pixels with a pixel value of 255. Extract the ordinate value y of the upper boundary of the rectangular detection box up0 , the ordinate value y of the lower boundary low0 , calculate the ordinate of the geometric center of the rectangle as According to the camera pixel size u and the bi-telecentric lens magnification A, the corresponding value α of the size of a single pixel and the real world is calculated.

[0056] S4. Establishing a mathematical model in the image coordinate system where y up ,y low Represents the vertical coordinate value of the upper and lower boundaries of the rectangular detection box, W B0 represents the width of the rectangular area, D represents the thickness of the metal substrate, and θ represents the inclination angle of the metal substrate relative to the horizontal plane. The inclination angle of the metal substrate relative to the horizontal plane is obtained by model calculation.

[0057] Building a mathematical model where y up ,y low Represents the vertical coordinate value of the upper and lower boundaries of the rectangular detection box, W B0 represents the width of the rectangular area, D represents the thickness of the metal substrate, and θ represents the inclination angle of the metal substrate relative to the horizontal plane. low ,y up , W B0 , D are substituted into the mathematical model to obtain the inclination angle θ0 of the metal substrate relative to the horizontal plane.

[0058] S5, take the i-th frame image A using the same method as that used to take the initial frame image i , and extract the detection area B i For image B i Gaussian filtering and binarization are performed to generate a minimum rectangular detection frame containing all pixel points with a pixel value of 255 in the binarized image, the ordinate values ​​of the upper and lower boundaries of the rectangular detection frame are extracted, and the ordinate of the geometric center of the rectangle is calculated. The mathematical model is used to obtain the inclination angle of the metal substrate relative to the horizontal plane.

[0059] The i-th frame image A is captured in the same way as the initial frame image. i , and extract the detection area B i For image B i Perform Gaussian filtering and binarization to generate a minimum rectangular detection frame in the binary image that contains all pixels with a pixel value of 255. Extract the ordinate value y of the upper boundary of the rectangular detection frame upi , the ordinate value y of the lower boundarylowi , then the ordinate of the geometric center of the rectangle is Using mathematical models Calculate the inclination angle θ of the metal substrate relative to the horizontal plane i .

[0060] S6. Calculate the pixel change of the i-th frame relative to the initial frame, convert the pixel points into real distances using the size corresponding values, and finally obtain the displacement change; calculate the angle change of the metal substrate of the i-th frame relative to the horizontal plane, and perform a difference operation with the angle change of the initial frame relative to the horizontal plane to obtain the change in the inclination angle.

[0061] Calculate the pixel change △y of the ordinate of the i-th frame relative to the initial frame pixels =y Pi -y P0 , use the size corresponding value α to convert the pixel point into the real distance, and finally get the displacement change △y real =α·△y pixels The angle change of the metal substrate of the i-th frame relative to the horizontal plane is θ i , and perform a difference operation with the angle change θ0 of the initial frame relative to the horizontal plane to obtain the change in inclination △θ.

[0062] Example:

[0063] like Figure 1 As shown, the industrial camera, light source module, and development board are connected to the power supply module through a power cord, and the industrial camera is connected to the development board through a data cable. The industrial camera and light source module are installed on the guide rail, and the bi-telecentric lens is installed on the camera. According to the size of the PECVD detection window, the industrial camera and light source module are fixed to the appropriate position on the guide rail by screws to ensure that the camera's field of view and the output light of the light source module can cover the detection window. Fix the linear polarizer in front of the light source module and rotate the linear polarizer in front of the bi-telecentric lens to a suitable angle to ensure that the captured image is not interfered by ambient light, metal substrate, and glare on the surface of the side observation window of the PECVD equipment.

[0064] like Figure 2 As shown, an initial frame image A0 is captured, and a detection area is defined in the image. The coordinates of the upper left corner vertex of the area are (x1, y1) = (300, 300), and the coordinates of the lower right corner vertex are (x2, y2) = (1800, 1800). The image in the area is cut out and recorded as B0.

[0065] like Figure 3As shown in the figure, the image B0 is Gaussian filtered, where the Gaussian kernel is 5*5 and the standard deviation is 1. The filtered image is binarized using the Otsu method, and a rectangle is generated in the binarized image. The rectangle is defined as the minimum rectangular detection box containing all pixels with a pixel value of 255. The ordinate value y of the upper boundary of the rectangular detection box is extracted. up0 =379, the ordinate value y of the lower boundary low0 =753, the ordinate of the geometric center of the rectangle is According to the camera pixel size u = 3.45μm, the bi-telecentric lens magnification A = 0.243, the corresponding value of the size of a single pixel and the real world is

[0066] like Figure 4 As shown, the mathematical model is established in the image coordinate system as follows:

[0067]

[0068] where y up ,y low Represents the vertical coordinate value of the upper and lower boundaries of the rectangular detection box, W B0 represents the width of the rectangular area, D represents the thickness of the metal substrate, and θ represents the inclination angle of the metal substrate relative to the horizontal plane. low =753,y up =379, W B0 Substituting =1500 and D=352 into the mathematical model, we can obtain the inclination angle θ0 of the metal substrate relative to the horizontal plane = 0.839°.

[0069] like Figure 5 As shown, the i-th frame image A is captured using the same method as the initial frame image. i , and extract the detection area B i .

[0070] like Figure 6 As shown, for image B i Perform Gaussian filtering and binarization to generate a minimum rectangular detection frame in the binary image that contains all pixels with a pixel value of 255. Extract the ordinate value y of the upper boundary of the rectangular detection frame upi =365, the ordinate value y of the lower boundary lowi =1034, then the ordinate of the geometric center of the rectangle is

[0071] Using mathematical models The inclination angle θ of the metal substrate relative to the horizontal plane i =11.661°.

[0072] The pixel change △y of the ordinate of the i-th frame relative to the initial framepixels =y Pi -y P0 =134, and use the size corresponding value α to convert the pixel point into the real distance, and finally get the displacement change △y real =α·△y pixels =1902.8 μm, that is, relative to the initial frame, the metal substrate in the i-th frame image is translated downward by 1902.8 μm.

[0073] The angle change of the metal substrate of the i-th frame relative to the horizontal plane is θ i =11.661°, and the difference between the angle change of the initial frame relative to the horizontal plane θ0=0.839° is calculated to obtain the change of the inclination angle △θ=θ i -θ0=10.822°, that is, relative to the initial frame, the metal substrate in the i-th frame image is tilted counterclockwise by 10.822°.

Claims

1. A high-precision measurement device for the position variation of a PECVD metal substrate, characterized in that: Including housing, development board, power supply module, light source module, linear polarizer, guide rail, bi-telecentric lens and industrial camera; The industrial camera is placed parallel to the light source module, facing the side of the housing, and installed on the guide rail; The bi-telecentric lens is installed on the industrial camera; The linear polarizer covers the bi-telecentric lens and the light source module; The power supply module and the development board are placed on the rear side of the light source module and the industrial camera respectively; The guide rail is composed of two vertical rods and a horizontal rod, the horizontal rod and the vertical rod are fixed by screws, and the height of the horizontal rod is adjusted by fixing screw holes at different heights; The industrial camera, light source module and development board are connected to the power supply module via a power cord; The industrial camera is connected to the development board via a data cable.

2. A high-precision measurement device for position variation of a PECVD metal substrate according to claim 1, characterized in that: The industrial camera is connected to the dual telecentric lens through a screw hole.

3. A high-precision measurement device for position variation of a PECVD metal substrate according to claim 1, characterized in that: The guide rail is fixed to the bottom of the casing, and the industrial camera and light source module are fixed to the guide rail by screws.

4. A high-precision measurement device for position variation of a PECVD metal substrate according to claim 1, characterized in that: There are multiple screw holes on the guide rail. The industrial camera and the light source module are fixed by screws according to the size of the side detection window of the PECVD equipment to ensure that the camera field of view and the light source module can cover the detection window.

5. A high-precision measurement method using the measuring device as claimed in claim 1, characterized in that: The steps include: Step 1: Install the bi-telecentric lens according to the size of the side inspection window of the PECVD equipment, adjust the industrial camera and light source module to the appropriate position, and fix them with screws; fix the linear polarizer in front of the light source module, and rotate the angle of the linear polarizer in front of the bi-telecentric lens to ensure that the captured image is not interfered by ambient light, metal substrate and glare on the surface of the side observation window of the PECVD equipment; Step 2: Take the image A0 of the initial frame, define a detection area in the image, cut out the image in the area and record it as B0; Step 3: Perform Gaussian filtering on image B0, binarize the filtered image, and generate a minimum rectangular detection frame containing all pixel points with a pixel value of 255 in the binarized image; Extract the ordinate values ​​of the upper and lower boundaries of the rectangular detection frame, and calculate the ordinate of the geometric center of the rectangular frame. According to the camera pixel size and the magnification of the bi-telecentric lens, calculate the corresponding value of the size of a single pixel point and the real world; Step 4: Establish a mathematical model in the image coordinate system where y up ,y low Represents the vertical coordinate value of the upper and lower boundaries of the rectangular detection box, W B0 represents the width of the rectangular area, D represents the thickness of the metal substrate, and θ represents the inclination angle of the metal substrate relative to the horizontal plane; the inclination angle of the metal substrate relative to the horizontal plane is obtained by the model calculation; Step 5: Use the same method as the initial frame image to capture the i-th frame image A i , and extract the detection area B i ; For image B i Gaussian filtering and binarization are performed to generate a minimum rectangular detection frame containing all pixel points with a pixel value of 255 in the binarized image, the ordinate values ​​of the upper and lower boundaries of the rectangular detection frame are extracted, and the ordinate of the geometric center of the rectangular frame is calculated, and the inclination angle of the metal substrate relative to the horizontal plane is obtained using a mathematical model; Step 6: Calculate the pixel change of the i-th frame relative to the initial frame, convert the pixel points into real distances using the size corresponding values, and finally obtain the displacement change; The angle change of the metal substrate in the i-th frame relative to the horizontal plane is calculated, and the difference between the angle change of the initial frame relative to the horizontal plane is calculated to obtain the change of the inclination angle.

6. A high-precision measurement method according to claim 5, characterized in that: The step 2 is specifically as follows: capture an image A0 of the initial frame, define a detection area in the image, the upper left corner vertex coordinates of the area are (x1, y1), the lower right corner vertex coordinates are (x2, y2), and the image in the area is cut out and recorded as B0.

7. A high-precision measurement method according to claim 6, characterized in that: The step 3 is specifically as follows: Step 3-1: Perform Gaussian filtering on image B0, binarize the filtered image using Otsu's method, and generate a rectangle in the binarized image. The rectangle is defined as the minimum rectangular detection frame that contains all pixels with a pixel value of 255; Step 3-2: Extract the ordinate value y of the upper boundary of the rectangular detection box up0 , the ordinate value y of the lower boundary low0 , calculate the ordinate of the geometric center of the rectangle as According to the camera pixel size u and the bi-telecentric lens magnification A, the corresponding value α of the size of a single pixel and the real world is calculated.

8. A high-precision measurement method according to claim 7, characterized in that: The step 4 is specifically as follows: establishing a mathematical model where y up ,y low Represents the vertical coordinate value of the upper and lower boundaries of the rectangular detection box, W B0 represents the width of the rectangular area, D represents the thickness of the metal substrate, and θ represents the inclination angle of the metal substrate relative to the horizontal plane; the variable y low ,y up , W B0 , D are substituted into the mathematical model to obtain the inclination angle θ0 of the metal substrate relative to the horizontal plane.

9. A high-precision measurement method according to claim 8, characterized in that: The step 5 is specifically as follows: Step 5-1: Use the same method as the initial frame image to capture the i-th frame image A i , and extract the detection area B i ; Step 5-2: Image B i Perform Gaussian filtering and binarization to generate a minimum rectangular detection frame containing all pixels with a pixel value of 255 in the binary image; Step 5-3: Extract the ordinate value y of the upper boundary of the rectangular detection box upi , the ordinate value y of the lower boundary lowi , then the ordinate of the geometric center of the rectangle is Using mathematical models Calculate the inclination angle θ of the metal substrate relative to the horizontal plane i .

10. A high-precision measurement method according to claim 9, characterized in that: The step 6 is specifically as follows: Step 6-1: Calculate the pixel change △y of the vertical coordinate of the i-th frame relative to the initial frame pixels =y Pi -y P0 , use the size corresponding value α to convert the pixel point into the real distance, and finally get the displacement change △y real =α·△y pixels ; Step 6-2: The angle change θ of the metal substrate of the i-th frame relative to the horizontal plane i , and perform a difference operation with the angle change θ0 of the initial frame relative to the horizontal plane to obtain the change in inclination △θ.

Images