An adjustable multi-through hole data acquisition system and a method for measuring aperture and hole spacing

Abstract

The present invention discloses an adjustable multi-through-hole data acquisition system and an aperture and hole spacing measurement method, which relate to the field of machine vision measurement technology; an adjustable, non-contact method is used to perform real-time measurement of all through-holes of the equipment under test on the production line. The system includes a rack component module, a horizontal beam module, a strip light source module, an industrial camera, a ceramic calibration plate, a power module and a host computer system; the through-hole image of the equipment under test and the ceramic calibration plate image are collected by the industrial camera, and the ceramic calibration plate image is used to calibrate the intrinsic parameter matrix, the rotation matrix and the translation matrix of the extrinsic parameter of the industrial camera, and the collected through-hole image is grayed, binarized, expanded, corroded, and bilaterally filtered. The through-hole diameter is calculated by Hough circle edge fitting, and the through-hole spacing is calculated by the distance between the two circle centers, and it is judged whether the equipment under test meets the set standard according to the through-hole diameter and the hole spacing.

Description

Technical Field

[0001] The invention relates to the technical field of machine vision measurement, and in particular to an adjustable multi-through-hole data acquisition system and an aperture and hole spacing measurement method. Background Art

[0002] With the continuous development of my country's aerospace and information transmission industries, aerospace enterprises have drilled hundreds of thousands of through holes in composite materials during the manufacturing process, most of which are at the millimeter level. Their processing accuracy and measurement accuracy are crucial to avoid the failure of satellite-borne electronic equipment and the disintegration of satellite structures. Among them, the through holes on the mechanical mounting surface of satellite-borne electronic equipment are important basic component connection parts of various types of aviation equipment. The accuracy, performance, life and reliability of the aperture and hole spacing of the through holes on the mounting surface play a decisive role in the accuracy, performance, life and reliability of the main controller. In aviation equipment, satellite-borne electronic equipment is a high-precision product that requires not only the comprehensive support of mathematics, physics and other disciplines, but also the services of materials science, heat treatment technology, precision machining and measurement technology, numerical control technology and effective numerical methods and powerful computer technology. Satellite-borne electronic equipment can be said to be a product that represents the scientific and technological strength of a country. Although the manufacturing scale of satellite-borne electronic equipment is constantly increasing, the quality inspection technology of the through holes on the mechanical mounting surface of satellite-borne electronic equipment is still very backward. In my country, most of the detection methods for the through holes on the mechanical mounting surface of satellite-borne electronic equipment still use mechanical and optical measuring instruments, which are relatively backward. This manual random sampling detection method has low detection efficiency, low precision, and is prone to human errors. It cannot accurately detect every through hole on the mechanical mounting surface of the onboard electronic equipment. When using the measuring instrument to detect the test piece, it is inevitable to contact the test piece, which may cause secondary damage to the test piece. The traditional detection method can no longer meet the needs. Therefore, in order to ensure the quality of onboard electronic equipment and solve the problems of human eyes being unable to recognize or low recognition efficiency, high labor costs, etc., a multi-through hole data acquisition system and aperture and spacing measurement method suitable for the mechanical mounting surface of onboard electronic equipment with the characteristics of adjustable, non-contact, high measurement accuracy, and high detection efficiency are proposed. Summary of the invention

[0003] In order to solve the problems existing in the prior art, the present invention proposes an adjustable multi-through-hole data acquisition system and an aperture and hole spacing measurement method, which adopts an adjustable and non-contact method to perform real-time measurement of all through-holes of the equipment under test on the production line.

[0004] In a first aspect, the present invention proposes an adjustable multi-through-hole data acquisition system, including a rack assembly module, a horizontal beam module, a strip light source module, an industrial camera, a ceramic calibration plate, a power module and a host computer system.

[0005] The rack assembly module includes a machine vision experiment frame and an optical experiment platform, threaded holes are arranged at equal intervals on the optical experiment platform, and the machine vision experiment frame is mounted on the optical experiment platform through a first fixing component;

[0006] The horizontal beam module includes a horizontal beam, a circular connector and a pan / tilt quick-release plate. The pan / tilt quick-release plate and the circular connector are mounted on the slide groove of the horizontal beam through a third fixing component. The industrial camera is mounted on the pan / tilt quick-release plate through a fourth fixing component.

[0007] The ceramic calibration plate is placed on the device under test, and the through-hole image of the device under test and the ceramic calibration plate image under the light source of the strip light source module are collected by an industrial camera and transmitted to the host computer system.

[0008] The power module provides power for industrial cameras, bar light modules and host computer systems;

[0009] Ceramic calibration plates with fixed-pitch pattern arrays for correcting industrial lens distortion and determining the conversion relationship between physical size and pixels;

[0010] The bar light source module includes a plurality of bar light sources and a light source controller. The bar light sources are respectively mounted on the light source bracket through a fifth fixing component. After each bar light source is connected to the light source controller, the brightness and angle of each bar light source are independently controlled.

[0011] Bar light source, built-in LED light, emitting direct light;

[0012] The power supply module includes a 220V AC power supply and a DC power supply that converts the input 220V AC power supply into a stable 12V output; the 12V DC power supply powers each industrial camera, and the 220V AC power supply powers the strip light source module and the host computer;

[0013] The host computer system stores the acquired through-hole images of the device under test and performs image processing on the through-hole images of the device under test through PyCharm software.

[0014] The machine vision experiment frame is provided with a main pole, a light source support fixture, a horizontal beam fixture and a light source support. The main pole is mounted on the optical experiment platform through a second fixing component. The light source support fixture and the horizontal beam fixture are mounted on the main pole through round holes. The light source support is mounted on the light source support fixture through a round hole. The horizontal beam is mounted on the horizontal beam fixture through a square hole.

[0015] The horizontal beam is provided with a horizontal beam tail arm, a horizontal beam tail arm screw and a horizontal beam fixing screw, and the light source bracket is provided with a light source bracket tail arm screw and a light source bracket fixing screw. The measuring area is controlled by adjusting the horizontal beam tail arm screw and the light source bracket tail arm screw up and down and by adjusting the horizontal beam fixing screw and the light source bracket fixing screw forward and backward.

[0016] In a second aspect, the present invention provides an adjustable aperture and hole spacing measurement method, using the adjustable multi-through-hole data acquisition system, comprising the following steps:

[0017] S1: Install and adjust the multi-through-hole data acquisition system according to the size of the device under test, and select a ceramic calibration plate to place on the device under test;

[0018] S2: The ceramic calibration plate images are collected by the industrial camera respectively, which are used to calibrate the intrinsic parameter matrix, the rotation matrix and the translation matrix of the extrinsic parameter of the industrial camera, and each ceramic calibration plate image is read in turn. According to the position of the ceramic calibration plate in the image, the orientation coordinates of the ceramic calibration plate and the center coordinates of the square pattern in the ceramic calibration plate are found in each ceramic calibration plate area;

[0019] S3: by adjusting the distance a between the industrial camera and the through hole of the device under test to control the diameter of each through hole in the through hole image to reach b pixels, the through hole image of the device under test is collected;

[0020] S4: grayscale, binarization, dilation, erosion, and bilateral filtering are performed on the through-hole image. The through-hole diameter is calculated by Hough circle edge fitting, and the through-hole distance is calculated by the distance between the two circle centers. Whether the device under test meets the set standard is determined based on the through-hole diameter and the through-hole distance.

[0021] The S1 specifically includes the following steps:

[0022] S1.1: Install the main rod of the machine vision experiment frame on the optical experiment platform through the first fixing assembly, and fix the light source bracket fixing device and the horizontal beam fixing device on the main rod through the circular hole;

[0023] S1.2: Install the strip light source symmetrically on the light source bracket fixture;

[0024] S1.3: Install several sets of pan-tilt quick-release plates and circular connectors on the slide slots of the horizontal beam through the horizontal beam adjustment assembly at a set spacing, and install industrial cameras vertically and symmetrically on each pan-tilt quick-release plate, and adjust the horizontal beam tail arm adjustment assembly until the industrial lens on the industrial camera is parallel to the optical experimental platform;

[0025] S1.4: Place the device under test statically on the optical experimental platform. According to the size of the device under test, adjust the light source bracket fixing component and the horizontal beam fixing component respectively until the device under test is in the center of the viewing range of each industrial camera and the projection of the strip light source. Each industrial camera captures all the through holes on the device under test. Adjust the clarity of the through hole image of the device under test by adjusting the focal length of the industrial lens.

[0026] S1.5: Select a ceramic calibration plate that is 5%-10% the size of the device under test and place it on the device under test;

[0027] After the multi-through-hole data acquisition system is installed and adjusted, a test board with the same material as the device under test is cut out by a CNC machine tool, and holes are punched on the test board in a manner of equal spacing, equal apertures in horizontal rows, and gradually increasing apertures in vertical rows to conduct a preliminary experiment, and collect experimental through-hole image data.

[0028] The S2 is specifically:

[0029] Get the internal and external parameters of the industrial camera for calibration. Set the number of ceramic calibration plate images to n. There are m square patterns on each ceramic calibration plate image. The center of the j-th square pattern on the i-th ceramic calibration plate image is d. ij The coordinates on the ceramic calibration plate image are: D(κ, φ i , μ i ,ξ ij )=κ(φ|μ)ξ ij , where φ i and μ i is the rotation matrix and translation vector corresponding to the i-th ceramic calibration plate image, κ is the intrinsic parameter matrix of the industrial camera, then the probability density function of the center coordinates of the square pattern in the ceramic calibration plate image is shown in formula (1):

[0030]

[0031] The likelihood function is obtained from the probability density function of formula (1), as shown in formula (2):

[0032]

[0033] Depend on get We can get κ and φ i , μ i , where κ=(f,p,r x , r y , l x , l y ) is the intrinsic parameter of the industrial camera, f is the focal length of the industrial camera, p is the distortion coefficient matrix of the industrial camera, l x , l y The actual physical value represented by a pixel in the x-axis direction and the y-axis direction, r x , r y is the number of pixels that differ between the pixel coordinates of the image center and the pixel coordinates of the image origin in the x-axis direction and the y-axis direction, respectively, i=(α, β, γ) is the rotation matrix of the external parameters of the industrial camera, α is the angle of rotation of the industrial camera around the x-axis of the industrial camera coordinate system, β is the angle of rotation of the industrial camera around the y-axis of the industrial camera coordinate system, γ is the angle of rotation of the industrial camera around the z-axis of the industrial camera coordinate system, μ i =(μ x , μ y , μ z ) is the translation vector of the external parameters of the industrial camera, μ x is the distance that the industrial camera translates along the x-axis in the world coordinate system, μ y is the distance that the industrial camera translates along the y-axis in the world coordinate system, μ z It is the distance that the industrial camera translates along the z-axis in the world coordinate system.

[0034] S4 specifically includes the following steps:

[0035] S4.1: grayscale the collected through-hole image to reduce it to a one-dimensional matrix; calculate the weighted average of the three components of RGB (three primary colors) of the collected through-hole image by formula (3) to obtain a through-hole grayscale image, as shown in formula (3):

[0036] N(i,j)=0.3R(i,j)+0.59G(i,j)+0.11B(i,j) (3)

[0037] Wherein, N(i, j) is the gray value of the converted through-hole grayscale image at (i, j), R(i, j) is the R component value of the converted through-hole grayscale image at (i, j), G(i, j) is the G component value of the converted through-hole grayscale image at (i, j), and B(i, j) is the B component value of the converted through-hole grayscale image at (i, j);

[0038] S4.2: Binarize the through-hole grayscale image by using the OTSU method (maximum between-class variance method), set t as the segmentation threshold of the through-hole in the through-hole grayscale image and the background other than the through-hole in the through-hole grayscale image, the ratio of the number of pixels of the through-hole in the through-hole grayscale image to the through-hole grayscale image is w1, the average grayscale of the through-hole in the through-hole grayscale image is v1, the ratio of the number of pixels of the background other than the through-hole in the through-hole grayscale image to the through-hole grayscale image is w2, the average grayscale of the background other than the through-hole in the through-hole grayscale image is v2, that is, the total average grayscale of the through-hole grayscale image is v=w1·v1+w2·v2; the variance of the through-hole in the through-hole grayscale image and the background other than the through-hole in the through-hole grayscale image is as shown in formula (4):

[0039] S = w1·(v1-v) 2+w2·(v2-v) 2 =w1·w2·(v1-v2) 2 (4)

[0040] Depend on When , v1 and v2 are calculated to obtain the through-hole binary image;

[0041] S4.3: Dilate and erode the through-hole binary image, as shown in formula (5):

[0042]

[0043] Among them, C x (x) represents the structural element, E x represents the working space, and x represents any pixel in the image;

[0044] Filling the through-holes in the through-hole binary image and the concave portions at the edges of the through-hole binary image by dilation;

[0045] The boundary points in the through-hole binary image are corroded, as shown in formula (6):

[0046]

[0047] Among them, C y (y) represents the structural element, E y represents the working space, and y represents any pixel in the image;

[0048] S4.4: Perform bilateral filtering on the through-hole binary image after expansion and erosion, and set the weight of any point s in the window centered on h in the through-hole binary image after expansion and erosion in Gaussian filtering. The pixel value weight is The spatial distance weight is Then the filtering effect W of the bilateral filter is as shown in formula (7):

[0049]

[0050] in, is the weighted sum of each pixel value in the filter serial port, used for weight normalization;

[0051] S4.5: Perform Hough circle edge fitting on the through-hole bilateral filter image to calculate the aperture, and set the Hough circle fitting equation to:

[0052] (xX) 2 +(yY) 2 =R xy 2 (8)

[0053] Convert formula (8) into the Hough circle fitting equation in the polar coordinate system, as shown in formula (9):

[0054]

[0055] Among them, X represents the horizontal coordinate of the center of the circle, Y represents the vertical coordinate of the center of the circle, and R xy represents the radius corresponding to the center of the circle, η represents the rotation angle, η∈[0°, 360°];

[0056] By using Canny edge detection on the through-hole bilateral filter image, Sebol u The operator formula is shown in (10):

[0057]

[0058] Calculate the derivatives H of the grayscale H(X, Y) of the through-hole bilateral filtering image along the two directions X and Y x =H(X, Y)*Sebol x (x, y) and H y =H(X, Y)*Sebol y (x, y), find the increase in gradient And the direction of the gradient Get all the circular lines Ω XY and normal F XY , non-maximum suppression is performed in the 4-neighborhood of Hough space Γ, and the circle center threshold is set to The cumulative sum of the cth circle center in the Hough space is O c , then the coordinates of the center of the circle are (X, Y) = Set the center (X, Y) to all the circumference lines Ω XY The distance is V XY , and the stable interval distance value is V range = {V∈Γ|V MIN ≤V≤V MAX}, where V MAX Indicates the maximum distance, V MIN Indicates the minimum distance, setting the radius threshold to The number of the same distance values ​​from the center of the circle to all the circumferences is Z XY , then the radius R corresponding to the center of the circle xy =

[0059] S4.6: Calculate the hole distance by the distance between the two points of the circle center. Set the two hole diameters to R1 and R2, and the coordinates of the circle center to (X1, Y1) and (X2, Y2). Then the hole distance L is xy for: The actual distance ω represented by one pixel obtained by industrial camera calibration is used to calculate the actual aperture r of the through hole of the device under test. xy =R xy ω, and the actual hole distance l xy =L xy ω;

[0060] S4.7: Determine the aperture and hole spacing results. Use the above steps to calculate the actual aperture and hole spacing of all through holes in the image. Take the average value through N1 measurements and compare it with the set size. Output whether the comparison result is within the absolute error range. If it is, it is qualified, otherwise it is unqualified.

[0061] Beneficial technical effects

[0062] 1. The present invention has the advantages of simple operation and portability. It can be taken and placed for static measurement, and multiple through holes can be measured simultaneously. It does not require professional knowledge, does not require mobile positioning, and is free from jitter interference.

[0063] 2. The present invention uses a ceramic calibration plate to calibrate the internal and external parameters of the industrial camera, avoiding the influence of image distortion on the measurement results, and the obtained calculation results have high accuracy and small errors;

[0064] 3. The aperture and hole spacing measurement system designed by the present invention is flexible and can adjust the position of the industrial camera to measure the aperture and hole spacing of the through-holes of the equipment to be tested with different sizes. It can measure multiple through-holes on the plane to be tested at the same time without positioning. Compared with the traditional manual vernier caliper measurement, it is non-destructive, fast, highly accurate, and not prone to false detection and missed detection; compared with other machine measurement methods, it avoids the disadvantages of low efficiency and high energy consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 A schematic diagram of the structure of an adjustable multi-through-hole data acquisition system provided by an embodiment of the present invention;

[0066] Figure 2 A block diagram of an adjustable multi-through-hole data acquisition system provided by an embodiment of the present invention;

[0067] Figure 3 A schematic diagram of a ceramic calibration plate and satellite-borne electronic equipment provided in an embodiment of the present invention;

[0068] Figure 4 A schematic diagram of a through-hole design for an aviation aluminum plate provided in an embodiment of the present invention;

[0069] Figure 5 A flowchart of a method for measuring aperture and spacing provided in an embodiment of the present invention;

[0070] Figure 6 A diagram showing the results of measuring the apertures of some through holes on the mechanical mounting surface of the onboard electronic equipment provided by an embodiment of the present invention;

[0071] Among them, 1. Rack assembly module; 11. Machine vision experiment frame; 12. Optical experiment platform; 121. Threaded hole; 111. Main rod; 112. Light source bracket fixing device; 113. Horizontal beam fixing device; 114. Light source bracket; 1141. Light source bracket tail arm screw; 1142. Light source bracket fixing screw; 2. Horizontal beam module; 21. Horizontal beam; 211. Horizontal beam tail arm; 212. Horizontal beam tail arm screw; 213. Horizontal beam fixing screw; 22. Circular connector; 23. Pan-tilt quick release plate; 3. Strip light source module; 31. Strip light source; 32. Light source controller; 4. Industrial camera; 5. Ceramic calibration board; 6. Power module; 7. Host computer system; 71. Computer; 8. Industrial lens; 9. Satellite-borne electronic equipment. DETAILED DESCRIPTION

[0072] The specific implementation of the present invention is further described in detail below in conjunction with the drawings and examples.

[0073] This embodiment proposes an adjustable multi-through-hole data acquisition system and an aperture diameter and spacing measurement method, which adopts an adjustable non-contact multi-through-hole data acquisition system and an aperture diameter and spacing measurement method suitable for the mechanical mounting surface of the satellite-borne electronic equipment 9. By performing real-time measurement of all through-holes on the mechanical mounting surface of the satellite-borne electronic equipment 9 on the production line, production time and resource costs are greatly saved, and the efficiency of measurement in the production process is improved.

[0074] In the first aspect, the present embodiment proposes an adjustable multi-through-hole data acquisition system including a rack assembly module 1, a horizontal beam module 2, a strip light source module 3, a Hikvision MV-CE200-11UM industrial camera 4, a WMO-045-3.0-C ceramic calibration board 5, a power module 6 and a host computer system 7, such as Figure 1 shown.

[0075] The horizontal beam module 2 and the strip light source module 3 are fixed to the rack component module 1, and the measurement area is controlled by the horizontal beam module 2 and the strip light source module 3. The Hikvision MV-CE200-11UM industrial camera 4 is fixed to the horizontal beam module 2, and the WMO-045-3.0-C ceramic calibration plate 5 is placed on the mechanical installation surface of the satellite-borne electronic device 9. The through-hole image of the mechanical installation surface of the satellite-borne electronic device 9 and the image of the WMO-045-3.0-C ceramic calibration plate 5 are collected by the Hikvision MV-CE200-11UM industrial camera 4. The image of the WMO-045-3.0-C ceramic calibration plate 5 is used to calibrate the intrinsic parameter matrix, the rotation matrix and the translation matrix of the extrinsic parameter of the Hikvision MV-CE200-11UM industrial camera 4. The power module 6 supplies power to the Hikvision MV-CE200-11UM industrial camera 4, the strip light source module 3 and the host computer system 7, as shown in FIG. Figure 2 shown.

[0076] The Hikvision MV-CE200-11UM industrial camera 4 converts the optical signal into an ordered electrical signal, and modulates the light beam through the installed Hikvision MVL-KF1228M-12MPE (neutral) industrial lens 8, images the target on the photosensitive surface of the CMOS image sensor, and collects the through-hole image and the WMO-045-3.0-C ceramic calibration plate 5 image. The model of the industrial camera 4 is Hikvision MV-CE200-11UM, and the model of the industrial lens 8 is Hikvision MVL-KF1228M-12MPE (neutral).

[0077] The rack assembly module 1 includes a machine vision experiment frame 11 and an optical experiment platform 12 . The optical experiment platform 12 has threaded holes 121 arranged at intervals of 2.5 cm. The machine vision experiment frame 11 is fixed to the optical experiment platform 12 by a first screw.

[0078] The machine vision experiment frame 11 has two main rods 111, a light source support fixture 112, a horizontal beam fixture 113 and a light source support 114. The two main rods 111 are mounted on the optical experiment platform 12 through second screws, the light source support fixture 112 and the horizontal beam fixture 113 are mounted on the two main rods 111 through round holes, and the light source support 114 is mounted on the light source support fixture 112 through a round hole. The horizontal beam 21 is mounted on the horizontal beam fixture 113 through a square hole.

[0079] The horizontal beam module 2 includes a horizontal beam 21, a circular connector 22 and a gimbal quick release plate 23. The gimbal quick release plate 23 and the circular connector 22 are mounted on the slide groove of the horizontal beam 21 through a third screw. Each Hikvision MV-CE200-11UM industrial camera 4 is fixed to the gimbal quick release plate 23 through a fourth screw.

[0080] The horizontal beam 21 is provided with a horizontal beam tail arm 211, a horizontal beam tail arm screw 212 and a horizontal beam fixing screw 213, and the light source bracket 114 is provided with a light source bracket tail arm screw 1141 and a light source bracket fixing screw 1142. The measuring area is controlled by adjusting the horizontal beam tail arm screw 212 and the light source bracket tail arm screw 1141 up and down and by adjusting the horizontal beam fixing screw 213 and the light source bracket fixing screw 1142 forward and backward.

[0081] Ceramic calibration plate 5, a flat plate with a fixed-pitch pattern array that can correct the distortion of Hikvision MVL-KF1228M-12MPE (neutral) industrial lens 8 and determine the conversion relationship between physical size and pixels. The model of ceramic calibration plate 5 is WMO-045-3.0-C, made of ceramic, with a thickness of 1mm, a size of 45mm×45mm, and an accuracy of ±1um.

[0082] The bar light source module 3 includes 4 Hikvision MV-LEVS-60-4-SY bar light sources 31 and 1 Hikvision MV-LEVS-60-4-SY light source controller 32. The 4 Hikvision MV-LEVS-60-4-SY bar light sources 31 are respectively fixed on the four sides of the light source bracket 114 by the fifth screws. After the Hikvision MV-LEVS-60-4-SY bar light source 31 is connected to the Hikvision MV-LEVS-60-4-SY light source controller 32, the brightness and adjustment angle are independently controlled. The model of the light source controller 32 is Hikvision MV-LEVS-60-4-SY.

[0083] Bar light source 31: Built-in LED light, emitting direct light, stable light emission and high brightness. The model of bar light source 31 is Hikvision MV-LLDS-237-38-W.

[0084] The power module 6 includes a 220V AC power supply and two DC power supplies that convert the input 220V AC power supply into a stable output of 12V; wherein each Hikvision MV-CE200-11UM industrial camera 4 requires a 12V DC power supply, and the 220V AC power supply powers the Hikvision MV-LEVS-60-4-SY bar light source 31 and the computer 71.

[0085] The host computer system 7 stores the collected through-hole images of the device under test in the computer 71 , and performs image processing on the through-hole images of the device under test through the PyCharm software.

[0086] In the second aspect, in the detection of the aperture and hole spacing of the through holes on the mechanical mounting surface of the onboard electronic equipment 9, the operation is difficult and requires high precision. This embodiment proposes an adjustable aperture and hole spacing measurement method, using the adjustable multi-through hole data acquisition system, including the following steps:

[0087] S1: Install and adjust the multi-through-hole data acquisition system according to the size of the device under test, and select a ceramic calibration plate to be placed on the device under test; including the following steps:

[0088] S1.1: Fix the two main rods 111 of the machine vision experiment frame 11 on the optical experiment platform 12 through the first screws, and fix the light source bracket fixing device 112 and the horizontal beam fixing device 113 on the two main rods 111 through the round holes;

[0089] S1.2: Install four Hikvision MV-LEVS-60-4-SY bar light sources 31 symmetrically on the light source bracket fixture 112;

[0090] S1.3: Fix three sets of pan / tilt quick-release plates 23 and circular connectors 22 to the slide groove of the horizontal beam 21 at a spacing of 15.00 cm through the horizontal beam fixing screws 213, vertically and symmetrically install the Hikvision MV-CE200-11UM industrial camera 4 on each pan / tilt quick-release plate 23, and adjust the horizontal beam tail arm screws 212 until the Hikvision MVL-KF1228M-12MPE (neutral) industrial lens 8 on the Hikvision MV-CE200-11UM industrial camera 4 is parallel to the optical experimental platform 12;

[0091] S1.4: Place the satellite-borne electronic device 9 statically on the optical experimental platform 12, and adjust the light source bracket fixing screws 1142 and the horizontal beam fixing screws 213 according to the size of the satellite-borne electronic device 9, until the mechanical mounting surface of the satellite-borne electronic device 9 is in the center of the framing range of each Hikvision MV-CE200-11UM industrial camera 4 and the projection of the Hikvision MV-LLDS-237-38-W bar light source 31, so that each Hikvision MV-CE200-11UM industrial camera 4 can capture all the through holes on the mechanical mounting surface of the satellite-borne electronic device 9, and adjust the clarity of the through hole image on the mechanical mounting surface of the satellite-borne electronic device 9 by adjusting the focal length of the Hikvision MVL-KF1228M-12MPE (neutral) industrial lens 8;

[0092] S1.5: Place the WMO-045-3.0-C ceramic calibration plate 5 with a thickness of 1 mm, a size of 45 mm × 45 mm, and an accuracy of ±1 um with a fixed spacing pattern array facing upward on the upper left, upper middle, upper right, middle left, center, middle right, lower left, lower middle, and lower right positions on the mechanical mounting surface of the onboard electronic equipment 9, such as Figure 3 As shown;

[0093] S2: The material of the mechanical mounting surface of the onboard electronic equipment 9 is made of aviation aluminum. Considering factors such as economy and processing difficulty, an aviation aluminum plate of the same material is selected to replace the mechanical mounting surface of the onboard electronic equipment 9; an aviation aluminum plate of 300mm×300mm in size and 8mm in thickness is cut out by a CNC machine tool, and then holes are punched on the aviation aluminum plate in a design in which the holes are equally spaced, the apertures are equal in the horizontal rows, and the apertures in the vertical rows increase one by one. The diameters of the through holes are 3.00mm, 3.10mm, 3.20mm, 4.00mm, 4.10mm, 4.20mm, 5.00mm, 5.10mm, 5.20mm, such as Figure 4 As shown;

[0094] S3: 20 images of WMO-045-3.0-C ceramic calibration plate 5 are collected by Hikvision MV-CE200-11UM industrial camera 4 respectively, which are used to calibrate the internal parameter matrix, external parameter rotation matrix and translation matrix of Hikvision MV-CE200-11UM industrial camera 4, read each image of WMO-045-3.0-C ceramic calibration plate 5 in turn, find the position of WMO-045-3.0-C ceramic calibration plate 5 in the image, and find the azimuth coordinates of WMO-045-3.0-C ceramic calibration plate 5 and the center coordinates of the square pattern in WMO-045-3.0-C ceramic calibration plate 5 in each WMO-045-3.0-C ceramic calibration plate 5 area, specifically:

[0095] Get the internal and external references of Hikvision MV-CE200-11UM industrial camera 4 for calibration. Set the number of WMO-045-3.0-C ceramic calibration plate 5 images to n. There are m square patterns on each WMO-045-3.0-C ceramic calibration plate 5 image. The center of the j-th square pattern on the i-th WMO-045-3.0-C ceramic calibration plate 5 image is d. ij The coordinates on the WMO-045-3.0-C ceramic calibration plate 5 image are: D(k, φ i , μ i ,ξ ij )=κ(φ|μ)ξ ij , where φ i and μ i is the rotation matrix and translation vector corresponding to the i-th WMO-045-3.0-C ceramic calibration plate 5 image, κ is the internal parameter matrix of Hikvision MV-CE200-11UM industrial camera 4, then the probability density function of the center coordinates of the square pattern in the WMO-045-3.0-C ceramic calibration plate 5 image is shown in formula (1):

[0096]

[0097] The likelihood function is obtained from the probability density function of formula (1), as shown in formula (2):

[0098]

[0099] Depend on get We can get κ and φ i , μ i , where κ=(f,p,r x , r y , l x , l y ) is the intrinsic parameter of Hikvision MV-CE200-11UM industrial camera 4, f is the focal length of Hikvision MV-CE200-11UM industrial camera 4, p is the distortion coefficient matrix of Hikvision MV-CE200-11UM industrial camera 4, l x , l y The actual physical value represented by a pixel in the x-axis direction and the y-axis direction, r x , r y is the number of pixels that differ between the pixel coordinates of the image center and the pixel coordinates of the image origin in the x-axis direction and the y-axis direction, respectively, i =(α, β, γ) is the rotation matrix of the external parameters of Hikvision MV-CE200-11UM industrial camera 4, α is the rotation angle of Hikvision MV-CE200-11UM industrial camera 4 around the x-axis direction of Hikvision MV-CE200-11UM industrial camera 4 coordinate system, β is the rotation angle of Hikvision MV-CE200-11UM industrial camera 4 around the y-axis direction of Hikvision MV-CE200-11UM industrial camera 4 coordinate system, γ is the rotation angle of Hikvision MV-CE200-11UM industrial camera 4 around the z-axis direction of Hikvision MV-CE200-11UM industrial camera 4 coordinate system, μ i =(μ x , μ y , μ z ) is the translation vector of the extrinsic parameters of Hikvision MV-CE200-11UM industrial camera 4, μ x is the translation distance of Hikvision MV-CE200-11UM industrial camera 4 along the x-axis in the world coordinate system, μ y is the translation distance of Hikvision MV-CE200-11UM industrial camera 4 along the y-axis in the world coordinate system, μ z is the translation distance of Hikvision MV-CE200-11UM industrial camera 4 along the z-axis in the world coordinate system;

[0100] S4: by adjusting the distance between the 20-megapixel Hikvision MV-CE200-11UM industrial camera 4 and the mechanical mounting surface of the satellite-borne electronic device 9 to 35.00 cm, the diameter of each through hole in the through hole image is controlled to reach 200 pixels, and the through hole image of the mechanical mounting surface of the satellite-borne electronic device 9 is collected;

[0101] S5: grayscale processing, binarization processing, expansion, corrosion, and bilateral filtering operations are performed on the through-hole image, the through-hole diameter is calculated by Hough circle edge fitting, and the through-hole distance is calculated by the distance between the two circle centers, and the satellite-borne electronic equipment 9 is judged whether it meets the set standards according to the through-hole diameter and the hole distance, such as Figure 5 As shown, the specific steps include:

[0102] S5.1: Since the size detection is independent of the image color, the collected through-hole image is grayed to reduce it to a one-dimensional matrix; since the human eye is most sensitive to green and least sensitive to blue, the pixel value of the collected through-hole image is weighted averaged by the three components of RGB (three primary colors) according to formula (3) to obtain the through-hole grayscale image, as shown in formula (3):

[0103] N(i,j)=0.3R(i,j)+0.59G(i,j)+0.11B(i,j) (3)

[0104] Wherein, N(i, j) is the gray value of the converted through-hole grayscale image at (i, j), R(i, j) is the R component value of the converted through-hole grayscale image at (i, j), G(i, j) is the G component value of the converted through-hole grayscale image at (i, j), and B(i, j) is the B component value of the converted through-hole grayscale image at (i, j);

[0105] S5.2: Binarize the through-hole grayscale image. The choice of threshold has a great influence on the effect of edge detection: if the threshold is too large, the edge detection will be incomplete; if the threshold is too small, the interference of non-real edge detection will increase. Since the aperture detection size of the through-hole on the mechanical mounting surface of the satellite electronic device 9 does not require background, the through-hole grayscale image is binarized by using the OTSU method (maximum between-class variance method), and t is set as the segmentation of the through-hole in the through-hole grayscale image and the background other than the through-hole in the through-hole grayscale image. The threshold, the ratio of the number of pixels of the through-hole in the through-hole grayscale image to the through-hole grayscale image is w1, the average grayscale of the through-hole in the through-hole grayscale image is v1, the ratio of the number of pixels of the background other than the through-hole in the through-hole grayscale image to the through-hole grayscale image is w2, the average grayscale of the background other than the through-hole in the through-hole grayscale image is v2, that is, the total average grayscale of the through-hole grayscale image is v = w1·v1+w2·v2; the variance of the through-hole in the through-hole grayscale image and the background other than the through-hole in the through-hole grayscale image is as shown in formula (4):

[0106] S = w1·(v1-v) 2 +w2·(v2-v) 2 =w1·w2·(v1-v2) 2 (4)

[0107] Depend on Calculate v1 and v2 to obtain the through-hole binary image;

[0108] S5.3: Perform dilation and corrosion operations on the through-hole binary image; as shown in formula (5):

[0109]

[0110] Among them, C x (x) represents the structural element, E x represents the working space, and x represents any pixel in the image;

[0111] Wherein, the through-holes in the through-hole binary image and the concave parts at the edges of the through-hole binary image are filled by dilation;

[0112] The boundary points in the through-hole binary image are corroded, as shown in formula (6):

[0113]

[0114] Among them, C y (y) represents the structural element, E y represents the working space, and y represents any pixel in the image;

[0115] S5.4: Perform bilateral filtering on the binary image of the through hole after expansion and corrosion. Since the binary image after expansion and corrosion contains high Gaussian noise and salt and pepper noise, bilateral filtering is performed on it to remove the noise. Bilateral filtering is a nonlinear filter for edge-preserving denoising that determines the filter coefficient by geometric space distance and pixel difference. Set the weight of any point s in the window centered on h in the binary image of the through hole after expansion and corrosion in the Gaussian filter. The pixel value weight is The spatial distance weight is Then the filtering effect W of the bilateral filter is as shown in formula (7):

[0116]

[0117] in, is the weighted sum of each pixel value in the filter serial port, used for weight normalization;

[0118] S5.5: Perform Hough circle edge fitting on the through-hole bilateral filter image to calculate the aperture, and set the Hough circle fitting equation to:

[0119] (xX) 2 +(yY) 2 =R xy 2 (8)

[0120] Convert formula (8) into the Hough circle fitting equation in the polar coordinate system, as shown in formula (9):

[0121]

[0122] Among them, X represents the horizontal coordinate of the center of the circle, Y represents the vertical coordinate of the center of the circle, and R xy represents the radius corresponding to the center of the circle, η represents the rotation angle, η∈[0°, 360°];

[0123] By using Canny edge detection on the through-hole bilateral filter image, Sebol u The operator formula is shown in (10):

[0124]

[0125] Calculate the derivatives H of the grayscale H(X, Y) of the through-hole bilateral filtering image along the two directions X and Y x =H(X, Y)*Sebol x (x, y) and Hy = H (X, Y) * Sebol y (x, y), find the increase in gradient And the direction of the gradient Get all the circular lines Ω XY and normal F XY , non-maximum suppression is performed in the 4-neighborhood of Hough space Γ, and the circle center threshold is set to The cumulative sum of the cth circle center in the Hough space is O c , then the coordinates of the center of the circle are (X, Y) = Set the center (X, Y) to all the circumference lines Ω Xy The distance is V XY , and the stable interval distance value is V range = {V∈Γ|V MIN ≤V≤V MAX}, where V MAX Indicates the maximum distance, V MIN Indicates the minimum distance, setting the radius threshold to The number of the same distance values ​​from the center of the circle to all the circumferences is Z XY , then the radius corresponding to the center of the circle

[0126] S5.6: The hole distance is calculated by the distance between the centers of two circles. There are two types of distance between pixels: one is parallel pixels, which are in the four directions of up, down, left, and right. The distance between these parallel pixels is 1 pixel; the second is the oblique direction, which is connected pixels. The oblique directions are the upper left corner, lower left corner, upper right corner, and lower right corner. The distance between these oblique directions is pixels; when measuring the aperture distance, it is necessary to calculate the distance according to the connection mode between pixels; suppose the two apertures are R1 and R2, and the coordinates of the center of the circle are (X1, Y1) and (X2, Y2), then the aperture distance L is xy for: The actual distance ω represented by one pixel obtained by the calibration of the Hikvision MV-CE200-11UM industrial camera 4 is used to calculate the actual aperture r of the through hole on the mechanical mounting surface of the satellite-borne electronic equipment 9 xy =R xy ω, and the actual hole distance l xy =L xy ω;

[0127] S5.7: Aperture and hole spacing result judgment, use the above steps to calculate the actual aperture and hole spacing of all through holes in the image, take the average value through multiple measurements, compare with the design size, and output whether the comparison result is within the absolute error range. If it is, it is qualified, otherwise it is unqualified; through multiple experiments, the aperture and hole spacing of some through holes on the mechanical mounting surface of the satellite electronic equipment 9 are measured, and the measurement results are shown in Table 1 below; by comparing the measured value and the actual value of the aperture of the through hole on the mechanical mounting surface of the satellite electronic equipment 9, the measurement error is within ±0.04mm, and the actual measured distance between the two holes is 50.024mm, which is also within the error of ±0.04, indicating that the measurement system can quickly and accurately measure the aperture and hole spacing of the through hole on the mechanical mounting surface of the satellite electronic equipment 9. In the later stage, the relevant algorithms will be improved or other algorithms will be tried for experiments to improve the robustness of the system. Some image data collected in the experiment (taking through hole diameters of 3.00mm, 4.00mm and 5.00mm as examples) are shown in Table 1. Figure 6 shown.

[0128] Table 1 Aperture measurement results of some through holes on the mechanical mounting surface of onboard electronic equipment 9 (unit: mm)

[0129]

[0130] This embodiment proposes an adjustable multi-through-hole data acquisition system and an aperture diameter and spacing measurement method, which measures the aperture of some through holes on the mechanical mounting surface of the satellite-borne electronic equipment 9. Experimental results show that the aperture diameter and spacing measurement has high accuracy and high efficiency, and the system is stable and reliable.

Claims

1. An adjustable multi-through-hole data acquisition system, characterized in that: It includes rack assembly module, horizontal beam module, strip light source module, industrial camera, ceramic calibration board, power module and host computer system; The rack assembly module includes a machine vision experiment frame and an optical experiment platform, threaded holes are arranged at equal intervals on the optical experiment platform, and the machine vision experiment frame is mounted on the optical experiment platform through a first fixing component; The horizontal beam module includes a horizontal beam, a circular connector and a pan-tilt quick-release plate. Several groups of pan-tilt quick-release plates and circular connectors are installed on the slide groove of the horizontal beam through a third fixing component at a set interval. An industrial camera is vertically and symmetrically installed on each pan-tilt quick-release plate. The industrial camera is installed on the pan-tilt quick-release plate through a fourth fixing component. The industrial lens on the industrial camera is adjusted to be parallel to the optical experimental platform. The device under test is placed statically on the optical experimental platform. The machine vision experimental frame is adjusted according to the size of the device under test so that the device under test is in the center of the viewing range of each industrial camera and the strip light source module. Each industrial camera is used to shoot all through holes on the device under test, and the clarity of the through hole image of the device under test is adjusted by adjusting the focal length of the industrial lens. The ceramic calibration plate is placed on the device under test, and the through-hole image of the device under test and the ceramic calibration plate image under the light source of the strip light source module are collected by an industrial camera and transmitted to the host computer system; The power module provides power for the industrial camera, bar light module and host computer system.

2. The adjustable multi-through-hole data acquisition system according to claim 1, characterized in that: Ceramic calibration plates with fixed-pitch pattern arrays for correcting industrial lens distortion and determining the conversion relationship between physical size and pixels; The bar light source module includes a plurality of bar light sources and a light source controller. The bar light sources are respectively mounted on the light source bracket through a fifth fixing component. After each bar light source is connected to the light source controller, the brightness and angle of each bar light source are independently controlled. Bar light source, built-in LED light, emitting direct light; The power supply module includes a 220V AC power supply and a DC power supply that converts the input 220V AC power supply into a stable 12V output; the 12V DC power supply powers each industrial camera, and the 220V AC power supply powers the strip light source module and the host computer; The host computer system stores the acquired through-hole images of the device under test and performs image processing on the through-hole images of the device under test through PyCharm software.

3. The adjustable multi-through-hole data acquisition system according to claim 2, characterized in that: The machine vision experiment frame is provided with a main pole, a light source support fixture, a horizontal beam fixture and a light source support. The main pole is mounted on the optical experiment platform through a second fixing component. The light source support fixture and the horizontal beam fixture are mounted on the main pole through round holes. The light source support is mounted on the light source support fixture through a round hole. The horizontal beam is mounted on the horizontal beam fixture through a square hole.

4. The adjustable multi-through-hole data acquisition system according to claim 2, characterized in that: The horizontal beam is provided with a horizontal beam tail arm, a horizontal beam tail arm screw and a horizontal beam fixing screw, and the light source bracket is provided with a light source bracket tail arm screw and a light source bracket fixing screw. The measuring area is controlled by adjusting the horizontal beam tail arm screw and the light source bracket tail arm screw up and down and by adjusting the horizontal beam fixing screw and the light source bracket fixing screw forward and backward.

5. A method for measuring adjustable aperture and hole spacing, using the adjustable multi-through-hole data acquisition system of claim 1, comprising the following steps: S1: Install and adjust the multi-through-hole data acquisition system according to the size of the device under test, and select a ceramic calibration plate to place on the device under test; S2: The ceramic calibration plate images are collected by the industrial camera respectively, which are used to calibrate the intrinsic parameter matrix, the rotation matrix and the translation matrix of the extrinsic parameter of the industrial camera, and each ceramic calibration plate image is read in turn. According to the position of the ceramic calibration plate in the image, the orientation coordinates of the ceramic calibration plate and the center coordinates of the square pattern in the ceramic calibration plate are found in each ceramic calibration plate area; S3: by adjusting the distance a between the industrial camera and the through hole of the device under test, the diameter of each through hole in the through hole image is controlled to reach b pixels, and the through hole image of the device under test is collected; S4: grayscale, binarization, dilation, erosion, and bilateral filtering are performed on the through-hole image. The through-hole diameter is calculated by Hough circle edge fitting, and the through-hole distance is calculated by the distance between the two circle centers. Whether the device under test meets the set standard is determined based on the through-hole diameter and the through-hole distance.

6. The adjustable aperture and aperture spacing measurement method according to claim 5, characterized in that: After installing and adjusting the multi-through-hole data acquisition system, a test board with the same material as the device under test is cut out by a CNC machine tool. Preliminary experiments are carried out on the test board with equal spacing, equal apertures in the horizontal rows, and gradually increasing apertures in the vertical rows, and the experimental through-hole image data is collected.

7. The adjustable aperture and aperture spacing measurement method according to claim 5, characterized in that: The S1 specifically includes the following steps: S1.1: Install the main rod of the machine vision experiment frame on the optical experiment platform through the first fixing assembly, and fix the light source bracket fixing device and the horizontal beam fixing device on the main rod through the circular hole; S1.2: Install the strip light source symmetrically on the light source bracket fixture; S1.3: Install several sets of pan-tilt quick-release plates and circular connectors on the slide slots of the horizontal beam through the horizontal beam adjustment assembly at a set spacing, and install industrial cameras vertically and symmetrically on each pan-tilt quick-release plate, and adjust the horizontal beam tail arm adjustment assembly until the industrial lens on the industrial camera is parallel to the optical experimental platform; S1.4: Place the device under test statically on the optical experimental platform. According to the size of the device under test, adjust the light source bracket fixing component and the horizontal beam fixing component respectively until the device under test is in the center of the viewing range of each industrial camera and the projection of the strip light source. Each industrial camera captures all the through holes on the device under test. Adjust the clarity of the through hole image of the device under test by adjusting the focal length of the industrial lens. S1.5: Select a ceramic calibration plate that is 5%-10% the size of the device under test and place it on the device under test.

8. The adjustable aperture and aperture spacing measurement method according to claim 5, characterized in that: The S2 specifically includes: Get the internal and external parameters of the industrial camera for calibration. Set the number of ceramic calibration plate images to n. There are m square patterns on each ceramic calibration plate image. The center of the j-th square pattern on the i-th ceramic calibration plate image is d. ij The coordinates on the ceramic calibration plate image are: D(κ, φ i , μ i ,ξ ij )=κ(φ|μ)ξ ij , where φ i and μ i is the rotation matrix and translation vector corresponding to the i-th ceramic calibration plate image, κ is the intrinsic parameter matrix of the industrial camera, then the probability density function of the center coordinates of the square pattern in the ceramic calibration plate image is shown in formula (1): The likelihood function is obtained from the probability density function of formula (1), as shown in formula (2): Depend on get We can get κ and φ i , μ i , where κ=(f,p,r x , r y , l x , l y ) is the intrinsic parameter of the industrial camera, f is the focal length of the industrial camera, p is the distortion coefficient matrix of the industrial camera, l x , l y The actual physical value represented by a pixel in the x-axis direction and the y-axis direction, r x , r y is the number of pixels that differ between the pixel coordinates of the image center and the pixel coordinates of the image origin in the x-axis direction and the y-axis direction, respectively, i =(α, β, γ) is the rotation matrix of the external parameters of the industrial camera, α is the angle of rotation of the industrial camera around the x-axis of the industrial camera coordinate system, β is the angle of rotation of the industrial camera around the y-axis of the industrial camera coordinate system, γ is the angle of rotation of the industrial camera around the z-axis of the industrial camera coordinate system, μ i =(μ x , μ y ,μ z ) is the translation vector of the external parameters of the industrial camera, μ x is the distance that the industrial camera translates along the x-axis in the world coordinate system, μ y is the distance that the industrial camera translates along the y-axis in the world coordinate system, μ z It is the distance that the industrial camera translates along the z-axis in the world coordinate system.

9. The adjustable aperture and aperture spacing measurement method according to claim 5, characterized in that: The S4 specifically includes the following steps: S4 specifically includes the following steps: S4.1: grayscale the collected through-hole image to reduce it to a one-dimensional matrix; perform weighted average calculation on the RGB three components of the pixel value of the collected through-hole image to obtain a through-hole grayscale image. S4.2: Binarizing the through-hole grayscale image to obtain a through-hole binary image; S4.3: performing expansion and corrosion operations on the through-hole binary image; filling the through-holes in the through-hole binary image and the concave parts at the edges of the through-hole binary image by expansion; S4.4: performing a bilateral filtering operation on the through-hole binary image after dilation and corrosion to obtain a through-hole bilateral filtered image; S4.5: performing Hough circle edge fitting on the through-hole bilateral filter image to calculate the aperture; performing Canny edge detection on the through-hole bilateral filter image; S4.6: Calculate the hole spacing by the distance between the centers of two circles; S4.7: Aperture and hole spacing result judgment: use the actual aperture and hole spacing of all through holes in the image, take the average value through N1 measurements, compare it with the set size, and output whether the comparison result is within the absolute error range. If it is, it is qualified, otherwise it is unqualified.

10. The adjustable aperture and aperture spacing measurement method according to claim 9, characterized in that: The binarization process uses the maximum inter-class variance method to perform binarization on the through-hole grayscale image, sets t as the segmentation threshold of the through-holes in the through-hole grayscale image and the background other than the through-holes in the through-hole grayscale image, the ratio of the number of pixels of the through-holes in the through-hole grayscale image to the through-hole grayscale image is w1, the average grayscale of the through-holes in the through-hole grayscale image is v1, the ratio of the number of pixels of the background other than the through-holes in the through-hole grayscale image to the through-hole grayscale image is w2, the average grayscale of the background other than the through-holes in the through-hole grayscale image is v2, that is, the total average grayscale of the through-hole grayscale image is v=w1·v1+w2·v2; the variance of the through-holes in the through-hole grayscale image and the background other than the through-holes in the through-hole grayscale image.

Images