High-precision calibration method, measurement method and system based on line-scan digital camera image

Abstract

The invention provides a line-scan digital camera image-based high-precision calibration method, a measurement method and a system, and relates to the field of automatic optical measurement of plates, and the calibration method comprises the following steps: constructing a hardware mechanism for line-scan digital camera image calibration; acquiring a linear array camera image based on a hardware mechanism; performing distortion calibration on each line-scan digital camera based on the line-scan digital camera images so as to obtain reference point X coordinates, X-axis average precision, distortion correction coefficient arrays, working distances, grid numbers, Y longitudinal view offset pixels and Y longitudinal average precision corresponding to different intervals, wherein the reference point X coordinates, the X-axis average precision, the distortion correction coefficient arrays, the working distances, the grid numbers, the Y longitudinal view offset pixels and the Y longitudinal average precision correspond to different intervals; and performing multi-machine joint calibration on the plurality of line-scan digital cameras based on the line-scan digital camera images so as to obtain the proportion of the blind area width corresponding to different intervals to the height difference of the calibration plane. Calibration data are obtained on the basis of high-precision calibration, automatic optical detection is carried out on the thickness of the plate in combination with the calibration data, and the problems that existing plate size measurement is low in efficiency, poor in precision and poor in repeatability are solved.

Description

Technical Field

[0001] The present invention relates to the field of automated optical measurement of plates, and more particularly, to a high-precision calibration method, measurement method, and system based on line array camera images. Background Art

[0002] In the production scenario of furniture cabinet plates, the plates need to go through processing procedures such as cutting, edge banding, and drilling before a finished plate can be obtained. Among them, there are processing size errors in each processing procedure of the plates; therefore, it is very important to verify whether the size of the finished plate meets the expected design size, which is a key factor to ensure the normal production of furniture cabinets.

[0003] Currently, most factories still mainly rely on manual measurement to control the size of the plates, using a tape measure to measure and recheck the length, width, hole positions, and slot specifications of the plates. This method has low efficiency, poor measurement accuracy, and poor repeatability. Summary of the Invention

[0004] The present invention provides a high-precision calibration method, measurement method, and system based on line array camera images, which solves the problems of low efficiency, poor accuracy, and poor repeatability in the existing measurement of plate sizes.

[0005] In a first aspect, an embodiment of the present invention provides a high-precision calibration method based on line array camera images, and the method includes the following processes:

[0006] Construct a hardware mechanism for calibrating line array camera images;

[0007] Obtain line array camera images based on the hardware mechanism for calibrating line array camera images;

[0008] Perform distortion calibration on each line array camera based on the line array camera images to obtain the reference point X coordinates, X-axis average accuracy, distortion correction coefficient array, working distance, number of grids, Y longitudinal field of view offset pixels, and Y longitudinal average accuracy corresponding to different intervals;

[0009] Perform multi-camera joint calibration on multiple line array cameras based on the line array camera images to obtain the ratio of the blind zone width to the calibration plane height difference corresponding to different intervals.

[0010] In the above embodiment, the present invention calibrates the hardware mechanism by means of distortion calibration of line scan cameras and multi-camera joint calibration, and can quickly achieve high-precision calibration.

[0011] As some optional embodiments of the present application, the hardware mechanism for calibrating line array camera images includes a checkerboard calibration board and multiple line array cameras; among them, multiple line array cameras are arranged side by side, and the lenses of the line array cameras are all vertically downward-mounted.

[0012] In some alternative embodiments of the present application, the process of obtaining the linear array camera image based on the hardware mechanism for calibrating the linear array camera image is as follows:

[0013] Place the checkerboard calibration plate on two calibration planes respectively;

[0014] When the checkerboard calibration plate is on two calibration planes, use the linear array camera to collect images of the checkerboard calibration plate to obtain the linear array camera images of the checkerboard calibration plate on the two calibration planes.

[0015] In some alternative embodiments of the present application, the process of performing distortion calibration on each linear array camera based on the linear array camera image is as follows:

[0016] Perform X-axis calibration on each linear array camera based on the linear array camera images of the checkerboard calibration plate on two calibration planes to obtain the X coordinates of the reference points corresponding to different intervals, the average accuracy in the X axis, the distortion correction coefficient array, the working distance, and the number of grids;

[0017] Perform Y-axis longitudinal calibration on each linear array camera based on the linear array camera images of the checkerboard calibration plate on two calibration planes to obtain the Y-axis longitudinal field-of-view offset pixels and the average accuracy in the Y axis corresponding to different intervals.

[0018] In some alternative embodiments of the present application, the process of performing X-axis calibration on each linear array camera based on the linear array camera images of the checkerboard calibration plate on two calibration planes is as follows:

[0019] Obtain the linear array camera images of the checkerboard calibration plate on plane 1 and plane 2, and splice and separate the field of view of the linear array camera images to obtain the X coordinates of the calibration points of the linear array camera in different intervals;

[0020] When the checkerboard calibration plate is on plane 1, perform distortion correction on different intervals respectively in combination with the X coordinates of the calibration points of the linear array camera in different intervals to obtain the X coordinates of the reference points corresponding to different intervals, the average accuracy in the X axis, the distortion correction coefficient array, and the number of grids;

[0021] When the checkerboard calibration points are on plane 2, perform thickness compensation on the structural parameters of the hardware mechanism in combination with the X coordinates of the calibration points of the linear array camera in different intervals to obtain the working distances corresponding to different intervals.

[0022] In some alternative embodiments of the present application, the process of performing multi-camera joint calibration on multiple linear array cameras based on the linear array camera images is as follows:

[0023] Obtain the physical length of the checkerboard calibration plate on plane 2 and the projected physical length projected onto plane 1;

[0024] Construct the projection relationship for each interval based on the physical length of the checkerboard calibration plate on plane 2, the projected physical length back-projected onto plane 1, and the number of grids in different intervals.

[0025] Based on the projection relationship for each interval, calculate the relationship between the blind zone width corresponding to different intervals and the height difference between plane 1 and plane 2, so as to obtain the ratio of the blind zone width corresponding to the linear array camera in different intervals to the height difference of the calibration plane.

[0026] In a second aspect, the present invention provides a high-precision measurement method based on linear array camera images, characterized in that the high-precision measurement method uses the calibration data obtained by the high-precision calibration method based on linear array camera images, and the high-precision measurement method includes the following processes:

[0027] Obtain the height difference between the plate to be detected and the calibration plane; wherein, the height difference between the plate to be detected and the calibration plane is the thickness of the plate to be detected.

[0028] Predict the coordinates of two pixel points corresponding to the two edges of the plate to be detected and the intervals corresponding to the two pixel points.

[0029] Obtain the calibration data for the corresponding interval, and calculate the physical distance between the two pixel points based on the calibration data for the interval; wherein, the physical distance is the width of the plate to be measured.

[0030] In a third aspect, the present invention provides a high-precision measurement system based on linear array camera images, and the system includes:

[0031] A plate thickness acquisition unit, which is used to obtain the height difference between the plate to be detected and the calibration plane; wherein, the height difference between the plate to be detected and the calibration plane is the thickness of the plate to be detected.

[0032] A placement interval prediction unit, which is used to predict the coordinates of two pixel points corresponding to the two edges of the plate to be detected and the intervals corresponding to the two pixel points.

[0033] A plate thickness calculation unit, which is used to obtain the calibration data for the corresponding interval, and calculate the physical distance between the two pixel points based on the calibration data for the interval; wherein, the physical distance is the width of the plate to be measured.

[0034] In a fourth aspect, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, it implements the high-precision measurement method based on linear array camera images.

[0035] In a fifth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the high-precision measurement method based on the linear array camera image is implemented.

[0036] The beneficial effects of the present invention are as follows: The present invention combines high-precision calibration data to perform automated optical detection on the thickness of the plate, with relatively high detection efficiency, relatively high detection accuracy, and good repeatability. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for use in the embodiments. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can also be obtained based on these drawings without creative efforts.

[0038] Figure 1 is the overall flowchart of the high-precision calibration method and measurement method described in the embodiments of the present invention;

[0039] Figure 2 is the front view of the hardware mechanism described in the embodiments of the present invention;

[0040] Figure 3 is the top view of the hardware mechanism described in the embodiments of the present invention;

[0041] Figure 4 is the comparison diagram of the image mosaics of four linear array cameras when the calibration plate is in plane 1 and plane 2 in the embodiments of the present invention;

[0042] Figure 5 is the curve graph of the distortion correction coefficient in interval 1 in the embodiments of the present invention;

[0043] Figure 6 is the projection relationship diagram of plane 1 and plane 2 in the embodiments of the present invention;

[0044] Figure 7 is the partial mosaic diagram of the images of camera 1 and camera 2 in the embodiments of the present invention;

[0045] Figure 8 is the partial enlarged view of the linear array camera image in the embodiments of the present invention;

[0046] Figure 9 is the physical schematic diagram of the linear array camera in the embodiments of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0048] To solve the problems of low efficiency, poor accuracy, and poor repeatability existing in the measurement of the size of existing plates, embodiments of the present invention provide a high-precision calibration method, a high-precision measurement method, and a system based on a linear array camera image. Please refer to Figure 1 , Figure 1 As shown in the overall flowchart of the high-precision calibration method and the measurement method, the specific process of the high-precision calibration method is as follows:

[0049] (1) Construct the hardware mechanism for calibrating the linear array camera image.

[0050] In the embodiments of the present invention, the hardware mechanism for calibrating the linear array camera image includes a conveyor belt, a checkerboard calibration plate, and a linear array camera. Among them, the number of linear array cameras is 4, and the 4 linear array cameras are arranged side by side, and the lenses of the linear array cameras are all installed vertically downward. Specifically, a guiding edge device is provided on the left side of the conveyor belt, and by setting the guiding edge device, it can always ensure that the plate to be measured is on the left. Please refer to Figure 2 , Figure 2 As shown in the front view of the hardware mechanism.

[0051] In the embodiments of the present invention, since zero error cannot be achieved in hardware debugging, hardware debugging errors are allowed within a certain range. Specifically, the hardware error allows the angle between the field of view of each linear array camera and the object movement direction to be within 90 ± 0.008°, and taking the field of view of the leftmost linear array camera (camera 1) as the reference baseline, the field of view offset pixels of the other 3 linear array cameras in the Y longitudinal direction are within ±5. Please refer to Figure 3 , Figure 3 As shown in the top view of the hardware mechanism.

[0052] (2) Obtain the linear array camera image based on the hardware mechanism for calibrating the linear array camera image.

[0053] Specifically, first, the checkerboard calibration plate is respectively placed on two calibration planes; then when the checkerboard calibration plate is on the two calibration planes, the linear array camera is used to collect images of the checkerboard calibration plate to obtain the linear array camera images of the checkerboard calibration plate on the two calibration planes.

[0054] In the embodiments of the present invention, the checkerboard calibration plate uses a square with a single grid of 20 mm, a total of 58 grids horizontally, and 59 calibration points (numbered from point 0 to point 58). At the same time, calibration plates of other specifications can also be selected for calibration, and the embodiments of the present invention do not limit this.

[0055] (3) Perform distortion calibration on each linear array camera. Among them, the distortion calibration includes X-axis calibration and Y-longitudinal calibration.

[0056] In the embodiments of the present invention, since the line array camera has an imaging principle of single-line acquisition and line-by-line stitching, theoretically, only one line of images containing all the calibration points needs to be acquired to complete the X-axis calibration; for the Y-axis calibration, the number of lines acquired by the line array camera is linearly related to the number of camera encoder signals, and the number of encoder information is linearly related to the moving distance of the line body. Therefore, the essence of the Y-axis calibration is to determine the moving distance of the line body represented by one line of images (i.e., the Y-axis single-pixel accuracy). Specifically, the Y-axis calibration of the line array camera is relatively easy to implement. In the embodiments of the present invention, only a simple description is given for the Y-axis calibration, but it should be noted that the Y-axis single-pixel accuracy should be close to the X-axis single-pixel accuracy to avoid obvious image stretching or compression phenomena.

[0057] Specifically, the process of the X-axis calibration is as follows:

[0058] (3.1) Obtain the images of the four line array cameras when the checkerboard calibration board is in plane 1 and plane 2, and splice and separate the field of view of the four line array camera images to obtain the X coordinates of the calibration points in different ranges.

[0059] First, collect images through the four line array cameras; then, splice the images of the four line array cameras. Please refer to Figure 4 , Figure 4 which is the comparison diagram of the splicing of the images of the four line array cameras when the checkerboard calibration board is in plane 1 and plane 2; then, take the X coordinates of the calibration points in the clearest row (such as the first row) from the spliced image; finally, separate the X coordinates of the calibration points according to the field of view of the four line array cameras to obtain the X coordinates of the calibration points in five ranges.

[0060] Please refer to Table 1. Table 1 is an example of the X coordinates of the calibration points within the field of view of the four line array cameras (since the X coordinates are relatively long, the decimal points of the X coordinates are omitted, but the decimals are not omitted during actual calculations); at this time, the height difference ΔH between plane 1 and plane 2 is 18.1 mm.

[0061]

[0062]

[0063] Table 1 is an example of the X coordinates of the calibration points within the field of view of the four line array cameras

[0064] In the embodiments of the present invention, since the four line array camera lenses are all vertically downward-mounted and the point closest to the center of the lens is used as the calibration reference point; therefore, the field of view of camera 1 needs to be separated into two parts (in the embodiments of the present invention, it is separated by point 8), so five intervals (interval 1, interval 2, interval 3, interval 4, interval 5) of the four line array cameras need to be separately calibrated for distortion.

[0065] (3.2) When the checkerboard calibration plate is on Plane 1, the X coordinates of the calibration points within the field of view of the 4 linear array cameras are combined, and distortion correction is performed on 5 different intervals respectively.

[0066] Define the number of grids within a certain interval as N, that is, without considering the overlapping part of the field of view, all calibration points of each linear array camera are involved in calibration.

[0067] Define the serial number of the calibration point as i, the X coordinate on Plane 1 as X[i], and the X coordinate on Plane 2 as X2[i]; define the serial number of the reference point as n, the X coordinate of the reference point on Plane 1 as X[n], and the X coordinate of the reference point on Plane 2 as X2[n].

[0068] On Plane 1, the average precision of this interval is:

[0069] precision_x = (N × 20) ÷ (X[n] - X[n - N]), applicable to Interval 1;

[0070] precision_x = (N × 20) ÷ (X[n + N] - X[n]), applicable to Intervals 2, 3, 4, 5;

[0071] On Plane 1, the average width of each grid in this interval is:

[0072] box_width = (X[n] - X[n - N]) ÷ N, applicable to Interval 1;

[0073] box_width = (X[n + N] - X[n]) ÷ N, applicable to Intervals 2, 3, 4, 5;

[0074] Define the actual pixel distance of the point with serial number i relative to the reference point and the actual cumulative grid width as:

[0075] length_a[i] = X[n] - X[i], applicable to Interval 1;

[0076] length_a[i] = X[i] - X[n], applicable to Intervals 2, 3, 4, 5;

[0077] Define the theoretical pixel distance of the point with serial number i relative to the reference point and the theoretical cumulative grid width as:

[0078] length_s[i] = boxwidth × (n - i), applicable to Interval 1;

[0079] length_s[i] = boxwidth × (i - n), applicable to Intervals 2, 3, 4, 5;

[0080] Define the distortion correction coefficient of the point with serial number i as:

[0081] rate[i] = length_a[i] / length_s[i];

[0082] Taking the left side of Camera 1 (i.e., Interval 1) as an example, where N = 8 and n = 8, then precision_x = (8 × 20) / (2478 - 479) = 0.080039 mm / pxiel, box_width = (2478 - 479) / 8 = 249.8775 pixel, and Table 2 is obtained. Table 2 is the distortion correction parameters for Interval 1.

[0083]

[0084] Table 2 is the distortion correction parameters for Interval 1

[0085] Then, draw a scatter plot of the X coordinate and the distortion correction coefficient on Plane 1, and use a 4th-order polynomial to fit the scatter trend line. Define the polynomial coefficients as the array factor1[], and obtain the curve equation graph. Please refer to Figure 5 , Figure 5 which is the distortion correction coefficient curve graph for Interval 1. Then the 5 values in the array factor1[] are: 7.42505E-17, -9.54029E-14, -9.42591E-10, 3.39681E-06, 0.99859397.

[0086] Finally, record the average precision Precision_x1, the X coordinate X[n] of the reference point, and the polynomial coefficients factor1[] into the calibration data. Similarly, complete the calculation and recording of the calibration data for Intervals 2 to 5.

[0087] (3.3) When the checkerboard calibration points are on Plane 2, combine the X coordinates of the calibration points within the field of view of the 4 linear array cameras to perform thickness compensation on the structural parameters of the hardware mechanism.

[0088] In the embodiment of the present invention, the physical meaning of the H value in the hardware mechanism is the value from the calibration plane 1 to the optical center of the lens. This H value cannot be accurately measured manually, and there are also tolerances in the lens itself and the hardware installation structure. In the actual scenario, a more accurate value can be obtained only by using the data back-calculation method.

[0089] Define the coordinate of point i within a certain interval as x[i], and its distortion correction coefficient is:

[0090] f[x[i]] = factor[0] × x[i] 4 + factor[1] × x[i] 3 + factor[2] × x[i] 2+ factor[3]

[0091] × x[i] + factor[4]

[0092] Then, the coordinate X2[n] of the reference point on Plane 2 has a distortion correction coefficient of f[X2[n]];

[0093] Then, for the point X2[n - N] (Interval 1) or X2[n + N] (Intervals 2, 3, 4, 5), the calculated distortion correction coefficient is f[X2[n - N]] or f[X2[n + N]];

[0094] Then, define the distance from the coordinate x[i] of point i in a certain interval to the reference point of that interval as:

[0095] g[x[i]] = (x[i] - X[n]) ÷ f[x[i]] × precision_x

[0096] Then, the physical distance from X2[n] to the reference point X[n] is g[X2[n]];

[0097] Then, the physical distance from the point X2[n - N] (Interval 1) or X2[n + N] (Intervals 2, 3, 4, 5) to the reference point X[n] is g[X2[n - N]] or g[X2[n + N]].

[0098] Please refer to Figure 6 , Figure 6 which is the projection relationship diagram of Plane 1 and Plane 2 (taking Interval 1 as an example); as can be seen from the figure:

[0099] It is applicable to Interval 1;

[0100] It is applicable to Intervals 2, 3, 4, 5;

[0101] Given the distance ΔH from Plane 1 to Plane 2, the value of H can be obtained. It should be noted that at this time, the point n + N may have exceeded the field of view of the line array camera, and it is necessary to perform the pre - processing of N = N - 1 (such as in Intervals 2 and 4). Taking the left side of Camera 1 (i.e., Interval 1) as an example, Table 3 is obtained, and Table 3 is the estimated value of the structural parameters of the hardware mechanism.

[0102]

[0103] Table 3 is the estimated value of the structural parameters in Interval 1 for the hardware mechanism

[0104] As can be seen from the table, when H = 500.5 mm, the physical grid width of the real calibration board on plane 2 can be restored more accurately, with an error within ±0.02 mm, meeting the accuracy requirements of this application scenario. (H - ΔH) / H can be understood as the thickness compensation coefficient, which varies with ΔH. Similarly, the H values in intervals 2 to 5 can be calculated and also need to be recorded in the calibration data.

[0105] Specifically, the process of the Y - longitudinal calibration is as follows:

[0106] (3.4) Obtain the images of the four linear array cameras when the checkerboard calibration board is on plane 1 or plane 2, splice and locally magnify the images of the four linear array cameras, and judge the image differences between adjacent images to obtain the pixel offset of the camera's Y - longitudinal field of view.

[0107] Please refer to Figure 7 , Figure 7 As shown in, which is a local spliced and magnified view of the images of camera 1 and camera 2. Among them, the left side is the linear array camera image of camera 1, and the right side is the linear array camera image of camera 2. It can be directly observed that the same line on the checkerboard calibration board is about 1 pixel behind in camera 1 compared to camera 2. Therefore, the pixel offset Offset_y1 of camera 2 in the Y - longitudinal field of view is equal to - 1. Similarly, the pixel offset Offset_y2 of camera 3 in the Y - longitudinal field of view and the pixel offset Offset_y3 of camera 4 in the Y - longitudinal field of view can be obtained.

[0108] (3.5) Obtain the pixel length of the linear array camera corresponding to the checkerboard calibration board grid in the Y - longitudinal direction, and obtain the Y - longitudinal accuracy of the linear array camera according to the grid length of the checkerboard calibration board.

[0109] Please refer to Figure 8 , Figure 8 As shown in, which is a local magnified view of a certain linear array camera image. Among them, it can be measured that the pixel length of 3 grids of the calibration board in the Y - direction is 753.7688 pixel. And we know that the grid of the calibration board is a square with a side length of 20 mm. So the physical length of 3 grids is 60 mm. Then the Y - longitudinal accuracy is precision_y = 60÷753.7688 = 0.0796 mm / pixel.

[0110] (4) Perform multi - camera joint calibration on the distorted - calibrated linear array cameras.

[0111] Specifically, the process of the multi - camera joint calibration is as follows:

[0112] In the embodiment of the present invention, when multi - camera joint measurement is performed, only the aforementioned calibration data is not enough. Please refer to Figure 9 , Figure 9This is the physical schematic diagram of the linear array camera. Among them, the physical length l (the total length of the checkerboard calibration board) on plane 2 is projected onto plane 1 with a physical length of L, and L = A + B + C + D + E. After performing distortion correction and thickness compensation through the aforementioned method, it can be calculated that the total length of the reverse projection of L onto plane 1 is a + b + c + d + e, lacking three parts, namely bc, cd, and de, which can be understood as the blind areas caused by the thickness change. As can be seen from the figure, these three parts all have a fixed proportional relationship with ΔH. Define these three proportional values as R1, R2, and R3 respectively, that is:

[0113] bc = ΔH × R1, cd = ΔH × R2, de = ΔH × R1;

[0114] As can be seen from the figure, the lengths of B, C, and D are the distances between adjacent interval reference points. After defining and removing the overlapping fields of view, the number of grids in each interval is N1, N2, N3, N4, N5 respectively, and the H values of each interval are H1, H2, H3, H4, H5 respectively. According to the projection relationship, there are:

[0115] B = N2 × 20, b = B × ((H2 - ΔH) / H2);

[0116] C = N3 × 20, c = C × ((H3 - ΔH) / H3);

[0117] D = N4 × 20, d = D × ((H4 - ΔH) / H4);

[0118] From the aforementioned definitions and projection relationships, it can be known that:

[0119] A = -g[X2[0]], a = A × ((H1 - ΔH) / H1);

[0120] Then there is:

[0121] a + b + bc + g[X2[n3]] × ((H3 - ΔH) / H3) = (N1 + N2) × 20;

[0122] That is, bc can be calculated, and then R1 can be calculated. Furthermore:

[0123] a + b + bc + c + cd + g[X2[n4]] × ((H5 - ΔH) / H5)

[0124] = (N1 + N2 + N3) × 20;

[0125] That is, cd can be calculated, and then R2 can be calculated. Furthermore:

[0126] a + b + bc + c + cd + d + g[X2[n5]] × ((H5 - ΔH) / H5)

[0127] =(N1 + N2 + N3 + N4) × 20;

[0128] Then de can be calculated, and thus R3 can be calculated. Record R1, R2, and R3 into the calibration data. So far, all the calibration contents have been completed. Now, how to use the calibration data will be introduced.

[0129] An embodiment of the present invention provides a high-precision measurement method based on linear array camera images. The method includes the following steps:

[0130] (5) Combine the calibration data to measure the width of the plate to be detected.

[0131] (5.1) Obtain the height difference between the plate to be detected and the calibration plane. Here, the height difference between the plate to be detected and the calibration plane is the thickness of the plate to be detected. In the embodiment of the present invention, now define the height difference of the product thickness relative to the calibration plane 1 as Δh;

[0132] (5.2) Predict the coordinates of two pixel points corresponding to the two edges of the plate to be detected and the intervals corresponding to the two pixel points; for example, the pixel coordinates of the two points whose distance needs to be measured are point A1(xa, ya) located in interval 1 and point B1(xb, yb) located in interval 5.

[0133] (5.3) Obtain the calibration data of the corresponding interval and calculate the physical distance between the two pixel points based on the calibration data of the interval; where the physical distance is the width of the plate to be measured. Please refer to

[0134] Table 4. Table 4 is the parameter definition of the calibration data.

[0135]

[0136] Table 4 is the parameter definition of the calibration data

[0137] Then the distortion correction coefficient of point A1 is:

[0138] f[xa] = factor1[0] × xa 4 + factor1[1] × xa 3 + factor1[2] × xa 2 + factor1[3]

[0139] × xa + factor1[4];

[0140] Then the distortion correction coefficient of point B1 is:

[0141] f[xb] = factor5[0] × xb 4 + factor5[1] × xb 3 + factor5[2] × xb2

[0142] +factor5]3]×xb+factor5[4[;

[0143] The physical distance of point A1 relative to the reference point of interval 1 is:

[0144] g[xa] = (xa - x1) ÷ f[xa] × precision_x1;

[0145] The physical distance of point B1 relative to the reference point of interval 5 is:

[0146] g[xb] = (xb - x5) ÷ f[xb] × precision_x5;

[0147] Then the physical distance of points A1 and B1 in the X direction is:

[0148]

[0149]

[0150] Then the physical distance between points A1 and B1 is:

[0151]

[0152] Among them, the physical distance lengthA1B1 between points A1 and B1 is the width of the plate to be detected.

[0153] In summary, in the embodiment of the present invention, within a certain range (limited by the imaging depth of field range of the optical system) where the plate thickness is higher or lower than the calibration plane, high-precision measurement can still be achieved, with an error within ±1 / 2 pixel, and the horizontal X and vertical Y of the camera image are independently calibrated. The checkerboard calibration plate used only needs its width to be close to the field of view width and its height to be at least 2 rows, and high-precision calibration data can be obtained quickly; and combined with the high-precision calibration data, automatic optical detection of the plate thickness is carried out, solving the problems of low efficiency, poor accuracy, and poor repeatability existing in the existing plate size measurement.

[0154] In addition, in an embodiment, based on the same inventive concept as the foregoing embodiment, the embodiment of the present invention provides a high-precision measurement system based on a linear array camera image. The system corresponds one-to-one to the foregoing method for high-precision measurement based on a linear array camera image in Embodiment 1. The system includes:

[0155] A plate thickness acquisition unit, which is used to acquire the height difference between the plate to be detected and the calibration plane; among them, the height difference between the plate to be detected and the calibration plane is the thickness of the plate to be detected;

[0156] A placement interval prediction unit, which is used to predict the coordinates of two pixel points corresponding to the two side edges of the to-be-detected sheet and the intervals corresponding to the two pixel points;

[0157] A sheet thickness calculation unit, which is used to obtain the calibration data of the corresponding interval and calculate the physical distance between the two pixel points based on the calibration data of the interval; wherein, the physical distance is the width of the to-be-measured sheet.

[0158] It should be noted that in this embodiment, each unit in the high-precision measurement system based on the line array camera image corresponds one by one to each step in the high-precision measurement method based on the line array camera image in the foregoing embodiment. Therefore, the specific implementation manner and the achieved technical effects of this embodiment can refer to the implementation manner of the foregoing high-precision measurement method based on the line array camera image, and will not be elaborated here.

[0159] In addition, in one embodiment, the present application further provides a computer device, which includes a processor, a memory, and a computer program stored in the memory. When the computer program is run by the processor, it implements the method in the foregoing embodiment.

[0160] In addition, in one embodiment, the present application further provides a computer storage medium, on which a computer program is stored. When the computer program is run by the processor, it implements the method in the foregoing embodiment.

[0161] In some embodiments, the computer-readable storage medium may be a memory such as FRAM, ROM, PROM, EPROM, EEPROM, flash memory, magnetic surface memory, optical disc, or CD-ROM; or it may be various devices including one or any combination of the above memories. The computer may be various computing devices including intelligent terminals and servers.

[0162] In some embodiments, the executable instructions may be in the form of a program, software, software module, script, or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including being deployed as an independent program or being deployed as a module, component, subroutine, or other unit suitable for use in a computing environment.

[0163] As an example, executable instructions may, but need not, correspond to a file in a file system, may be stored as part of a file that stores other programs or data, such as in one or more scripts in a HyperText Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files storing one or more modules, subroutines, or code portions).

[0164] By way of example, executable instructions may be deployed to be executed on one computing device, or on multiple computing devices located at one site, or on multiple computing devices distributed across multiple sites and interconnected by a communication network.

[0165] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0166] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0167] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory / random access memory, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a multimedia terminal device (which can be a mobile phone, a computer, a television receiver, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0168] The above are only preferred embodiments of the present application, and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A high-precision calibration method based on linear array camera images, characterized in that The calibration method includes the following processes: Construct the hardware mechanism for linear array camera image calibration; Obtain the linear array camera images based on the hardware mechanism for linear array camera image calibration; Perform distortion calibration on each linear array camera based on the linear array camera images to obtain the reference point X coordinates, X-axis average accuracy, distortion correction coefficient array, working distance, grid number, Y-axis vertical field-of-view offset pixels, and Y-axis vertical average accuracy corresponding to different intervals; Perform multi-camera joint calibration on multiple linear array cameras based on the linear array camera images to obtain the ratio of the blind zone width to the height difference of the calibration plane corresponding to different intervals.

2. The high-precision calibration method based on line array camera images according to claim 1, characterized in that, The hardware mechanism for linear array camera image calibration includes a checkerboard calibration board and multiple linear array cameras; among them, multiple linear array cameras are arranged side by side, and the lenses of the linear array cameras are all installed vertically downward.

3. A high-precision calibration method based on linear array camera images according to claim 1, characterized in that, The process of obtaining the linear array camera images based on the hardware mechanism for linear array camera image calibration is as follows: Place the checkerboard calibration board on two calibration planes respectively; When the checkerboard calibration board is on the two calibration planes, use the linear array camera to collect images of the checkerboard calibration board to obtain the linear array camera images when the checkerboard calibration board is on the two calibration planes.

4. A high-precision calibration method based on linear array camera images according to claim 3, characterized in that, The process of performing distortion calibration on each linear array camera based on the linear array camera images is as follows: Perform X-axis calibration on each linear array camera based on the linear array camera images when the checkerboard calibration board is on the two calibration planes to obtain the reference point X coordinates, X-axis average accuracy, distortion correction coefficient array, working distance, and grid number corresponding to different intervals; Perform Y-axis vertical calibration on each linear array camera based on the linear array camera images when the checkerboard calibration board is on the two calibration planes to obtain the Y-axis vertical field-of-view offset pixels and Y-axis vertical average accuracy corresponding to different intervals.

5. A high-precision calibration method based on linear array camera images according to claim 4, characterized in that, The process of performing X-axis calibration on each linear array camera based on the linear array camera images when the checkerboard calibration board is on the two calibration planes is as follows: Obtain the linear array camera images when the checkerboard calibration board is on Plane 1 and Plane 2, and splice and divide the field of view of the linear array camera images to obtain the X coordinates of the calibration points of the linear array camera in different intervals; When the checkerboard calibration board is on Plane 1, combine the X coordinates of the calibration points of the linear array camera in different intervals, and perform distortion correction on different intervals respectively to obtain the reference point X coordinates, X-axis average accuracy, distortion correction coefficient array, and grid number corresponding to different intervals; When the checkerboard calibration points are on Plane 2, combine the X coordinates of the calibration points of the linear array camera in different intervals, and perform thickness compensation on the structural parameters of the hardware mechanism to obtain the working distance corresponding to different intervals.

6. A high-precision calibration method based on linear array camera images according to claim 5, characterized in that, The process of performing multi-camera joint calibration on multiple linear array cameras based on the linear array camera images is as follows: Obtain the physical length of the checkerboard calibration board on Plane 2 and the projected physical length projected onto Plane 1; Construct the projection relationship of each interval based on the physical length of the checkerboard calibration board on Plane 2, the projected physical length projected onto Plane 1, and the grid number of different intervals; Calculate the relationship between the blind zone width and the height difference between Plane 1 and Plane 2 corresponding to different intervals based on the projection relationship of each interval to obtain the ratio of the blind zone width to the height difference of the calibration plane corresponding to different intervals of the linear array camera.

7. A high-precision measurement method based on linear array camera images, characterized in that, The high-precision measurement method uses the calibration data obtained by the high-precision calibration method based on the linear array camera image as described in any one of claims 1 to 6. The high-precision measurement method includes the following processes: Obtain the height difference between the to-be-detected board and the calibration plane; wherein, the height difference between the to-be-detected board and the calibration plane is the thickness of the to-be-detected board; Predict the coordinates of two pixel points corresponding to the two edges of the to-be-detected board and the intervals corresponding to the two pixel points; Obtain the calibration data of the corresponding interval, and calculate the physical distance between the two pixel points based on the calibration data of the interval; wherein, the physical distance is the width of the to-be-measured board.

8. A high-precision measurement system based on linear array camera images, characterized in that, The system includes: A board thickness acquisition unit, which is used to obtain the height difference between the to-be-detected board and the calibration plane; wherein, the height difference between the to-be-detected board and the calibration plane is the thickness of the to-be-detected board; A placement interval prediction unit, which is used to predict the coordinates of two pixel points corresponding to the two edges of the to-be-detected board and the intervals corresponding to the two pixel points; A board thickness calculation unit, which is used to obtain the calibration data of the corresponding interval, and calculate the physical distance between the two pixel points based on the calibration data of the interval; wherein, the physical distance is the width of the to-be-measured board.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, it implements the high-precision measurement method based on the linear array camera image as described in claim 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, it implements the high-precision measurement method based on the linear array camera image as described in claim 7.

Images