A positioning method based on a vector array target simulator device

Abstract

The application provides a positioning method based on a vector array target simulator device, which measures 13 rows and 13 columns of circle dot templates to calculate average focal length data of a camera, then measures 18 rows and 18 columns of circle dot templates to calculate average values of external rotation parameters of the military camera by combining the camera pixel data into a 9-dimensional array. Then, 100 rows and 100 columns of circle dot templates are used to collect adaptive basic pixel position data, first, a self-adaptive adjustment rule of the translation parameters of the military camera is designed, and the three translation parameters are self-adaptively corrected and calibrated, then nine self-adaptive rules of the rotation parameters are designed, and the nine rotation parameters between the coordinate system of the military camera and the world coordinate system are self-adaptively corrected and calibrated, so that the final nine rotation parameters of the military camera are obtained. Finally, a vector array target composed of two targets is designed, and a double-target positioning position calculation algorithm is designed to calculate the position coordinates of the measured point.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of camera calibration positioning, and in particular, to a vector array target simulator positioning method based on data self-adaption. BACKGROUND

[0002] With the continuous improvement of hardware conditions and computer performance, machine vision technology is widely used in artificial intelligence, visual measurement, robot technology and image processing. Camera is the most important sensor in visual measurement technology, and camera calibration technology is a key step in visual measurement technology. Through the calibration process of the camera, the accurate positional relationship between different coordinate systems can be quickly established, and further the position tracking, size measurement and surface reconstruction of the measured object can be realized. Visual measurement using a camera is not only widely used in civilian fields, but also has a wide range of applications in military and aerospace fields, such as precise non-contact measurement and automatic assembly of cabin segment pose in missile production process, visual measurement and pose solution of large parts of military aircraft, pose measurement and adjustment of fighter parts, and photographic measurement of attitude position information of moving military aircraft in wind tunnel according to target, etc. Traditional camera calibration positioning technology has Zhang's calibration positioning method, direct linear transformation method and Tsai two-step method, which are characterized by calibrating based on a limited number of known points, using less known point information and simple processing and solving. However, the accuracy of the data and the camera model is highly dependent. Based on the above reasons, the present application proposes a military camera calibration positioning method based on a large amount of data self-adaptive calibration, which has low requirements for the camera model, can stably approach the optimal parameters through a large amount of data self-adaptive solving to improve the accuracy of calibration positioning, and the method can continuously adjust and improve the positioning accuracy with the continuous increase and training of data, which is an effect that cannot be achieved by traditional methods.

[0003] It should be noted that the information disclosed in the above background section of the application is only used to strengthen the understanding of the background of the present application, and therefore can include information that does not constitute prior art known to those of ordinary skill in the art. SUMMARY

[0004] The present application aims to provide a positioning method based on a vector array target simulator device, thereby overcoming the problems of low data utilization efficiency and low calibration positioning accuracy caused by the limitations and defects of related technologies.

[0005] According to one aspect of the present application, a positioning method based on a vector array target simulator device is provided, comprising the following steps:

[0006] Step S10, print a template with 13 rows and 13 columns of dots, and place the template at two different known positions, then set the military camera at a special position to meet the condition that the military camera coordinate system coincides with the world coordinate system, record the three-dimensional coordinates of the dot centers on the template in the world coordinate system and store them as an array; then take pictures of the template at the above two positions using the military camera, and detect the obtained images to detect the two-dimensional coordinates of the pixel positions of all dot centers and store them as an array; and calculate the initial value of the focal length parameter according to the above data array, and further solve the average focal length parameter of the military camera.

[0007] Step S20, print 18 rows and 18 columns of template paper and paste it on a plane, place the military camera at a fixed working position, and preliminarily determine the translation parameters between the military camera coordinate system and the world coordinate system by measurement; then place the template at two known positions, obtain the coordinate data of the dots on the template in the world coordinate system and store them as an array; then take pictures of the template at the above two positions using the military camera, and detect the obtained images to detect the pixel position coordinate data of all dot centers and store them as an array; calculate the external parameter column vector matrix and the external parameter related matrix according to the above world coordinate system data and pixel data and the average focal length parameter of the military camera, and solve its left inverse matrix and the external parameter matrix; then average the calculated external parameters to obtain the average value of the external rotation parameters of the military camera.

[0008] Step S30, print 100 rows and 100 columns of template paper and paste it on a plane, place the military camera at a fixed working position, place the template at ten known positions, obtain the coordinate data of the dots on the template in the world coordinate system and store them as an array, and record them as parameter adaptive coordinate data; then take pictures of the template at the above two positions using the military camera, and detect the obtained images to detect the pixel position coordinate data of all dot centers and store them as an array, and obtain the parameter adaptive pixel data.

[0009] Step S40, according to the parameter adaptive coordinate data and the parameter adaptive pixel data, design a military camera translation parameter adaptive adjustment rule to adaptively correct and calibrate the translation parameters between the military camera coordinate system and the world coordinate system, thereby obtaining the final three military camera external translation parameters.

[0010] Step S50, according to the average value of the military camera external parameters, the final three military camera external translation parameters, and the adaptive coordinate data and the parameter adaptive pixel data, design nine rotation parameter adaptive rules respectively to adaptively correct the nine rotation parameters between the military camera coordinate system and the world coordinate system, thereby obtaining the final nine rotation parameters of the military camera.

[0011] Step S60, design left and right staggered diagonal type 2 target, consisting of vector array target. Using probe contact measured point position, and then using military camera to take pictures of the target, and detect the two target in the picture pixel position, and then design double target positioning position solution algorithm, to solve the measured point position coordinates.

[0012] In an example embodiment of the application, a template with 13 rows and 13 columns of dots is printed and placed in two different known positions, then set the military camera at a special position, meet the condition of the military camera coordinate system and the world coordinate system coincide, record the position of the dot center in the world coordinate system three-dimensional coordinates and record as an array; again using the military camera to take pictures of the template in the above two positions, and detect the obtained image, detect the two-dimensional coordinates of the pixel position of all dot centers, also record as an array; and according to the above data array, solve the focal length parameter initial value, further solve the average focal length parameter of the military camera, including:

[0013]

[0014] Where f is the average focal length parameter of the military camera, c x , c y is the actual size of each pixel. a0, b0 represents the center of the image coordinate system in the pixel position coordinate of the pixel coordinate system, the above parameters c x , c y , a0, b0 are related to the resolution of the image, which are constant parameters. f1(j) and f2(j) are the initial value of the focal length parameter. a 31 (k), a 32 (k), a 33 (k) is the three-dimensional coordinate data of the template dot, b 31 (m), b 32 (m) is the two-dimensional coordinate data of the pixel position of the template dot, and its acquisition process is as follows: first, print a template with 13 rows and 13 columns of dots and paste the template on the plane. The size of the template is 100 cm long and 100 cm wide. Second, set the military camera at a special position, meet the condition of the military camera coordinate system and the world coordinate system coincide position, then place the template in the first different known position, get the coordinates of the template in the world coordinate system, record the position of the dot in the world coordinate system as a 11 (i,j), a 12 (i,j), a 13 (i,j), where i = 1, 2, …, 13, j = 1, 2, …, 13, a 11 (i,j) is the x-axis coordinate of the dot center position of the template in the world coordinate system, a 12(i,j) is the x-axis coordinate of the center of the circle dot in the i-th row and j-th column of the template in the world coordinate system, a 13 (i,j) is the z-axis coordinate of the center of the circle dot in the i-th row and j-th column of the template in the world coordinate system. Then, the template is placed in a second different known position, the coordinates of the above template in the world coordinate system are obtained, and the coordinates of the positions of the points on the template in the world coordinate system are recorded as a 21 (i,j), a 22 (i,j), a 23 (i,j), a 21 (i,j) is the x-axis coordinate of the center of the circle dot in the i-th row and j-th column of the template in the world coordinate system, a 22 (i,j) is the y-axis coordinate of the center of the circle dot in the i-th row and j-th column of the template in the world coordinate system, a 23 (i,j) is the z-axis coordinate of the center of the circle dot in the i-th row and j-th column of the template in the world coordinate system. Then, the above two two-dimensional arrays a 11 (i,j), a 21 (i,j) are merged and connected as a one-dimensional array a 31 (k), where k = 1, 2, …, 338. a 12 (i,j), a 22 (i,j) are merged and connected as a one-dimensional array a 32 (k), a 13 (i,j), a 23 (i,j) are merged and connected as a one-dimensional array a 33 (k). Again, the template at the above two positions is photographed by a military camera, and the obtained image is detected to detect the pixel positions of all the circle dot centers, and the pixel position data are one-to-one corresponding to a 31 (k). The x-coordinate of the pixel position data is recorded in a one-dimensional array b 31 (m), m = 1, 2, …, 338, and the y-coordinate of the pixel position data is recorded in a one-dimensional array b 32 (m).

[0015] In an example embodiment of the present application, a 18x18 template paper is printed and pasted on a plane, a military camera is placed in a fixed working position, and translation parameters between the military camera coordinate system and the world coordinate system are preliminarily determined by measurement; then the template is placed at two known positions, coordinate data of the dots on the template in the world coordinate system are obtained and stored as an array; then the template at the two positions is photographed by the military camera, and the obtained image is detected to detect pixel position coordinate data of the centers of all the dots and store the data as an array; external parameter column vector matrix and external parameter correlation matrix are solved according to the world coordinate system data and pixel data and the average focal length parameter of the military camera, a left inverse matrix thereof is solved, and external parameter matrix parameters are solved; then the solved external parameters are summarized and averaged to obtain the average value of the external rotation parameters of the military camera, including:

[0016]

[0017] wherein f is the average focal length parameter of the military camera, c x , c y is the actual size of each pixel. a0, b0 represent the center of the image coordinate system in the pixel position coordinate of the pixel coordinate system, j1=1, 2, 3, 4, 5, 6, 7, 8, 9.

[0018]

[0019] When i1=1,

[0020]

[0021] wherein j1=1, 2, 3, 4, 5, 6, 7, 8, 9, and l1=1, 2, 3, 4, 5, 6, 7, 8, 9.

[0022]

[0023]

[0024] When i1=2, 3, …, 72, W i (l1, j1) is constructed in the same way. a (1)=(W1(l1, j1)) -1 ; wherein W a (1) is the left inverse matrix of W1(l1, j1). When i1=1,

[0025]

[0026] When i1=2, 3, …, 72,

[0027]

[0028] where W a (i) = (W i (l1,j1)) -1 , which is also the left inverse matrix. If the inverse matrix of W i (l1,j1) does not exist, select the next i1 = 2, 3… to calculate the next group of data until all 72 groups of data are calculated.

[0029]

[0030] where r a1 , r a2 , r a3 , r a4 , r a5 , r a6 , r a7 , r a8 , r a9 are the average values of the nine external rotation parameters of the military camera. Where t x0 , t y0 , t z0 and the two-dimensional array d 32a (i1,j1), d 33a (i1,j1), g 31a (i1,j1), g 32a (i1,j1) are obtained as follows: First, print a template with 18 rows and 18 columns of dots and paste it on a plane, and the template size is 100 cm in length and width. Place the military camera in a fixed working position, and preliminarily determine the translation parameters between the camera coordinate system and the world coordinate system, denoted as t x0 , t y0 , t z0 , which is a preliminary data and does not need to be accurately measured and will be corrected later. Then place the template in the first different known position, obtain the coordinates of the template in the world coordinate system, and record the coordinates of the points on the template in the world coordinate system as d 11 (i,j), d 12 (i,j), d 13 (i,j), where i = 1, 2, …, 18, j = 1, 2, …, 18, d 11 (i,j), d 12 (i,j) d 13 (i,j) are the x, y, z axis coordinates of the center of the circle of the i-th row and j-th column of the template in the world coordinate system. Then place the template in the second different known position, obtain the coordinates of the template in the world coordinate system, and record the coordinates of the points on the template in the world coordinate system as d 21 (i,j), d 22(i,j), d 23 (i,j), d 21 (i,j), d 22 (i,j), d 23 (i,j) are the x, y, z axis coordinates of the center of the circle in the world coordinate system, respectively. Then the two two-dimensional arrays d 11 (i,j), d 21 (i,j) are merged into one one-dimensional array d 31 (k), where k = 1, 2, …, 648. The d 12 (i,j), d 22 (i,j) are merged into one one-dimensional array d 32 (k). The d 13 (i,j), d 23 (i,j) are merged into one one-dimensional array d 33 (k). Finally, the templates at the above two positions are photographed by a military camera, and the obtained images are detected to detect the pixel positions of all the circle centers, which are one-to-one corresponding to d 31 (k), the x coordinates of the pixel position data are recorded in a one-dimensional array g 31 (m), m = 1, 2, …, 648, and the y coordinates of the pixel position data are recorded in a one-dimensional array g 32 (m). Then the above data d 31 (k), d 32 (k), d 33 (k), g 31 (m), g 32 (m) are grouped, for example, d 31 (k) are selected 9 by 9 to obtain a two-dimensional array d 31a (i1,j1), where j1 = 1, 2, 3, 4, 5, 6, 7, 8, 9, and i1 = 1, 2, 3, …, 72. Similarly, a two-dimensional array d 32a (i1,j1), d 33a (i1,j1), g 31a (i1,j1), g 32a (i1,j1) is obtained.

[0031] In an example embodiment of the present application, a 100 row by 100 column template paper is printed and pasted on a plane, a military camera is placed in a fixed working position, the template is placed in ten known positions, the coordinates of the dots on the template in the world coordinate system are obtained and stored as an array, denoted as parameter adaptive coordinate data; then the template at the above two positions is photographed using the military camera, and the obtained image is detected to detect the pixel position coordinate data of the center positions of all the dots and store them as an array to obtain parameter adaptive pixel data, which includes: first, a 100 row by 100 column dot template is printed and pasted on a plane, and the template has a size of 100 cm in length and width; second, the military camera is placed in a fixed working position, and then the template is placed in a first different known position to obtain the coordinates of the template in the world coordinate system, and the coordinates of the positions of the dots on the template in the world coordinate system are recorded as e 11 (i,j), e 12 (i,j), e 13 (i,j), e 11 (i,j) is the x-axis coordinate of the center position of the dot in the template i row j column in the world coordinate system, e 12 (i,j) is the y-axis coordinate of the center position of the dot in the template i row j column in the world coordinate system, e 13 (i,j) is the z-axis coordinate of the center position of the dot in the template i row j column in the world coordinate system; third, the template is placed in a second different known position to obtain the coordinates of the template in the world coordinate system, and the coordinates of the positions of the dots on the template in the world coordinate system are recorded as e 21 (i,j), e 22 (i,j), e 23 (i,j), e 21 (i,j) is the x-axis coordinate of the center position of the dot in the template i row j column in the world coordinate system, e 22 (i,j) is the y-axis coordinate of the center position of the dot in the template i row j column in the world coordinate system, e 23 (i,j) is the z-axis coordinate of the center position of the dot in the template i row j column in the world coordinate system; third, the template is placed in a second different known position to obtain the coordinates of the template in the world coordinate system, and the coordinates of the positions of the dots on the template in the world coordinate system are recorded as e 31 (i,j), e 101 (i,j), e 32 (i,j), e 102 (i,j), e 33 (i,j), e 103 (i,j) data; thereafter, the above two two-dimensional arrays e 11 (i,j), e 21(i,j) until e 101 (i,j) are merged into one one-dimensional array f 31 (k), where k = 1, 2, …, 100000. e 12 (i,j), e 22 (i,j) until e 102 (i,j) are merged into one one-dimensional array f 32 (k), e 13 (i,j), e 23 (i,j) until e 103 (i,j) are merged into one one-dimensional array f 33 (k), the above f 31 (k), f 32 (k), f 33 (k) are recorded as parameter adaptive coordinate data; finally, the templates of the above ten positions are photographed by using a military camera, and the pixel positions of all the circular dot centers are detected from the obtained images, and are one-to-one corresponding with f 31 (k), the x coordinates of the pixel position data are recorded in one-dimensional array h 31 (m), m = 1, 2, …, 100000, and the y coordinates of the pixel position data are recorded in one-dimensional array h 32 (m), h 31 (m), h 32 (m) are recorded as parameter adaptive pixel data.

[0032] In an example embodiment of the present application, according to the parameter adaptive coordinate data and the parameter adaptive pixel data, a military camera translation parameter adaptive adjustment rule is designed, the translation parameters between the military camera coordinate system and the world coordinate system are adaptively corrected and revised, so that the final three military camera external translation parameters are obtained, which include:

[0033] e(n) = [h 31a (n) - a0+ h 32a (n) - b0](r a7 f 31 (n) + r a8 f 32 (n) + r a9 f 33 (n) + t z (n)

[0034]

[0035] where f 31 (k), f 32 (k), f 33 (k) are parameter adaptive coordinate data, h31a (m), h 32a (m) is parameter adaptive pixel data, k w10 , k w11 , k w12 is a constant adaptive parameter, t x The initial value of t x0 is selected as t y measured by measurement; and the initial value of t y0 is selected as t z measured by measurement; the initial value of t x0 is selected as t x measured by measurement; the initial value of t y , t z is selected as t xf , t yf and t zf , and is recorded as the final three military camera external translation parameters obtained.

[0036] In an example embodiment of the application, according to the average of the military camera external rotation parameters, the final three military camera external translation parameters, the adaptive coordinate data and the parameter adaptive pixel data, nine rotation parameter adaptive rules are designed respectively to adaptively correct the nine rotation parameters between the military camera coordinate system and the world coordinate system, so as to obtain the final nine rotation parameters of the military camera, which include:

[0037]

[0038] Where f 31 (k), f 32 (k), f 33 (k) is parameter adaptive coordinate data, h 31a (m), h 32a (m) is parameter adaptive pixel data, and r a1t The initial values of r

[0039] r a1t (1) = r a1 ; r a2t (1) = r a2 ; r a3t (1) = r a3 ;

[0040] r a4t (1) = r a4 ; r a5t (1) = r a5 ; r a6t (1) = r a6 ;

[0041] r a7t (1)=r a7 ;r a8t (1)=r a8 ;r a9t (1)=r a9 ;

[0042]

[0043]

[0044] wherein k w1 , k w2 , k w3 , k w4 , k w5 , k w6 , k w7 , k w8 , k w9 are constant adaptive parameters, and r a1t , r a2t , r a3t , r a4t , r a5t , r a6t , r a7t , r a8t , r a9t obtained after the above iterative calculation are respectively denoted as r a1f , r a2f , r a3f , r a4f , r a5f , r a6f , r a7f , r a8f , r a9f , and are denoted as the final nine military camera external rotation parameters obtained.

[0045] In an example embodiment of the present application, two targets of a diagonal type are designed to be staggered left and right to form a vector array target. A probe is used to contact the position of the measured point, and then a military camera is used to take a photo of the target, and the pixel positions of the two targets in the photo are detected, and then a double-target positioning position solving algorithm is designed to solve the coordinates of the measured point position, which comprises:

[0046]

[0047] c1=-(a0r a7f +mr a1f );d1=-(b0r a7f +nr a4f );

[0048] c2=-(a0r a8f+mr a2f )d2= - (b0r a8f +nr a5f )

[0049] c3= - (a0r a9f +mr a3f )d3= - (b0r a9f +nr a6f )

[0050] c4= -a0t zf -mt xf d4= -b0t zf -nt yf

[0051]

[0052] l1= -c4-u1t zf -u1r a7f x1-u1r a8f y1-u1r a9f z1-c1x1-c2y1-c3z1;

[0053] l2= -d4-v1t zf -v1r a7f x1-v1r a8f y1-v1r a9f z1-d1x1-d2y1-d3z1;

[0054] l3= -c4-u1t zf -u1r a7f x2-u1r a8f y2-u1r a9f z2-c1x2-c2y2-c3z2;

[0055] l4= -d4-v1t zf -v1r a7f x2-v1r a8f y2-v1r a9f z2-d1x2-d2y2-d3z2;

[0056]

[0057] W n1 = M n1 L1; W n2 = M n2 L2;

[0058] The target design and the data acquisition mode of u1, v1 and u2, v2 and x1, y1, z1, x2, y2, z2 are as follows: two diagonal targets are designed to be staggered left and right to form a vector array target; the target is mainly composed of a non-modulated high-power emission source (light source) and a heat sink, and the heat sink is made of copper-aluminum composite material. The function is to radiate light pulses that meet the shooting requirements of military cameras, provide sufficient light sources for military camera shooting, and clearly identify the target pixel position in the photo. The total size of the target is 10mmx10mm; the total size of a single target is 10mmx10mm, the center wavelength is 625nm, it is suitable for a distance of less than or equal to 10m, the beam width of the LED lamp is 60°, the divergence angle is 120°, and the average emission power is not less than 1.6W. When measuring, the preset measured point coordinates are x a , y a , z a , and the coordinates of the two targets relative to the measured point group are marked as x a +x1, y a +y1, z a +z1 and x a +x2, y a +y2, z a +z2. x1, y1, z1, x2, y2, z2 are constant values obtained according to the relative position between the target and the probe. Then a military camera is used to take a photo of the target, and the pixel positions of the two targets in the photo are detected and marked as u1, v1 and u2, v2. m, n are intermediate variables, c x , c y are the actual sizes of each pixel. a0, b0 represent the pixel position coordinates of the center of the image coordinate system in the pixel coordinate system, r a1f , r a2f , r a3f , r a4f , r a5f , r a6f , r a7f , r a8f , r a9f , and the obtained final nine military camera external rotation parameters are denoted as; L1, L2 are coordinate column vectors calculated according to the above intermediate variables and target pixel position data; W1, W2 are coordinate matrices calculated according to the above intermediate variables and target pixel position data. W n1 and W n2 are the left inverse matrices of the coordinate matrices W1, W2, A n1 and A n2 are the first and second estimated values of the measured point coordinates, and A n is the measured point coordinates, which is a three-row-one-column column vector, and the elements are the three-dimensional coordinates x a , y a, z a , and then the three-dimensional position coordinates of the measured point can be solved by solving A n , and then the three-dimensional position coordinates of the measured point can be solved by solving A

[0059] Advantages

[0060] The positioning method based on the vector array target simulator device provided by the application has the following two main innovations: first, a method for self-adapting correction and revision of the external translation parameters and nine rotation parameters of the military camera based on a large amount of known calibration data is proposed, the method has the advantage of absorbing the advantages of big data, and the precision of the calibration of the military camera is continuously improved by using massive data, of course, compared with the traditional positioning calculation method, the calculation amount and the calculation complexity are greatly increased, but for the current computer software and hardware level and the calculation speed, it is not a problem at all. Second, a military camera position calculation method based on double-target positioning is proposed, in the case of high calibration precision, the double-target positioning calculation method has the advantages of simplicity, feasibility and reliable precision, of course, if the conditions permit, the method can also be extended to multi-target positioning method, and the precision will be further improved.

[0061] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the application. BRIEF DESCRIPTION OF DRAWINGS

[0062] The accompanying drawings, which are incorporated into and form part of the specification, illustrate embodiments consistent with the application and, together with the specification, serve to explain the principles of the application. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained from these drawings without creative labor for those skilled in the art.

[0063] Figure 1 is a flow chart of the positioning method based on the vector array target simulator device provided by the application;

[0064] Figure 2 is a 13-row 13-column setting printing template diagram of the method provided by the embodiment of the application;

[0065] Figure 3 is a 100-row 100-column setting printing template electronic diagram of the method provided by the embodiment of the application;

[0066] Figure 4 is a simple structure diagram of the target of the method provided by the embodiment of the application;

[0067] Figure 5 is a simple position diagram of the vector array target of the method provided by the embodiment of the application;

[0068] Figure 6 is the first data self-adapting calibration chart of the external rotation parameters of a military camera provided by the method of the embodiment of the present application;

[0069] Figure 7 is the second data self-adapting calibration chart of the external rotation parameters of a military camera provided by the method of the embodiment of the present application;

[0070] Figure 8 is the eighth data self-adapting calibration chart of the external rotation parameters of a military camera provided by the method of the embodiment of the present application;

[0071] Figure 9 is the ninth data self-adapting calibration chart of the external rotation parameters of a military camera provided by the method of the embodiment of the present application. DETAILED DESCRIPTION

[0072] Example implementations will now be described more fully with reference to the accompanying drawings. Example implementations can be implemented in any

[0073] The present application provides a method of calibrating a military camera based on data self-adaptation and positioning calculation using double-target light emission. First, the average focal length data of the camera is calculated by measuring a 13-row 13-column dot template, then the average value of the external rotation parameters of the military camera is calculated by measuring an 18-row 18-column dot template and the pixel data of the camera to form a 9-dimensional array. Then, the basic pixel position data for parameter self-adaptation is collected by using a 100-row 100-column dot template. First, the self-adaptation adjustment rule of the translation parameters of the military camera is designed, the basic pixel position data is used to adaptively correct and calibrate the three translation parameters, then the self-adaptation rule of the nine rotation parameters is designed, the nine rotation parameters between the coordinate system of the military camera and the world coordinate system are adaptively corrected and calibrated, so as to obtain the final nine rotation parameters of the military camera. Finally, a vector array target composed of two targets is designed, and a double-target positioning position calculation algorithm is designed to calculate the position coordinates of the measured point position coordinates.

[0074] The positioning method based on a vector array target simulator device of the present application will be further explained and described below in combination with the accompanying drawings. Referring to Figure 1 The method can comprise the following steps:

[0075] In step S10, a template provided with 13 rows and 13 columns of dots is printed, and the template is placed at two different known positions. Then, a military camera is set at a special position to meet the condition that the military camera coordinate system coincides with the world coordinate system. The positions of the dot centers on the template in the three-dimensional coordinates of the world coordinate system are recorded and stored as an array. Then, the template at the above two positions is photographed by using the military camera, and the obtained images are detected to detect the two-dimensional coordinates of the pixel positions of all the dot centers, which are also recorded and stored as an array. The preliminary value of the focal length parameter is calculated according to the above data array, and the average focal length parameter is further solved.

[0076] Specifically, a template as shown in Figure 2 is printed first, and the template is pasted on a plane. The template is provided with 13 rows and 13 columns of dots, and the size of the template is 100 cm in length and width.

[0077] Secondly, the military camera is set at a special position to meet the condition that the military camera coordinate system coincides with the world coordinate system. Then, the template is placed at a first different known position to obtain the coordinates of the template in the world coordinate system, and the positions of the points on the template in the world coordinate system are recorded as a 11 (i,j), a 12 (i,j), a 13 (i,j), a 11 (i,j) is the x-axis coordinate of the dot center position of the i-th row and the j-th column of the template in the world coordinate system, a 12 (i,j) is the y-axis coordinate of the dot center position of the i-th row and the j-th column of the template in the world coordinate system, a 13 (i,j) is the z-axis coordinate of the dot center position of the i-th row and the j-th column of the template in the world coordinate system. Then, the template is placed at a second different known position to obtain the coordinates of the template in the world coordinate system, and the positions of the points on the template in the world coordinate system are recorded as a 21 (i,j), a 22 (i,j), a 23 (i,j), a 21 (i,j) is the x-axis coordinate of the dot center position of the i-th row and the j-th column of the template in the world coordinate system, a 22 (i,j) is the y-axis coordinate of the dot center position of the i-th row and the j-th column of the template in the world coordinate system, a 23(i,j) is the z-axis coordinate of the dot center position of the i-th row and j-th column of the template in the world coordinate system. Then, the two two-dimensional arrays a 11 (i,j), a 21 (i,j) are merged and connected into a one-dimensional array a 31 (k), where k = 1, 2, …, 338. The a 12 (i,j), a 22 (i,j) are merged and connected into a one-dimensional array a 32 (k), the a 13 (i,j), a 23 (i,j) are merged and connected into a one-dimensional array a 33 (k).

[0078] Again, the templates at the above two positions are photographed by using the military camera, and the obtained images are detected to detect the pixel positions of all dot centers, which are one-to-one corresponding to a 31 (k). The x-coordinate of the pixel position data is recorded in a one-dimensional array b 31 (m), m = 1, 2, …, 338, and the y-coordinate of the pixel position data is recorded in a one-dimensional array b 32 (m).

[0079] Then, the following calculation formula is used to solve the preliminary values f1(j) and f2(j) of the focal length parameters as follows:

[0080]

[0081] where c x and c y are the actual size of each pixel. a0 and b0 represent the pixel position coordinates of the center of the image coordinate system in the pixel coordinate system. The above parameters c x , c y , a0 and b0 are related to the resolution of the image and are constant parameters.

[0082] Then, the average value of the preliminary values of the focal length parameters is taken as the average focal length parameter as follows:

[0083]

[0084] where f is the average focal length parameter of the military camera to be solved.

[0085] Step S20, print 18 rows and 18 columns of template paper and paste it on a plane, place the military camera in a fixed working position, preliminarily determine the translation parameters between the military camera coordinate system and the world coordinate system by measurement; then place the template at two known positions, obtain the coordinate data of the dots on the template in the world coordinate system and store them as an array; then take pictures of the template at the above two positions by using the military camera, detect the obtained images, detect the pixel position coordinate data of the centers of all the dots and store them as an array; solve the external parameter column vector matrix and the external parameter related matrix according to the above world coordinate system data and pixel data, obtain the left inverse matrix thereof, and solve the external parameter matrix; then average the obtained external parameters to obtain the average value of the nine external rotation parameters of the military camera.

[0086] Specifically, it is divided into the following 10 small steps. First, print a template and paste the template on a plane. The template is provided with 18 rows and 18 columns of dots, and the size of the template is 100 cm in length and width.

[0087] Second, place the military camera in a fixed working position, preliminarily determine the translation parameters between the military camera coordinate system and the world coordinate system by measurement, and record them as t x0 , t y0 , t z0 . The data is preliminary data, which does not need to be measured accurately and will be corrected later. Then, place the template at a first different known position, obtain the coordinates of the template in the world coordinate system, and record the coordinates of the points on the template in the world coordinate system as d 11 (i,j), d 12 (i,j), d 13 (i,j), where i = 1, 2, …, 18, j = 1, 2, …, 18, d 11 (i,j) is the x-axis coordinate of the center of the dot in the world coordinate system, d 12 (i,j) is the y-axis coordinate of the center of the dot in the world coordinate system, and d 13 (i,j) is the z-axis coordinate of the center of the dot in the world coordinate system.

[0088] Third, place the template at a second different known position, obtain the coordinates of the template in the world coordinate system, and record the coordinates of the points on the template in the world coordinate system as d 21 (i,j), d 22 (i,j), d 23 (i,j), d 21 (i,j) is the x-axis coordinate of the center of the dot in the world coordinate system, d 22(i,j) is the y-axis coordinate of the center of the circle dot in the world coordinate system in the i-th row and j-th column of the template, d 23 (i,j) is the z-axis coordinate of the center of the circle dot in the world coordinate system in the i-th row and j-th column of the template.

[0089] Fourthly, the two two-dimensional arrays d 11 (i,j), d 21 (i,j) are merged and connected as a one-dimensional array d 31 (k), where k = 1, 2, …, 648. d 12 (i,j), d 22 (i,j) are merged and connected as a one-dimensional array d 32 (k), d 13 (i,j), d 23 (i,j) are merged and connected as a one-dimensional array d 33 (k).

[0090] Fifthly, the templates at the above two positions are photographed by a military camera, and the obtained images are detected to detect the pixel positions of all the circle dot centers, which are one-to-one corresponding to d 31 (k), the x-coordinate of the pixel position data is recorded in a one-dimensional array g 31 (m), m = 1, 2, …, 648, and the y-coordinate of the pixel position data is recorded in a one-dimensional array g 32 (m).

[0091] Sixthly, the above data d 31 (k), d 32 (k), d 33 (k), g 31 (m), g 32 (m) are grouped, for example, d 31 (k) are selected as a group of 9, to obtain a two-dimensional array d 31a (i1,j1), where j1 = 1, 2, 3, 4, 5, 6, 7, 8, 9, and i1 = 1, 2, 3, …, 72. Similarly, a two-dimensional array d 32a (i1,j1), d 33a (i1,j1), g 31a (i1,j1), g 32a (i1,j1) is obtained.

[0092] Seventhly, the above two-dimensional arrays are firstly calculated to obtain an external parameter column vector matrix Y(i1,j1), where i1 = 1, Y1(1,j1) is a column vector formed by the matrix Y(i1,j1) when i1 = 1, and the calculation is as follows:

[0093]

[0094] wherein

[0095]

[0096] j1 = 1, 2, 3, 4, 5, 6, 7, 8, 9;

[0097] The eighth step, taking i1 = 1 as an example, the external parameter correlation matrix W1(l1,j1) is calculated, using d 31a (1,j1), d 32a (1,j1), d 33a (i1,j1), g 31a (1,j1), g 32a (1,j1) to calculate W1(l1,j1), wherein j1 = 1, 2, 3, 4, 5, 6, 7, 8, 9, and l1 = 1, 2, 3, 4, 5, 6, 7, 8, 9. The elements of W1(l1,j1) are as follows:

[0098]

[0099] When i1 = 2, 3, …, 72, W i (l1,j1) is constructed in the same way. Wherein

[0100]

[0101]

[0102] The ninth step is to find the left inverse matrix of W1(l1,j1), denoted as W a (1) = (W1(l1,j1)) -1 The external generation matrix parameters are calculated as follows, wherein i1 = 1:

[0103]

[0104] And for i1 = 2, 3, …, 72,

[0105]

[0106] Wherein W a (i) = (W i (l1,j1)) -1 , which is also a left inverse matrix.

[0107] If the inverse matrix of W i (l1,j1) does not exist, the next i1 = 2, 3, … is selected to calculate the next group of data until all 72 groups of data are calculated.

[0108] The tenth step is to collect and average the nine external rotation parameters of the military camera 9 to obtain the average of the external parameters as follows:

[0109]

[0110]

[0111] wherein r a1 , r a2 , r a3 , r a4 , r a5 , r a6 , r a7 , r a8 , r a9 are the average values of the nine external rotation parameters of the military camera.

[0112] In step S30, a 100*100 template paper is printed and pasted on a plane, the military camera is placed in a fixed working position, the template is placed at ten known positions, the coordinates of the dots on the template in the world coordinate system are obtained and stored as an array, denoted as parameter adaptive coordinate data; then the template at the above two positions is photographed by using the military camera, and the pixel position coordinate data of the center positions of all the dots on the obtained image are detected and stored as an array, denoted as parameter adaptive pixel data.

[0113] Specifically, the following six sub-steps are included. In the first step, a template is printed and pasted on a plane. The template is provided with 100*100 dots, and the size of the template is 100*100 cm.

[0114] In the second step, the military camera is placed in a fixed working position, and then the template is placed at a first different known position to obtain the coordinates of the template in the world coordinate system, and the coordinates of the dots on the template in the world coordinate system are recorded as e 11 (i,j), e 12 (i,j), and e 13 (i,j), wherein i=1, 2, …, 100, j=1, 2, …, 100, e 11 (i,j) is the x-axis coordinate of the center position of the dot in the world coordinate system, e 12 (i,j) is the y-axis coordinate of the center position of the dot in the world coordinate system, and e 13 (i,j) is the z-axis coordinate of the center position of the dot in the world coordinate system.

[0115] In the third step, the template is placed at a second different known position to obtain the coordinates of the template in the world coordinate system, and the coordinates of the dots on the template in the world coordinate system are recorded as e21 (i,j), e 22 (i,j), e 23 (i,j), e 21 (i,j) represents the x-coordinate of the center of the circle in the i-th row and j-th column of the template in the world coordinate system. 22 (i,j) represents the y-axis coordinate of the center of the circle in the i-th row and j-th column of the template in the world coordinate system. 23 (i,j) represents the z-axis coordinate of the center of the circle in the i-th row and j-th column of the template in the world coordinate system. Using the same method, the template is placed sequentially at the third to tenth different known positions to obtain e. 31 (i,j),…,e 101 (i,j), e 32 (i,j),…,e 102 (i,j), e 33 (i,j),…,e 103 (i,j) data.

[0116] The fourth step is to combine the two two-dimensional arrays e mentioned above. 11 (i,j), e 21 (i,j) up to e 101 (i,j) are merged and concatenated into a one-dimensional array f. 31 (k), where k = 1, 2, ..., 100000. Let e 12 (i,j), e 22 (i,j) up to e 102 (i,j) are merged and concatenated into a one-dimensional array f. 32 (k), e 13 (i,j), e 23 (i,j) up to e 103 (i,j) are merged and concatenated into a one-dimensional array f. 33 (k), the above f 31 (k), f 32 (k), f 33 (k) is denoted as the parameter adaptive coordinate data.

[0117] The fifth step involves using a military camera to photograph the templates at the ten locations mentioned above, and then performing detection on the obtained images to determine the pixel positions of the center positions of all dots, and comparing them with f. 31 (k) One-to-one correspondence, recording the x-coordinate of the pixel position data in a one-dimensional array h 31a In array (m), m = 1, 2, ..., 100000, the y-coordinates of the pixel position data are recorded in a one-dimensional array h. 32a In (m), h 31a (m), h 32a (m) is denoted as parameter adaptive pixel data.

[0118] Step S40, according to the parameter adaptive coordinate data and the parameter adaptive pixel data, a parameter adaptive adjustment rule of the military camera translation parameter is designed, the translation parameter between the military camera coordinate system and the world coordinate system is adaptively corrected and revised, and finally three military camera external translation parameters are obtained.

[0119] Specifically, the following 6 sub-steps are included. First, the x-direction translation parameter t x between the military camera coordinate system and the world coordinate system is adaptively corrected, and the initial value of t x is selected as t x0 measured in step two; the y-direction translation parameter t x between the military camera coordinate system and the world coordinate system is adaptively corrected, and the initial value of t y is selected as t y0 measured in step two; the z-direction translation parameter t z between the military camera coordinate system and the world coordinate system is adaptively corrected, and the initial value of t z is selected as t x0

[0120] Second, the error data e(n) of the current point is calculated as follows:

[0121]

[0122] Third, the parameter adaptive correction rule of the x-direction translation parameter t x is designed as follows:

[0123]

[0124] Where k w10 is a constant adaptive parameter, and its selection is described in the following case implementation.

[0125] Fifth, the parameter adaptive correction rule of the y-direction translation parameter t y is designed as follows:

[0126]

[0127] Where k w11 is a constant adaptive parameter, and its selection is described in the following case implementation.

[0128] Sixth, the parameter adaptive correction rule of the z-direction translation parameter t z is designed as follows:

[0129]

[0130] where k w12 is a constant adaptive parameter, which is selected as follows.

[0131] After the above iterative calculation, the final t x , t y , and t z are denoted as t xf , t yf , and t zf , respectively, and are denoted as the final three military camera external translation parameters.

[0132] Step S50, according to the three military camera external translation parameters, adaptive coordinate data, and parameter adaptive pixel data, nine military camera rotation parameter adaptive rules are designed respectively to adaptively correct the nine rotation parameters between the military camera coordinate system and the world coordinate system, thereby obtaining the final nine rotation parameters of the military camera.

[0133] Specifically, the following two sub-steps are included. First, according to the military camera external translation parameters, the error data e1(n) of the current point is recalculated, and the calculation method is as follows:

[0134]

[0135] where the initial values of r a1t , etc. are selected as the average values of the nine external parameters of the military camera obtained in the second step, as follows:

[0136] r a1t (1) = r a1 ; r a2t (1) = r a2 ; r a3t (1) = r a3 ;

[0137] r a4t (1) = r a4 ; r a5t (1) = r a5 ; r a6t (1) = r a6 ;

[0138] r a7t (1) = r a7 ; r a8t (1) = r a8 ; r a9t (1) = r a9 ;

[0139] Second, the data adaptive correction iteration rule of the nine rotation parameters between the military camera coordinate system and the world coordinate system is as follows:

[0140]

[0141] Where k w1 k w2 k w3 k w4 k w5 k w6 k w7 k w8 k w9 This is a constant adaptive parameter; its selection will be discussed in the case study below.

[0142] The final r obtained after the above iterative calculation a1t r a2t r a3t r a4t r a5t r a6t r a7t r a8t r a9t Let them be denoted as r a1f r a2f r a3f r a4f r a5f r a6f r a7f r a8f r a9f , denoted as the final nine external rotation parameters of the military camera.

[0143] Step S60: Design two diagonally offset targets to form a vector array target. Use a probe to contact the measured point on the military component, then use a military camera to photograph the targets and detect the pixel positions of the two targets in the photograph. Then, design a dual-target positioning algorithm to calculate the coordinates of the measured point.

[0144] Specifically, it is divided into the following 6 steps. Step 1: Design two diagonally offset targets to form a vector array target. The target mainly consists of an unmodulated high-power emission source (light source) and a heat sink, which will be made of copper-aluminum composite material. Its function is to radiate light pulses that meet the requirements of military camera shooting, providing sufficient energy for the camera and ensuring that the target pixel positions are clearly identifiable in the photograph. The total size of the target is 10mm × 10mm, and the overall structural design is shown in the figure below. Figure 4 As shown, the topmost element is a lens. The total size of a single target is 10mm × 10mm, with a center wavelength of 625nm. It is suitable for distances of 10m or less, with an LED beamwidth of 60°, a divergence angle of 120°, and an average emission power of not less than 1.6W. During measurement, assume the coordinates of the measured point are x... a ya , z a , while the two targets' coordinates relative to the group of measured points are marked as x a + x1, y a + y1, z a + z1 and x a + x2, y a + y2, z a + z2. And x1, y1, z1, x2, y2, z2 are constants, which are obtained according to the relative position between the targets and the probe.

[0145] Secondly, take pictures of the targets by using the military camera, and detect the pixel positions of the two targets in the pictures, which are marked as u1, v1 and u2, v2.

[0146] Thirdly, according to the final external rotation parameter of the military camera and the final external translation parameter of the military camera, the following intermediate variables are calculated:

[0147]

[0148] Fourthly, according to the above intermediate variables and the pixel position data of the targets, the coordinate column vectors L1, L2 are calculated.

[0149] Wherein

[0150] l1 = -c4 - u1t zf - u1r a7f x1 - u1r a8f y1 - u1r a9f z1 - c1x1 - c2y1 - c3z1;

[0151] l2 = -d4 - v1t zf - v1r a7f x1 - v1r a8f y1 - v1r a9f z1 - d1x1 - d2y1 - d3z1;

[0152] l3 = -c4 - u1t zf - u1r a7f x2 - u1r a8f y2 - u1r a9f z2 - c1x2 - c2y2 - c3z2;

[0153] l4 = -d4 - v1t zf - v1r a7f x2 - v1r a8f y2 - v1r a9f z2 - d1x2 - d2y2 - d3z2;

[0154] The fifth step, according to the above intermediate variable and target pixel position data to solve the coordinate matrix W1, W2.

[0155]

[0156]

[0157] The sixth step, according to the coordinate column vector and coordinate matrix, the final measured point coordinate data is solved. Wherein the left inverse matrix of the coordinate matrix W1, W2 is solved, respectively recorded as W n1 And W n2 Then the first and second estimates of the final measured point coordinates are solved, respectively recorded as A n1 And A n2 Its calculation method is as follows:

[0158] W n1 =M n1 L1; W n2 =M n2 L2;

[0159] Then the measured point coordinates are calculated, recorded as A n Its calculation method is as follows:

[0160]

[0161] Wherein A n is a column vector of three rows and one column, and the elements are measured point coordinates x a , y a , z a .

[0162] Case implementation and computer simulation results analysis

[0163] In step S10, according to the 13 rows and 13 columns of the printing template as shown in Figure 2 According to the operation, the average value of the focal length parameter is finally obtained as f=0.0197.

[0164] In step S20, according to the operation, the average value of the military camera external parameter is finally obtained

[0165] r a1 =0.991, r a2 =-0.0944, r a3 =-0.0866, r a4 =0.0868, r a5 =0.9924, r a6 =0.0874, r a7 =0.0945, r a8 =0.0789, r a9= 0.9926.

[0166] In step S30, according to the operation, 100 rows and 100 columns of circle dots are printed as shown in Figure 3

[0167] In step S40, according to the operation, k w10 = 0.0009, k w11 = 0.0008, k w11 = 0.0007, three military camera external translation parameters t xf = 25.4967, t yf = 3.5025 and t zf = -2.5178 are finally solved.

[0168] In step S50, according to the operation, k w1 = 0.0005, k w2 = 0.0008, k w3 = 0.0013, k w4 = 0.0007, k w5 = 0.0011, k w6 = 0.0008, k w7 = 0.0006, k w8 = 0.0009, k w9 = 0.0007, nine final military camera external rotation parameters r a1f = 0.988, r a2f = -0.0935, r a3f = -0.0858, r a4 = 0.0857, r a5 = 0.9916, r a6 = 0.0879, r a7 = 0.0936, r a8 = 0.0794, r a9 = 0.9935 are finally obtained. The adaptive calibration of the first military camera external rotation parameter r a1f is shown in Figure 6 . The adaptive calibration of the second military camera external rotation parameter r a2f is shown in Figure 7 . The adaptive calibration of the eighth military camera external rotation parameter r a1f is shown in Figure 8 . The adaptive calibration of the ninth military camera external rotation parameter r a2f is shown in Figure 9 .

[0169] In step S60, a target is designed as shown in Figure 4 ​As shown, then according to left and right staggered diagonal 2 target vector array target as shown Figure 5 As shown. Then the probe contact military aircraft measured point position, can also be fixed with the characteristics of the military aircraft vector array target position, using military camera to take pictures of the target, and detect two target pixel position in the picture, recorded as u1=960.2551, v1=540.0557 and u2=960.2820, v2=540.0622, ultimately calculated to get the measured point coordinates x a =45.0031, y a =7.9968, z a =3.0025. Through the experiment can be seen that the method based on data adaptive can realize convergence in the case of more data, and ultimately achieve steady state value. At the same time, by comparing with the standard value can be seen that the accuracy of the final positioning is very high, and it is worth mentioning that by using more data for calibration, the method proposed in the present application can further improve the accuracy of calibration through adaptive algorithm, and can further improve the accuracy of the final positioning, so that the method provided by the present application not only has high theoretical value, but also has high military engineering practical value. It is worth mentioning that the embodiment of the present application only measures a certain fixed point of the military aircraft, but it can also be applied to the relative position measurement between multiple points, such as measuring the length of a certain component, etc. Since there is no special difficulty in principle, therefore, it is not necessary to enumerate one by one.

Claims

1. A positioning method based on a vector array target simulator device, characterized in that... The following steps are required: Step S10: Print a template with 13 rows and 13 columns of dots, and place the template in two different known positions. Then, set the military camera in a special position to ensure that the military camera coordinate system coincides with the world coordinate system. Record the three-dimensional coordinates of the center position of the dots on the template in the world coordinate system and store them as an array. Next, use the military camera to take pictures of the template at the two positions, and perform detection on the obtained images to detect the two-dimensional coordinates of the pixel positions of all dot centers, which are also recorded as an array. Based on the above data array, calculate the preliminary value of the focal length parameter, and further solve the average focal length parameter of the military camera as follows: Where a 31 (k), a 32 (k), a 33 (k) represents the three-dimensional coordinate data of the template circle point, b 31 (m), b 32 (m) represents the two-dimensional coordinate data of the pixel position of the template dots. The acquisition process is as follows: First, print a template with 13 rows and 13 columns of dots and paste it on a plane; the template size is 100cm in both length and width; second, set the military camera in a specific position to ensure that the military camera coordinate system coincides with the world coordinate system, then place the template in the first different known position, obtain the coordinates of the template in the world coordinate system, and record the coordinates of the points on the template in the world coordinate system as a. 11 (i,j), a 12 (i,j), a 13 (i,j), where i=1,2,…,13, j=1,2,…,13, a 11 (i,j) represents the x-coordinate of the center of the circle in the i-th row and j-th column of the template in the world coordinate system. 12 (i,j) represents the y-axis coordinate of the center of the circle in the i-th row and j-th column of the template in the world coordinate system. 13 (i,j) represents the z-axis coordinate of the center of the circle in the i-th row and j-th column of the template in the world coordinate system; then, the template is placed in a second, different known position, and the coordinates of the template in the world coordinate system are obtained. The coordinates of the point on the template in the world coordinate system are recorded as a. 21 (i,j), a 22 (i,j), a 23 (i,j), a 21 (i,j) represents the x-coordinate of the center of the circle in the i-th row and j-th column of the template in the world coordinate system. 22 (i,j) represents the y-axis coordinate of the center of the circle in the i-th row and j-th column of the template in the world coordinate system. 23 (i,j) represents the z-axis coordinate of the center of the circle in the i-th row and j-th column of the template in the world coordinate system; then, the two two-dimensional arrays a above are... 11 (i,j), a 21 The elements (i,j) are merged and concatenated into a one-dimensional array a. 31 (k), where k = 1, 2, ..., 338; a 12 (i,j), a 22 The elements (i,j) are merged and concatenated into a one-dimensional array a. 32 (k), a 13 (i,j), a 23 The elements (i,j) are merged and concatenated into a one-dimensional array a. 33 (k); Next, a military camera is used to photograph the templates at the two locations mentioned above, and the obtained images are analyzed to detect the pixel positions of the center positions of all dots, and compared with a 31 (k) One-to-one correspondence, recording the x-coordinate of the pixel position data in a one-dimensional array b 31 In (m), m = 1, 2, ..., 338, the y-coordinates of the pixel position data are recorded in a one-dimensional array to obtain b. 32 (m); f is the desired average focal length parameter of the military camera, c x c y The actual size of each pixel; a0 and b0 represent the pixel position coordinates of the center of the image coordinate system in the pixel coordinate system, and the above parameter c x c y a0 and b0 are related to the image resolution and are constant parameters; f1(j) and f2(j) are preliminary values ​​of the focal length parameter. Step S20: Print an 18-row, 18-column template paper and paste it onto a flat surface. Place the military camera in a fixed working position and determine the translation parameters between the military camera coordinate system and the world coordinate system through measurement. Then, place the template at two known locations, obtain the coordinate data of the dots on the template in the world coordinate system, and store them as an array. Next, use the military camera to take pictures of the template at the two locations, and detect the obtained images to obtain the pixel position coordinate data of the center positions of all dots, storing them as an array. Based on the world coordinate system data, pixel data, and the average focal length parameter of the military camera, solve for the external parameter column vector matrix and the external parameter correlation matrix, obtain their left inverse matrix, and solve for the external generation matrix parameters. Then, summarize and average the obtained external parameters to obtain the average value of the military camera's external rotation parameters as follows: Where f is the desired average focal length parameter of the military camera, and c x c y The actual size of each pixel; a0 and b0 represent the pixel position coordinates of the center of the image coordinate system in the pixel coordinate system, j1 = 1, 2, 3, 4, 5, 6, 7, 8, 9; When i1 = 1 Where j1 = 1, 2, 3, 4, 5, 6, 7, 8, 9, l1 = 1, 2, 3, 4, 5, 6, 7, 8, 9; When i1 = 2, 3, ..., 72, W is constructed in the same way. i (l1,j1); W a (1)=(W1(l1,j1)) -1 ;W a (1) is the left inverse matrix of W1(l1,j1); when i1=1 When i1 = 2, 3, ..., 72 Among them W a (i)=(W i (l1,j1)) -1 It is also a left inverse matrix; if W i If the inverse matrix (l1,j1) does not exist, then select the next i1 = 2, 3... to solve the next set of data, until all 72 sets of data have been solved; Where t x0 t y0 t z0 With two-dimensional array d 32a (i1,j1),d 33a (i1,j1),g 31a (i1,j1),g 32a The method for obtaining (i1,j1) is as follows: First, print a template with 18 rows and 18 columns of dots and paste it on a flat surface. The template size is 100cm in both length and width. Place the military camera in a fixed working position and preliminarily determine the translation parameters between the military camera coordinate system and the world coordinate system by measurement, denoted as t. x0 t y0 t z0 This data is preliminary and does not require precise measurement; it will be corrected later. Then, the template is placed at the first different known location, and the coordinates of the template in the world coordinate system are obtained. The coordinates of the points on the template in the world coordinate system are recorded as d. 11 (i,j),d 12 (i,j),d 13 (i,j), where i = 1, 2, ..., 18, j = 1, 2, ..., 18, d 11 (i,j),d 12 (i,j)d 13 (i, j) represent the x, y, and z coordinates of the center of the circle in the i-th row and j-th column of the template, respectively, in the world coordinate system. Then, the template is placed in a second, different known position, and the coordinates of the template in the world coordinate system are obtained. The coordinates of the points on the template in the world coordinate system are recorded as d. 21 (i,j),d 22 (i,j),d 23 (i,j), d 21 (i,j),d 22 (i,j),d 23 (i,j) represent the x, y, and z coordinates of the center of the circle in the i-th row and j-th column of the template, respectively, in the world coordinate system; then, the two two-dimensional arrays d above are... 11 (i,j), d 21 (i,j) are merged and concatenated into a one-dimensional array d. 31 (k), where k = 1, 2, ..., 648; d 12 (i,j), d 22 (i,j) are merged and concatenated into a one-dimensional array d. 32 (k), d 13 (i,j), d 23 (i,j) are merged and concatenated into a one-dimensional array d. 33 (k); Finally, a military camera is used to photograph the templates at the two locations mentioned above, and the obtained images are analyzed to detect the pixel positions of the center positions of all dots, and then compared with d. 31 (k) One-to-one correspondence, recording the x-coordinate of the pixel position data in a one-dimensional array g 31 In (m), m = 1, 2, ..., 648, the y-coordinates of the pixel position data are recorded in a one-dimensional array g. 32 (m); then take the above data d 31 (k), d 32 (k), d 33 (k), g 31 (m), g 32 (m) grouped, such as d 31 (k) Select 9 items in sequence to form a group, and obtain a two-dimensional array d. 31a (i1,j1), where j1=1,2,3,4,5,6,7,8,9, i1=1,2,3,…,72; the two-dimensional array d is obtained in the same way. 32a (i1,j1),d 33a (i1,j1),g 31a (i1,j1),g 32a (i1,j1); r a1 r a2 r a3 r a4 r a5 r a6 r a7 r a8 r a9 This represents the average of nine external rotation parameters of a military camera. Step S30: Print a 100-row, 100-column template paper and paste it onto a flat surface. Place a military camera in a fixed working position and position the template at ten known locations. Obtain the coordinate data of the dots on the template in the world coordinate system and store it as an array, denoted as the parameter adaptive coordinate data. Then, use the military camera to take pictures of the template at the above two positions and perform detection on the obtained images to detect the pixel position coordinate data of the center positions of all dots and store it as an array. The parameter adaptive pixel data is as follows: First, print a dot template with 100 rows and 100 columns and paste it onto a flat surface. The template dimensions are 100cm in both length and width. Second, place a military camera in a fixed working position, and then place the template in a first, different, known position. Obtain the coordinates of the template in the world coordinate system and record the coordinates of the points on the template in the world coordinate system as e. 11 (i,j), e 12 (i,j), e 13 (i,j), where i = 1, 2, ..., 100, j = 1, 2, ..., 100, e 11 (i,j) represents the x-coordinate of the center of the circle in the i-th row and j-th column of the template in the world coordinate system. 12 (i,j) represents the y-axis coordinate of the center of the circle in the i-th row and j-th column of the template in the world coordinate system. 13 (i,j) represents the z-axis coordinate of the center of the circle in the i-th row and j-th column of the template in the world coordinate system. Next, the template is placed in a second, different known position, and the coordinates of the template in the world coordinate system are obtained. The position of the point on the template in the world coordinate system is recorded as e. 21 (i,j), e 22 (i,j), e 23 (i,j), e 21 (i,j) represents the x-coordinate of the center of the circle in the i-th row and j-th column of the template in the world coordinate system. 22 (i,j) represents the y-axis coordinate of the center of the circle in the i-th row and j-th column of the template in the world coordinate system. 23 (i,j) represents the z-axis coordinate of the center of the circle in the i-th row and j-th column of the template in the world coordinate system. Using the same method, the template is placed in the third to tenth different known positions to obtain e. 31 (i,j),…,e 101 (i,j), e 32 (i,j),…,e 102 (i,j), e 33 (i,j),…,e 103 (i,j) data; subsequently, the above two-dimensional arrays e 11 (i,j), e 21 (i,j) up to e 101 (i,j) are merged and concatenated into a one-dimensional array f. 31 (k), where k = 1, 2, ..., 100000; e 12 (i,j), e 22 (i,j) up to e 102 (i,j) are merged and concatenated into a one-dimensional array f. 32 (k), e 13 (i,j), e 23 (i,j) up to e 103 (i,j) are merged and concatenated into a one-dimensional array f. 33 (k), the above f 31 (k), f 32 (k), f 33 (k) is denoted as the parameter adaptive coordinate data; finally, a military camera is used to photograph the templates at the above ten locations, and the obtained images are used for detection to determine the pixel positions of the center positions of all dots, and then compared with f. 31 (k) One-to-one correspondence, recording the x-coordinate of the pixel position data in a one-dimensional array h 31a In array (m), m = 1, 2, ..., 100000, the y-coordinates of the pixel position data are recorded in a one-dimensional array h. 32a In (m), h 31a (m), h 32a (m) is denoted as parameter-adaptive pixel data; Step S40: Based on the aforementioned adaptive coordinate data and adaptive pixel data, design an adaptive adjustment rule for the translation parameters of the military camera, and adaptively correct and calibrate the translation parameters between the military camera coordinate system and the world coordinate system, thereby obtaining the final three external translation parameters of the military camera, including: Where f 31 (k), f 32 (k), f 33 (k) represents the parameter adaptive coordinate data, h 31a (m), h 32a (m) represents the parameter adaptive pixel data, k w10 k w11 k w12 For constant adaptive parameters, t x The initial value is selected as t obtained through measurement. x0 ;and t y The initial value is selected as t obtained through measurement. y0 ;t z The initial value is selected as t obtained through measurement. x0 The final t obtained after the above iterative calculation x t y t z Let them be t respectively xf t yf With t zf , denoted as the final three external translation parameters of the military camera; Step S50: Based on the average value of the military camera's external parameters, the final three external translation parameters of the military camera, the adaptive coordinate data, and the parameter adaptive pixel data, nine adaptive rotation parameter rules are designed to adaptively correct the nine rotation parameters between the military camera coordinate system and the world coordinate system, thereby obtaining the final nine rotation parameters of the military camera, including: Where f 31 (k), f 32 (k), f 33 (k) represents the parameter adaptive coordinate data, h 31a (m), h 32a (m) represents the parameter adaptive pixel data, and r is calculated. a1t The initial values ​​for parameters are selected as the average values ​​of the nine external rotation parameters of the military camera obtained in the second step, as follows: r a1t (1)＝r a1 ；r a2t (1)＝r a2 ；r a3t (1)＝r a3 ； r a4t (1)＝r a4 ；r a5t (1)＝r a5 ；r a6t (1)＝r a6 ； r a7t (1)＝r a7 ；r a8t (1)＝r a8 ；r a9t (1)＝r a9 ； Where k w1 k w2 k w3 k w4 k w5 k w6 k w7 k w8 k w9 As a constant adaptive parameter, r is the final value obtained after the above iterative calculation. a1t r a2t r a3t r a4t r a5t r a6t r a7t r a8t r a9t Let them be denoted as r respectively a1f r a2f r a3f r a4f r a5f r a6f r a7f r a8f r a9f , denoted as the final nine external rotation parameters of the military camera; Step S60: Design two diagonally offset targets to form a vector target array; use a probe to contact the measured point, then use a military camera to photograph the targets and detect the pixel positions of the two targets in the photograph; then design a dual-target positioning algorithm to calculate the coordinates of the measured point, including: c1=-(a0r a7f +mr a1f );d1=-(b0r a7f +nr a4f , c2=-(a0r a8f +mr a2f d2=-(b0r a8f +nr a5f ) c3=-(a0r a9f +mr a3f )d3=-(b0r a9f +nr a6f ) c4=-a0t zf -mt xf d4=-b0t zf -nt yf l1=-c4-u1t zf -u1r a7f x1-u1r a8f y1-u1r a9f z1-c1x1-c2y1-c3z1; l2=-d4-v1t zf -v1r a7f x1-v1r a8f y1-v1r a9f z1-d1x1-d2y1-d3z1; l3=-c4-u1t zf -u1r a7f x2-u1r a8f y2-u1r a9f z2-c1x2-c2y2-c3z2; l4=-d4-v1t zf -v1r a7f x2-v1r a8f y2-v1r a9f z2-d1x2-d2y2-d3z2; The target design and data acquisition method for u1, v1, u2, v2, and x1, y1, z1, x2, y2, z2 are as follows: Two diagonally offset targets are designed to form a target array. Each target mainly consists of an unmodulated high-power emitter and a heat sink, with the heat sink to be made of copper-aluminum composite material. Its function is to radiate light pulses that meet the requirements of military camera photography, providing sufficient energy for the camera and ensuring that the target pixel positions are clearly identifiable in the photograph. During measurement, the preset coordinates of the measured point are x... a y a z a The coordinates of the two targets relative to the test point group are plotted as x a +x1、y a +y1、z a +z1 and x a +x2、y a +y2、z a +z2; while x1, y1, z1, x2, y2, z2 are constant values, designed based on the relative positions between the target and the probe; then, a military camera is used to photograph the target, and the pixel positions of the two targets in the photograph are detected, denoted as u1, v1 and u2, v2; m and n are intermediate variables, c x c y The actual size of each pixel; a0 and b0 represent the pixel position coordinates of the center of the image coordinate system in the pixel coordinate system, r a1f r a2f r a3f r a4f r a5f r a6f r a7f r a8f r a9f Let L1 and L2 be the final nine external rotation parameters of the military camera; L1 and L2 are the coordinate column vectors calculated based on the above intermediate variables and target pixel position data; W1 and W2 are the coordinate matrices calculated based on the above intermediate variables and target pixel position data; W n1 With W n2 Let A be the left inverse of coordinate matrices W1 and W2. n1 With A n2 For the first and second estimated values ​​of the final measured point coordinates, A n The coordinates of the measured point are represented by a 3x1 column vector, where each element represents the three-dimensional coordinates (x, y) of the measured point. a y a z a Thus, by solving A n The three-dimensional position coordinates of the measured point can then be obtained.

Images