Calibration method for X-ray imaging device combined with parallel external fixation support

Abstract

The invention discloses an X-ray imaging device calibration method combined with a parallel external fixation support. The method comprises the following steps: constructing a mathematical model of an X-ray imaging system; acquiring three-dimensional coordinates of the marking points; pixel coordinates of the mark points are obtained; solving internal parameters; and solving external parameters to complete calibration of internal and external parameters of the camera. According to the method, the linear pinhole model is used as the calibration model, the calculation formula is simple, operation is easy and convenient, implementation is convenient in the operation, the method is suitable for the operation scene with little image distortion, the random sorting evolutionary algorithm is adopted to optimize and calculate the internal and external parameters of the camera, the robustness is good, and the solving precision is high. The X-ray image calibration device can meet the requirement for X-ray image calibration in clinical operation.

Description

Technical Field

[0001] The invention belongs to the technical field of X-ray imaging device calibration, and in particular relates to a calibration method for an X-ray imaging device combined with a parallel external fixing bracket. Background Art

[0002] With the rapid development of social economy and the improvement of urbanization level, fractures caused by traffic accidents, high-altitude falls caused by infrastructure construction, and falls in overweight people and the elderly have shown an increasing trend year by year, with the annual incidence of accidental (unintentional) non-fatal falls being 2831 / 100000. Fracture patients usually need fracture treatment to restore normal limb function.

[0003] Fracture treatment can be divided into three parts: reduction, fixation, and rehabilitation, among which fracture reduction is the key. Traditional reduction methods include closed reduction and open reduction. Closed reduction refers to the reduction of the fracture ends by direct or instrumental traction under non-direct vision conditions. Its advantages are small damage and simple operation. Its disadvantages are low reduction accuracy, frequent need for secondary reduction, repeated X-ray fluoroscopy of doctors and patients, high radiation dose, and difficulty in treating complex fractures. Open reduction refers to the reduction of the fracture ends by doctors under direct vision conditions under which the fracture site is exposed by surgical means. Its advantages are high accuracy and intuitive operation. Its disadvantages are high requirements for comprehensive anatomy and large surgical trauma. It is easy to cause complications such as tissue adhesion and fracture end necrosis, which affect fracture healing. Traditional reduction methods mainly rely on the doctor's human eyes to locate the reduction target of the broken bone. Its empirical positioning has the disadvantages of low accuracy and low efficiency, and the reduction effect is not ideal.

[0004] Parallel external fixators have been widely used in the field of fracture treatment due to their high precision and high rigidity. Since the invention of the Taylor space stent in 1994, scientific research institutions and enterprises have continuously improved and innovated on its basis, and derived similar products such as the Orthofix stent and the Ortho-SUV stent. The auxiliary diagnosis and treatment software of the parallel external fixator is a necessary condition for the stent to play its fracture reduction function. Through the calculation and planning of the software, it guides the doctor to adjust the length of the branch chain, thereby achieving accurate reduction of the fracture ends. Among them, determining the relative position of the affected bone and the external fixator is a key step in guiding reduction.

[0005] Existing parallel external fixator auxiliary diagnosis and treatment software can be divided into two categories according to the reference data source. The first category uses the flat film marking method to determine the relative position of the affected bone and the external fixator. Lines are drawn in the anteroposterior and lateral flat films to represent the plane of the proximal bone axis, the distal bone axis and the static platform, and a set of corresponding points are selected at the proximal and distal ends of the fracture (the corresponding points will coincide after reduction); the angle between the proximal bone axis and the distal bone axis, the angle between the proximal bone axis and the reference platform, and the distance between the corresponding points are measured to obtain the relative position of the reference platform and the proximal bone, and the relative position of the proximal bone and the distal bone. Finally, the relative position of the static platform and the dynamic platform is obtained through kinematics. The marking positioning method requires too many parameters to be measured and the operation is cumbersome. In addition, its mathematical model replaces the bone model with the bone axis, ignoring the axial rotation deformity between the broken bones, and the measurement accuracy is poor.

[0006] The second type determines the relative position of the affected bone and the external fixator based on CT three-dimensional reconstruction, and uses identifiable feature points on the CT data for registration and positioning. By calculating the mapping of the marker points from the CT coordinate system to the bracket coordinate system, the position of the affected bone in the bracket coordinate system is determined, so as to plan the reduction path and perform the reduction operation. The positioning method based on CT reconstruction increases the intensity of radiation exposure to the patient. At the same time, postoperative CT reconstruction usually prolongs the time of fracture treatment and increases the risk of infection.

[0007] The positional relationship between the affected bones can be determined using 2D / 3D registration, which links the image of the image acquisition device (such as a C-arm) with the model of the affected bone obtained by preoperative computed tomography (CT). Based on the coordinates of the image acquisition device, the positional relationship between the instrument and the patient is calculated through registration based on the marker points, and then the positional relationship between the affected bones is obtained using the correct solution of the near and far platforms. The method of localizing the affected bones with the help of preoperative CT and intraoperative X-rays can not only accurately obtain the axial parameters of the bone, but also complete the operation at a low cost during the operation. It is a method with great application potential. Among them, the calibration of the imaging device is the basis for establishing accurate spatial coordinate mapping, and plays a decisive role in improving the accuracy of 2D-3D registration.

[0008] So far, many researchers have proposed different calibration methods for imaging equipment calibration, among which the more classic methods include Tsai calibration method, Faugeras calibration method, Zhang Zhengyou calibration method, etc. Although the above methods can achieve good calibration effects, they still have certain limitations. Zhang Zhengyou calibration method requires taking multiple images, which is time-consuming; Faugeras calibration method is prone to fall into local optimal solutions rather than global optimal solutions; Tsai two-step calibration method has complex nonlinear optimization steps and slow calculations. In addition, there is a lack of specific calibration methods for the application scenarios described in this patent. In order to meet the needs of accurate, efficient and robust X-ray image calibration, a calibration method based on the scale invariance of the distance between markers is proposed. The image calibration problem of the combined external fixator is solved by simplifying the mathematical model. Summary of the invention

[0009] The present invention is proposed to solve the problems existing in the prior art, and its purpose is to provide a calibration method for an X-ray imaging device combined with a parallel external fixation bracket.

[0010] The technical solution of the present invention is: a calibration method of an X-ray imaging device combined with a parallel external fixation bracket, comprising the following steps:

[0011] A. Construct a mathematical model of the X-ray imaging system;

[0012] B. Get the three-dimensional coordinates of the marker points;

[0013] C. Get the pixel coordinates of the marker point;

[0014] D. Solve internal parameters;

[0015] E. Solve the external parameters and complete the calibration of the internal and external parameters of the camera.

[0016] Furthermore, step A constructs a mathematical model of the X-ray imaging system, and the specific process is as follows:

[0017] First, the X-ray imaging process is similar to the pinhole camera model in terms of imaging principle;

[0018] Then, the pinhole camera model is used as the mathematical model of the C-arm X-ray imaging system.

[0019] Furthermore, step B obtains the three-dimensional coordinates of the marking point, and the specific process is as follows:

[0020] First, select the marker.

[0021] Then, the representation of the motion platform marking point in the motion platform coordinate system is obtained;

[0022] Then, the representation of the fixed platform marking point in the fixed platform coordinate system is obtained;

[0023] Then, the kinematics is solved according to the length of each branch link of the parallel external fixed bracket to calculate the relative position and posture of the moving platform and the fixed platform;

[0024] Finally, obtain the coordinates of the marker point in the world coordinate system.

[0025] Furthermore, the coordinates of the marker points in the world coordinate system are obtained. The specific process is as follows:

[0026] First, calculate the representation of all marker points in the fixed platform coordinate system;

[0027] Then, assume that the world coordinate system coincides with the fixed platform coordinate system;

[0028] Finally, obtain the coordinates of the marker point in the world coordinate system.

[0029] Furthermore, step C obtains the pixel coordinates of the marker point, and the specific process is as follows:

[0030] First, perform Canny edge detection on the input image to obtain a binary image of edge detection;

[0031] Then, the Sobel operator is performed on the input image to calculate the neighborhood gradient values ​​of all pixels;

[0032] Then, initialize an accumulator array to record possible circle center positions;

[0033] Then, traverse the non-zero pixel points obtained by Canny edge detection, and for each edge point, vote in the accumulator along the gradient direction to record the possible center position;

[0034] Finally, the coordinates of the marker point in the pixel coordinate system are obtained.

[0035] Furthermore, the coordinates of the marker points in the pixel coordinate system are obtained. The specific process is as follows:

[0036] First, count the values ​​in the accumulator;

[0037] Then, sort all possible circle center positions;

[0038] Then, find the coordinates of the center of the circle whose number of votes is greater than the specified threshold, and obtain the coordinates of the marked point in the pixel coordinate system through screening and optimization.

[0039] Furthermore, step D performs internal parameter solving, and the specific process is as follows:

[0040] First, the internal parameter equations are constructed;

[0041] Then, the random sorting evolutionary algorithm is used to solve the camera internal parameters.

[0042] Furthermore, the internal parameter equations are constructed. The specific process is as follows:

[0043] First, the transformation from the world coordinate system to the camera coordinate system is a rigid transformation;

[0044] Then, the distances between the markers are kept constant and a set of equations to be solved regarding the internal parameters is constructed.

[0045] Furthermore, the random sorting evolutionary algorithm is used to solve the internal parameters of the camera. The specific process is as follows:

[0046] First, define the fitness function consisting of the objective function and the penalty function;

[0047] Then, the random sorting evolutionary algorithm is used to avoid premature convergence to the local optimal solution, and the internal parameters are solved through continuous iterative optimization.

[0048] Furthermore, step E solves the external parameters to complete the calibration of the internal and external parameters of the camera. The specific process is as follows:

[0049] First, after solving the internal parameters, the coordinates of the marker point in the camera coordinate system can be obtained;

[0050] Then, given the coordinates of the corresponding world coordinate system, the equations are listed according to the relationship between the corresponding points and solved using the least squares method;

[0051] Finally, complete the calibration of all camera internal and external parameters.

[0052] The beneficial effects of the present invention are as follows:

[0053] The present invention utilizes a linear pinhole model as a calibration model, which has a simple calculation formula, is easy to operate, and is convenient to implement during surgery. It is suitable for surgical scenarios with small image distortion, and uses a randomly sorted evolutionary algorithm to optimize and calculate the internal and external parameters of the camera, which has good robustness and high solution accuracy.

[0054] The camera calibration method based on the scale invariance of the distance between marker points of the present invention can meet the demand for X-ray image calibration in clinical surgery. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 Schematic diagram of the camera calibration pinhole model in the present invention;

[0056] Figure 2 This is a structural diagram of a six-branch parallel external fixing bracket in one embodiment of the present invention;

[0057] Figure 3 A schematic diagram of a two-dimensional point coordinate calculation process in a pixel coordinate system in one embodiment of the present invention;

[0058] Figure 4 The figure is a flow chart of an evolutionary algorithm based on random sorting in one embodiment of the present invention. DETAILED DESCRIPTION

[0059] Hereinafter, the present invention will be described in detail with reference to the accompanying drawings and embodiments:

[0060] like Figures 1 to 4 As shown, a calibration method for an X-ray imaging device combined with a parallel external fixator comprises the following steps:

[0061] A. Construct a mathematical model of the X-ray imaging system;

[0062] B. Get the three-dimensional coordinates of the marker points;

[0063] C. Get the pixel coordinates of the marker point;

[0064] D. Solve internal parameters;

[0065] E. Solve the external parameters and complete the calibration of the internal and external parameters of the camera.

[0066] Step A constructs a mathematical model of the X-ray imaging system. The specific process is as follows:

[0067] First, the X-ray imaging process is similar to the pinhole camera model in terms of imaging principle;

[0068] Then, the pinhole camera model is used as the mathematical model of the C-arm X-ray imaging system.

[0069] Step B obtains the three-dimensional coordinates of the marking point. The specific process is as follows:

[0070] First, click on the marker;

[0071] Then, the representation of the motion platform marking point in the motion platform coordinate system is obtained;

[0072] Then, the representation of the fixed platform marking point in the fixed platform coordinate system is obtained;

[0073] Then, the kinematics is solved according to the length of each branch link of the parallel external fixed bracket to calculate the relative position and posture of the moving platform and the fixed platform;

[0074] Finally, obtain the coordinates of the marker point in the world coordinate system.

[0075] Get the coordinates of the marker point in the world coordinate system. The specific process is as follows:

[0076] First, calculate the representation of all marker points in the fixed platform coordinate system;

[0077] Then, assume that the world coordinate system coincides with the fixed platform coordinate system;

[0078] Finally, obtain the coordinates of the marker point in the world coordinate system.

[0079] Step C obtains the pixel coordinates of the marker point. The specific process is as follows:

[0080] First, perform Canny edge detection on the input image to obtain a binary image of edge detection;

[0081] Then, the Sobel operator is performed on the input image to calculate the neighborhood gradient values ​​of all pixels;

[0082] Then, initialize an accumulator array to record possible circle center positions;

[0083] Then, traverse the non-zero pixel points obtained by Canny edge detection, and for each edge point, vote in the accumulator along the gradient direction to record the possible center position;

[0084] Finally, the coordinates of the marker point in the pixel coordinate system are obtained.

[0085] Get the coordinates of the marker point in the pixel coordinate system. The specific process is as follows:

[0086] First, count the values ​​in the accumulator;

[0087] Then, sort all possible circle center positions;

[0088] Then, find the coordinates of the center of the circle whose number of votes is greater than the specified threshold, and obtain the coordinates of the marked point in the pixel coordinate system through screening and optimization.

[0089] Step D is to solve the internal parameters, and the specific process is as follows:

[0090] First, the internal parameter equations are constructed;

[0091] Then, the random sorting evolutionary algorithm is used to solve the camera internal parameters.

[0092] Construct the internal parameter equation group. The specific process is as follows:

[0093] First, the transformation from the world coordinate system to the camera coordinate system is a rigid transformation;

[0094] Then, the distances between the markers are kept constant and a set of equations to be solved regarding the internal parameters is constructed.

[0095] The random sorting evolutionary algorithm is used to solve the internal parameters of the camera. The specific process is as follows:

[0096] First, define the fitness function consisting of the objective function and the penalty function;

[0097] Then, the random sorting evolutionary algorithm is used to avoid premature convergence to the local optimal solution, and the internal parameters are solved through continuous iterative optimization.

[0098] Step E solves the external parameters and completes the calibration of the internal and external parameters of the camera. The specific process is as follows:

[0099] First, after solving the internal parameters, the coordinates of the marker point in the camera coordinate system can be obtained;

[0100] Then, given the coordinates of the corresponding world coordinate system, the equations are listed according to the relationship between the corresponding points and solved using the least squares method;

[0101] Finally, complete the calibration of all camera internal and external parameters.

[0102] Specifically, the pinhole camera model in step A regards the propagation of X-rays as passing through an idealized point light source (pinhole) and penetrating the object, forming an inverted real image on the imaging plane, which is highly consistent with the imaging mechanism of the pinhole camera. At the same time, the pinhole camera model provides a simplified physical framework, which abstracts the imaging process into a pure geometric projection problem by ignoring complex optical distortion, thereby simplifying the mathematical description of the imaging mechanism. And its simplified geometric relationship can effectively reduce the computational complexity of the C-arm X-ray imaging system model parameters while maintaining sufficient accuracy.

[0103] Specifically, step A constructs a mathematical model of the X-ray imaging system, as follows:

[0104] like Figure 1 As shown, an ideal point model is established, where O w -X w Y w Z w is the world coordinate system, O c -X c Y c Z c is the camera coordinate system, O g -xy is the image coordinate system, O I -uv is the pixel coordinate system. p of any point in the approximate space w The imaging position in the image can be approximated by the pinhole model, that is, any point p w The projection position p in the image g Optical center O c With p w The line connecting the points O c p w The point of intersection with the image plane.

[0105] The conversion relationship between the world coordinate system, camera coordinate system and pixel coordinate system is as follows: In the pixel coordinate system, let the origin O gThe coordinates in the u, v coordinate system are (u 0 ,v 0 ), the physical size of each pixel in the x-axis and y-axis directions is dx, dy, then the coordinate relationship of any pixel in the image in the two coordinate systems is:

[0106]

[0107] Expressed in homogeneous coordinate and matrix form, we have

[0108]

[0109] Space point p w (x w ,y w ,z w ) has a coordinate of p in the camera coordinate system. c (x c ,y c ,z c ), the projection point on the camera plane is p g (x,y), according to the proportional relationship:

[0110]

[0111] Expressed in homogeneous coordinate and matrix form, we have:

[0112]

[0113] The following relationship exists between the world coordinate system and the camera coordinate system:

[0114]

[0115] Where, the rotation matrix R is a 3×3 orthogonal unit matrix, t is a three-dimensional translation vector, 0 = (0, 0, 0), M 2 is a 4×4 matrix. Substituting equation (2) and equation (5) into equation (4), we have:

[0116]

[0117] Thus, the relationship between the image plane coordinate system and the world coordinate system is established, where M 1 Including equivalent focal length f, image center point u 0 、v 0 , pixel size dx, dy parameters, because they are only related to the internal structure of the camera, they are called camera internal parameters. 2 Determined by the orientation of the camera relative to the world coordinate system, it is called the camera's external parameters. In general, dx = dy and u 0 、v 0 As a known condition, u 0、v 0 They are half of the image pixel width and pixel height respectively. Therefore, the camera needs to calibrate 2 camera internal parameters and 6 camera external parameters, a total of 8 parameters.

[0118] Specifically, step B obtains the three-dimensional coordinates of the marking point as follows:

[0119] like Figure 2 As shown, Figure 2 is a six-branch parallel external fixed bracket structure diagram, O A -X A Y A Z A and O B -X B Y B Z B Represent the central coordinate systems of the fixed platform and the moving platform respectively. represents the coordinates of the six hinge points on the fixed platform under the fixed platform, Indicates the coordinates of the six hinge points on the motion platform under the motion platform. The length of the six driving legs is L i It is also known that, then the problem of solving the position (x, y, z) and posture (α, β, γ) of the motion platform in the Cartesian coordinate system is:

[0120]

[0121] Where R is the coordinate system O A to B The rotation matrix contains the posture information of the motion platform, and t is the coordinate system O A to B The translation matrix contains the position information of the motion platform.

[0122] According to the length of the six branches, a six-dimensional nonlinear equation group can be listed, and the coordinate system O can be solved by Jacobi matrix method. A to B The rotation matrix R and translation matrix t.

[0123] The position of the marker point in the motion platform coordinate system Where n≥4, then Calculate the coordinates in the fixed platform coordinate system.

[0124] Assume that the world coordinate system coincides with the fixed platform coordinate system, and the coordinates of the marking point in the world coordinate system are:

[0125] Specifically, step C obtains the two-dimensional position of the marking point in the pixel coordinate system, see Appendix Figure 3 , as follows:

[0126] c1. Preprocess the original image and remove points with pixel values ​​less than 125.

[0127] c2. Use the Canny edge detection algorithm to obtain the edge point set E of the image.

[0128] c3. Use the Sobel operator to calculate the gradient direction of each edge point.

[0129] c4. Count the intersection values ​​of all gradient directions (x k ,y k ). For each intersection value, the distance value between it and the edge point is the radius candidate value. The radius value that appears the most times is set as the radius r k (5≤r k ≤50).

[0130] c5. At all intersection values ​​(x k ,y k ) Find the points where the number of votes is greater than q (usually set to 80), and the parameters (x k ,y k ,r k ) is the parameter of the circle existing in the image, and is added to the circle queue. The center coordinates of circles with a center distance less than 5 pixels are set to the same center value.

[0131] c6. Judge (x k ,y k ) Whether the pixel value of the original image corresponding to the point is less than 30, if so, it is considered as an interference point and removed from the circle queue.

[0132] c7. According to the detected circle queue parameters (x k ,y k ,r k ), draw the corresponding circle in the image and filter it. Get the coordinate position of the two-dimensional point

[0133] Specifically, step D performs internal parameter solving, as follows:

[0134] d1. Use the scale invariance of the distance between the markers to construct the internal parameter equations and solve the camera internal parameters. Suppose the coordinates of the markers in the camera coordinate system are

[0135]

[0136] Since the world coordinate system is converted to the camera coordinate system, M 2 Convert to a rigid transformation, so the distance between the markers remains unchanged.

[0137]

[0138] In the formula, and Respectively expressed in the world coordinate system Marking points and The distance of the marker point and the camera system Marking points and The distance of the marked point. There are a total of (2+n) parameters to be determined: f, dx,

[0139] d2. Use the improved random sorting evolutionary algorithm to solve the parameters, see Figure 4 .

[0140] d21. In the search space The fitness function is defined as:

[0141] ψ(x)=f(x)+φ(x) (11)

[0142]

[0143] Where x is the camera internal parameter f, dx and the z-axis coordinate value in the camera coordinate system The vector formed, f(x) is the objective function, that is, the Euclidean distance of the sum of (n*(n-1) / 2) distance differences, φ(x) is the penalty function, representing (n*(n-1) / 2) nonlinear equality constraints. Given a population size of 120, a crossover rate of 0.45, and a mutation rate of 0.002, the maximum number of iterations is 10,000.

[0144] d22. Randomly generate 120 individuals from U to form an initial population S = {s 1 ,s 2 ,…,s 120}, set the algebraic counter iter = 0.

[0145] d23. Calculate the fitness of each individual in S and sort them according to the random sorting algorithm. Perform a maximum of 240 rounds of sorting. Each round of sorting requires traversing and comparing the fitness of two adjacent solutions. After 120 iterations, if the ranking order does not change in a complete traversal, exit the sorting.

[0146] The sorting rules are as follows: given any pair of adjacent individuals, if both individuals are feasible solutions, compare their objective functions, and the one with a smaller objective function is moved to a higher position, indicating that its fitness is high. If at least one of them is an infeasible solution, find a random number from 0 to 1. If the random number is less than 0.4, compare the objective functions, and place the one with a smaller objective function at a higher position; otherwise, compare the penalty functions of the two, and place the one with a smaller penalty function at a higher position.

[0147] d24. If the maximum number of iterations has been reached or the individual with the largest fitness in S meets the acceptable error range of the objective function, that is, f(x) < 10 -4 , the algorithm ends.

[0148] d25. Select the first 24 individuals with the highest fitness values ​​and use the league selection algorithm to obtain the values ​​of the other 96 individuals to form a new group of 120 individuals, recorded as S1.

[0149] d26. Randomly select 54 individuals from S1 with a crossover rate of 0.45, pair them up for crossover operation, and replace the original individuals with the generated new individuals to obtain population S2.

[0150] d27. According to the mutation number 1 determined by the mutation rate 0.002, randomly select an individual from S2, perform a mutation operation on one of the solution values ​​and replace the original individual with the generated new individual, and obtain the population S3.

[0151] d28. Use population S3 as the new generation population, that is, replace S with S3, iter=iter+1, and repeat step d23.

[0152] Specifically, step E solves the external parameters and completes the calibration of the internal and external parameters of the camera, as follows:

[0153] After the internal parameters are solved, the camera coordinates of the marker point are obtained. It is also known that its corresponding world coordinate system coordinates are The corresponding point equations are expressed as follows:

[0154] E·A=C (14)

[0155] In the formula,

[0156]

[0157] A 12×1 =[r 1 r 2 r 3 r 4 r 5 r 6 r 7 r 8 r 9 t 1 t 2 t 3 ] T (16)

[0158]

[0159] The external parameters are calculated according to the least squares method formula A = (E T E)-1 E T C is solved. So far, the calibration of all camera internal and external parameters has been completed.

[0160] The present invention utilizes a linear pinhole model as a calibration model, which has a simple calculation formula, is easy to operate, and is convenient to implement during surgery. It is suitable for surgical scenarios with small image distortion, and uses a randomly sorted evolutionary algorithm to optimize and calculate the internal and external parameters of the camera, which has good robustness and high solution accuracy.

[0161] The camera calibration method based on the scale invariance of the distance between marker points of the present invention can meet the demand for X-ray image calibration in clinical surgery.

[0162] The above description of the present invention is merely illustrative and not restrictive, so the embodiments of the present invention are not limited to the above specific embodiments. If a person skilled in the art is inspired by the above, without departing from the scope of protection of the present invention and the claims, other changes or modifications are made, which belong to the protection scope of the present invention.

Claims

1. A method for calibrating an X-ray imaging device in conjunction with a parallel external fixator, characterized in that: The following steps are involved: A. Construct a mathematical model of the X-ray imaging system; B. Get the three-dimensional coordinates of the marker points; C. Get the pixel coordinates of the marker point; D. Solve internal parameters; E. Solve the external parameters and complete the calibration of the internal and external parameters of the camera.

2. The calibration method of an X-ray imaging device combined with a parallel external fixator according to claim 1, characterized in that: Step A constructs a mathematical model of the X-ray imaging system. The specific process is as follows: First, the X-ray imaging process is similar to the pinhole camera model in terms of imaging principle; Then, the pinhole camera model is used as the mathematical model of the C-arm X-ray imaging system.

3. The calibration method of an X-ray imaging device combined with a parallel external fixator according to claim 1, characterized in that: Step B obtains the three-dimensional coordinates of the marking point. The specific process is as follows: First, click on the marker; Then, the representation of the motion platform marking point in the motion platform coordinate system is obtained; Then, the representation of the fixed platform marking point in the fixed platform coordinate system is obtained; Then, the kinematics is solved according to the length of each branch chain of the parallel external fixed bracket to calculate the relative position and posture of the moving platform and the fixed platform; Finally, obtain the coordinates of the marker point in the world coordinate system.

4. The calibration method of an X-ray imaging device combined with a parallel external fixator according to claim 3, characterized in that: Get the coordinates of the marker point in the world coordinate system. The specific process is as follows: First, calculate the representation of all marker points in the fixed platform coordinate system; Then, assume that the world coordinate system coincides with the fixed platform coordinate system; Finally, obtain the coordinates of the marker point in the world coordinate system.

5. The calibration method of an X-ray imaging device combined with a parallel external fixator according to claim 1, characterized in that: Step C obtains the pixel coordinates of the marker point. The specific process is as follows: First, perform Canny edge detection on the input image to obtain a binary image of edge detection; Then, the Sobel operator is performed on the input image to calculate the neighborhood gradient values ​​of all pixels; Then, initialize an accumulator array to record possible circle center positions; Then, traverse the non-zero pixel points obtained by Canny edge detection, and for each edge point, vote in the accumulator along the gradient direction to record the possible center position; Finally, the coordinates of the marker point in the pixel coordinate system are obtained.

6. A calibration method for an X-ray imaging device combined with a parallel external fixator according to claim 5, characterized in that: Get the coordinates of the marker point in the pixel coordinate system. The specific process is as follows: First, count the values ​​in the accumulator; Then, sort all possible circle center positions; Then, find the coordinates of the center of the circle whose number of votes is greater than the specified threshold, and obtain the coordinates of the marked point in the pixel coordinate system through screening and optimization.

7. The calibration method of an X-ray imaging device combined with a parallel external fixator according to claim 1, characterized in that: Step D is to solve the internal parameters, and the specific process is as follows: First, the internal parameter equations are constructed; Then, the random sorting evolutionary algorithm is used to solve the camera internal parameters.

8. The calibration method of an X-ray imaging device combined with a parallel external fixator according to claim 7, characterized in that: Construct the internal parameter equation group. The specific process is as follows: First, the transformation from the world coordinate system to the camera coordinate system is a rigid transformation; Then, the distances between the markers are kept constant and a set of equations to be solved regarding the internal parameters is constructed.

9. The calibration method of an X-ray imaging device combined with a parallel external fixator according to claim 7, characterized in that: The random sorting evolutionary algorithm is used to solve the internal parameters of the camera. The specific process is as follows: First, define the fitness function consisting of the objective function and the penalty function; Then, the random sorting evolutionary algorithm is used to avoid premature convergence to the local optimal solution, and the internal parameters are solved through continuous iterative optimization.

10. The calibration method of an X-ray imaging device combined with a parallel external fixator according to claim 7, characterized in that: Step E solves the external parameters and completes the calibration of the internal and external parameters of the camera. The specific process is as follows: First, after solving the internal parameters, the coordinates of the marker point in the camera coordinate system can be obtained; Then, given the coordinates of the corresponding world coordinate system, the equations are listed according to the relationship between the corresponding points and solved using the least squares method; Finally, complete the calibration of all camera internal and external parameters.

Images