Industrial CT pose parameter calibration method based on microsphere motif

Abstract

The invention provides an industrial CT (Computed Tomography) pose parameter calibration method based on a microsphere motif, which relates to the field of industrial CT imaging and comprises the following calibration steps: acquiring perspective images of the microsphere motif at different positions; extracting circle center coordinates of a microsphere projection circle in the perspective image; calculating the coordinates of two projection points on the rotating shaft according to the circle center coordinates of the projection circle so as to calibrate the in-plane deflection angle of the detector; establishing a mathematical relationship model among pose parameters, model body parameters and projection image coordinates, and calibrating the distance between a radiation source and a rotating shaft of a rotating table and the distance between the radiation source and a detector; and establishing a mathematical relationship model among the circle center coordinates of the projection circle of the sphere a, the offset of the detector and the parameters of the industrial CT system, and realizing the calculation of the offset of the detector. The rotary table does not need to be rotated by too many angles, the influence of the motion error of the rotary table on the calibration result is avoided, and meanwhile the machining difficulty of the die body is reduced.

Description

Technical fields:

[0001] This invention relates to a method for calibrating pose parameters in industrial CT scanners. Background technology:

[0002] Industrial CT (Computed Tomography) technology utilizes X-rays to penetrate a sample. The X-rays, carrying sample information, are received by a detector to create an image, and a reconstruction algorithm is used to detect the three-dimensional structure inside the sample. Before using industrial CT for inspection, the pose parameters of the X-ray source, rotating stage, and detector involved in the industrial CT process must be calibrated to improve the imaging performance of the industrial CT system.

[0003] Currently, methods for calibrating pose parameters in industrial CT systems can be mainly divided into single-sphere rotational imaging calibration and single-angle imaging calibration using a specially designed phantom. The single-sphere rotational imaging calibration method utilizes a single microsphere for multi-angle rotational imaging to establish a relationship model between the system's pose parameters and the projected image, thereby obtaining the desired pose parameters. However, this method, due to the need for multi-angle imaging, is susceptible to the influence of rotary table motion errors, resulting in inaccurate pose parameter calibration. The single-angle imaging calibration method using a specially designed phantom utilizes a single perspective image of the phantom to analyze the mathematical relationship between pose parameters and the projected image, overcoming the influence of rotary table motion errors on pose parameter accuracy. However, it requires high precision in phantom machining. Therefore, a technical solution is desired to overcome or at least mitigate the shortcomings of existing technologies, achieving simple, convenient, and high-precision calibration of pose parameters in industrial CT systems. Summary of the Invention:

[0004] The purpose of this invention is to provide a method for calibrating pose parameters in industrial CT scans to overcome or at least mitigate the deficiencies in existing technologies. The technical solution is as follows:

[0005] A method for calibrating pose parameters in industrial CT based on microsphere phantoms, comprising the following steps:

[0006] (1) Move the rotary table to the initial position (denoted as position A) via the motion axis, place the microsphere phantom on the rotary table, and set appropriate imaging parameters to obtain a perspective image of the microsphere phantom. Specifically, this includes:

[0007] ① Position the microsphere phantom plane as perpendicular as possible to the X-ray beam direction (this position is denoted as A1), turn on the industrial CT system, and acquire the phantom's perspective image P at this angle. A1 Using the Hough transform to apply the projection image P A1 The sub-pixel centers of the four microsphere projections are extracted, denoted as (u Aa1 ,v Aa1 ), (u Ab1 ,vAb1 ), (u Ac1 ,v Ac1 ), (u Ad1 ,v Ad1 );

[0008] ② Rotate the stage counterclockwise by 60° (this position is denoted as A2) and obtain the perspective image P of the model at this angle. A2 Using the Hough transform to apply the projection image P A2 The sub-pixel centers of the four microsphere projections are extracted, denoted as (u Aa2 ,v Aa2 ), (u Ab2 ,v Ab2 ), (u Ac2 ,v Ac2 ), (u Ad2 ,v Ad2 );

[0009] ③ Rotate the stage clockwise by 120° (this position is denoted as A3) and obtain the perspective image P of the model at this angle. A3 Using the Hough transform to apply the projection image P A3 The sub-pixel centers of the four microsphere projections are extracted, denoted as (u Aa3 ,v Aa3 ), (u Ab3 ,v Ab3 ), (u Ac3 ,v Ac3 ), (u Ad3 ,v Ad3 );

[0010] (2) Calculate the coordinates of the projection point of the rotary table axis on the detector at position A; the lines connecting the projections of spheres a and b at positions A1, A2, and A3 intersect at a point, whose coordinates are: (u A-up ,v A-up ); The lines connecting the projections of spheres c and d at positions A1, A2, and A3 intersect at a single point, denoted as: (u A-down ,v A-down The line connecting the two intersection points represents the projection of the rotation axis onto the detector; (u A-up ,v A-up ), (u A-down ,v A-down It can be obtained from the following formula:

[0011]

[0012] (3) Calculate the relative deflection angle between the rotation axis and the detector when the rotary table is at position A, that is, the in-plane deflection angle of the detector at this time, denoted as γ. A :

[0013]

[0014] (4) Use the motion axis to move the rotary table to positions B, C, D, and E, and repeat steps (1) to (3) to calculate γ respectively. B γ C γ D γ E Then the in-plane deflection angle γ of the detector can be obtained:

[0015]

[0016] (5) Calculation of the distance SOD between the X-ray source and the rotating stage axis, and the distance SDD between the X-ray source and the detector. Specifically, this includes:

[0017] ① When the rotary table is at position A, the projection distance on the image between the lines containing spheres a and b and the lines containing spheres c and d is denoted as h. pro-A :

[0018]

[0019] ② Similarly, when the rotary table is located at positions B, C, D, and E, the corresponding projection spacing is h. pro-B h pro-C h pro-D h pro-E ;

[0020] ③ By establishing a mathematical relationship model between pose parameters, phantom parameters, and projected images, the distance SOD between the X-ray source and the rotation axis of the rotary stage, and the distance SDD between the X-ray source and the detector can be calculated, as shown in the following formula:

[0021]

[0022] Where l is the distance between the lines containing spheres a and b and the lines containing spheres c and d in the model, and Δ is the distance between positions A, B, C, D, and E.

[0023] (6) Calculation of detector offsets u and v. The specific steps are as follows:

[0024] ① Use the following formula to transform the coordinates of the center of sphere a when the rotary table is in various positions:

[0025]

[0026] Among them, u ia1 v ia1 Let x and y be the coordinates of the center of the projection of sphere a when the rotary table is at position i. Here are the x and y coordinates of the center of the projected circle of the transformed sphere a;

[0027] ②Establish a mathematical relationship model between the coordinates of the projection center of sphere a, the detector offset, and the parameters of the industrial CT system, and calibrate the detector offsets u and v.

[0028]

[0029] Where PixelSize is the detector pixel size, Resolution is the number of pixels in a column or row of the detector, and d is an intermediate variable that does not require a solution.

[0030] The technical features and innovations of this invention are as follows: the microsphere phantom is used to calibrate the pose parameters of industrial CT. This method is simple, convenient and highly accurate. The calibration of pose parameters can effectively improve the imaging performance of industrial CT. Attached image description:

[0031] Figure 1 Schematic diagram of industrial CT system and its pose parameters; specifically including: X-ray source 1; microsphere phantom 2; rotary stage 3; detector 4.

[0032] Figure 2 Schematic diagram of microsphere phantom

[0033] Figure 3 Top view of the imaging position Detailed implementation method:

[0034] The purpose of this invention is to provide a method for calibrating pose parameters of industrial CT based on microsphere phantoms, thereby improving the imaging performance of industrial CT. The method is simple, practical and reliable.

[0035] refer to Figures 1-3 This invention is achieved through the following technical solution: an industrial CT pose parameter calibration method based on a microsphere phantom, which can be implemented through the following scheme:

[0036] Step 1: In Figure 1 The industrial CT system shown has a microsphere phantom 3 mounted on the rotary stage 2. The rotary stage is moved to a position via a motion axis as shown. Figure 3 At location A, set appropriate industrial CT imaging parameters to obtain a perspective image of the microsphere phantom, and extract the center of the projection circle of the perspective image. Specifically, the required perspective image includes:

[0037] Step 11: Position the plane of the microsphere phantom as perpendicular as possible to the direction of the ray beam, such as... Figure 3 At position 1 shown, denoted as A1, the industrial CT system is activated to acquire the perspective image P of the microsphere phantom at this angle. A1 ;

[0038] Step 12: Rotate the mold body counterclockwise by 60° using a rotary table, until... Figure 3At position 2 shown, denoted as A2, obtain the perspective image P of the phantom at this angle. A2 ;

[0039] Step 13: Rotate the mold body 120° clockwise using a rotary table until it reaches the desired position. Figure 3 At position 3 shown, denoted as A3, obtain the perspective image P of the phantom at this angle. A3 ;

[0040] Step 14: Apply the Hough transform to the projected image P A1 P A2 P A3 The sub-pixel centers of the four microsphere projections are extracted and denoted as (u Aa1 ,v Aa1 ), (u Ab1 ,v Ab1 ), (u Ac1 ,v Ac1 ), (u Ad1 ,v Ad1 ),

[0041] (u Aa2 ,v Aa2 ), (u Ab2 ,v Ab2 ), (u Ac2 ,v Ac2 ), (u Ad2 ,v Ad2 ), (u Aa3 ,v Aa3 ), (u Ab3 ,v Ab3 ), (u Ac3 ,v Ac3 ), (u Ad3 ,v Ad3 ).

[0042] Step 2: Using the coordinates of the projection center of the microsphere obtained in Step 1, calculate the coordinates of the projection points of the two points on the rotation axis. The lines connecting the projections of spheres a and b at positions A1, A2, and A3 intersect at a single point, denoted as: (u A-up ,v A-up ); The lines connecting the projections of spheres c and d at positions A1, A2, and A3 intersect at a single point, denoted as: (u A-down ,v A-down The intersection of the two phases is the projection of the two points on the axis of rotation.

[0043]

[0044]

[0045] Step 3: Using the intersection point coordinates obtained in Step 2, calculate the in-plane deflection angle γ of the detector. A :

[0046]

[0047] Step 4: Using the motion axis, move the rotary table to positions B, C, D, and E respectively, and repeat steps 1 to 3 to calculate γ for each position. B γ C γ D γ E Then the in-plane deflection angle γ of the detector can be expressed as:

[0048]

[0049] Step 5: Calculation of the distance SOD between the X-ray source and the rotary stage axis, and the distance SDD between the X-ray source and the detector. The specific calculation process includes:

[0050] Step 51: When the rotary table is at position A, use the coordinates of the projection points of the rotation axis on the detector obtained in step 2 to calculate the projection point spacing h. pro-A :

[0051]

[0052] Step 52: When the rotary table is located at positions B, C, D, and E respectively, the corresponding projection distance of the rotation axis is recorded as h. pro-B h pro-C h pro-D h pro-E ;

[0053] Step 53: Establish a mathematical relationship model between the standard positioning attitude parameters, phantom parameters, and rotation axis projection distance, thereby realizing the calculation of the distance SOD between the X-ray source and the rotation axis of the rotary table, and the distance SDD between the X-ray source and the detector, as shown in the following formula:

[0054]

[0055] Where, l is the distance between the lines containing spheres a and b and the lines containing spheres c and d in the model, as shown below. Figure 2 As shown; Δ is the distance between positions A, B, C, D, and E, as... Figure 3 As shown.

[0056] Step 6: Using the known parameters such as the center of the projection circle of sphere a at positions A1, B1, C1, D1, and E1, and the parameters of the industrial CT system, calculate the detector offsets u and v. The specific steps are as follows:

[0057] Step 61: Transform the coordinates of the center of sphere a at each obtained position using the following formula:

[0058]

[0059] Among them, u ia1 v ia1 Let x and y be the coordinates of the center of the projection of sphere a when the rotary table is at position i. Here are the x and y coordinates of the center of the projected circle of the transformed sphere a;

[0060] Step 62: Establish a mathematical relationship model between the coordinates of the projection center of sphere a, the detector offset, and the parameters of the industrial CT system, and realize the calculation of detector offsets u and v.

[0061]

[0062] Where PixelSize is the detector pixel size, Resolution is the number of pixels in a column or row of the detector, and d is an intermediate variable that does not require a solution.

Claims

1. A microsphere phantom-based industrial CT pose parameter calibration method, comprising the following steps: (1) Moving the rotating table to the initial position (denoted as position A) through the motion axis, placing the microsphere phantom on the rotating table, setting appropriate imaging parameters to obtain the perspective image of the microsphere phantom, and extracting the microsphere projection circle center. Specifically, it comprises: ①The microsphere model plane is perpendicular to the direction of the ray beam as much as possible (this position is recorded as A1), the industrial CT system is turned on, and the perspective image P of the model at this angle is obtained A1 ; The centers of the four microspheres in the projection image P A1 are sub-pixel extracted by using Hough transform, recorded as (u Aa1 , v Aa1 ), (u Ab1 , v Ab1 ), (u Ac1 , v Ac1 ), (u Ad1 , v Ad1 ) ② Rotate the stage counterclockwise by 60° (this position is denoted as A2) and obtain the perspective image P of the model at this angle. A2 Using the Hough transform to apply the projection image P A2 The sub-pixel centers of the four microsphere projections are extracted, denoted as (u Aa2 ,v Aa2 ), (u Ab2 ,v Ab2 ), (u Ac2 ,v Ac2 ), (u Ad2 ,v Ad2 ); ③ Rotate the stage 120° clockwise (this position is recorded as A3), and acquire the perspective image P of the model at this angle A3 ; use Hough transform to extract the sub-pixel centers of the four microspheres in the projection image P A3 , recorded as (u Aa3 ,v Aa3 ), (u Ab3 ,v Ab3 ), (u Ac3 ,v Ac3 ), (u Ad3 ,v Ad3 ) (2) Calculate the coordinates of the projection point of the rotating table's rotation axis on the detector at position A; the connecting lines of the projection figures of the a, b spheres at positions A1, A2, A3 intersect at a point, and the coordinates are denoted as: (u A-up ,v A-up ); the connecting lines of the projection figures of the c, d spheres at positions A1, A2, A3 intersect at a point, and the coordinates are denoted as: (u A-down ,v A-down ); the connecting line of the two intersection points represents the projection of the rotation axis on the detector; (u A-up ,v A-up ), (u A-down ,v A-down ) can be calculated by the following formula: (3) Calculate the relative angle of rotation axis and detector when the rotation table is at position A, that is, the in-plane angle of the detector at this time, denoted as γ A : (4) Using the motion axis, move the rotating table to positions B, C, D, E, repeat steps (1)-(3), and calculate γ B , γ C , γ D , γ E , then the in-plane deflection angle γ of the detector can be obtained. (5) Calculation of the distance SOD between the ray source and the rotating table rotation axis, and the distance SDD between the ray source and the detector. (6) Calculation of the detector offset u, v.

2. The method of claim 1, wherein, Step (5) specifically includes: When the rotating table is located at position A, the distance between the straight line on which the a and b balls are located and the straight line on which the c and d balls are located is denoted as h pro-A : Similarly, when the rotating table is located at positions B, C, D and E, the corresponding projection intervals are h pro-B , h pro-C , h pro-D , h pro-E ; ③ Establishing a mathematical relationship model among the pose parameters, the phantom parameters, and the projection image, the distance SOD between the ray source and the rotating table rotation axis, and the distance SDD between the ray source and the detector can be obtained, as follows: Wherein, l is the distance between the straight line where a, b balls are located and the straight line where c, d balls are located, and Δ is the distance between positions A, B, C, D, and E.

3. The method of claim 1, wherein, Step (6) specifically includes: ① Transform the obtained ball a circle center coordinates when the rotating table is at each position using the following formula: wherein u ia1 , v ia1 are the horizontal and vertical coordinates of the center of the circle of the projection of the sphere a when the rotating table is in position i, are the horizontal and vertical coordinates of the center of the circle of the projection of the sphere a after the corresponding transformation. ② Establish a mathematical relationship model among the ball a projection circle center coordinates, the detector offset, and the industrial CT system parameters, and realize the calibration of the detector offset u, v. Wherein, PixelSize is the detector pixel size, Resolution is the number of pixels included in a certain column or row of the detector, and d is an intermediate variable that does not need to be solved.

Images