Calculation method for projection position of rotating shaft of CL system

Abstract

The invention relates to the technical field of CL imaging, in particular to a CL system rotating shaft projection position calculation method. The method comprises the steps that a sample is placed on an objective table, the height is adjusted to ensure complete projection capture, the objective table is controlled to rotate by a circle to collect a plurality of projection images with marked angles and coordinates at equal angle intervals, all the projections are superposed to generate a composite image, and fixed feature points form an elliptical orbit; delimiting a row coordinate interval containing effective sample projection points in the row direction of the composite image as a candidate symmetry axis search range, traversing the row position of each hypothetical symmetry axis in the range, constructing a symmetric row pair combination, and calculating the sum of pixel value differences of all row pairs under each hypothetical symmetry axis; and selecting the row coordinate with the minimum difference sum as the projection position of the rotating shaft in the Y direction of the detector. According to the method, high-precision positioning is realized through symmetry analysis of the projection superposition image, a die body does not need to be calibrated, the stability of a calculation result is remarkably superior to that of a traditional technology, and the definition of a reconstructed image is effectively improved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of CL imaging, and particularly relates to a CL system rotating shaft projection position calculation method. BACKGROUND

[0002] Computer laminography (CL) has important application in three-dimensional imaging and internal defect detection of plate-shaped parts; the accuracy of the rotating center of the CL system detector has important influence on the accuracy of the CL reconstructed image; one coordinate of the rotating center is the coordinate corresponding to the projection line of the system rotating shaft; inaccurate rotating shaft leads to inaccurate rotating center, which further causes geometric artifacts in the CL reconstructed image, and seriously affects the quality of the reconstructed image; at present, the method for solving the problem is mainly a geometric analysis method; however, in actual application, the projection position of the rotating shaft calculated by the method is affected by the imaging sample, and the projection position of the rotating shaft calculated under different samples has a certain deviation.

[0003] A combined scanning CL imaging method is disclosed in Chinese Invention Patent No. CN113533392B, which comprises the following steps: S1: constructing a linear-circular combined scanning CL imaging system and obtaining a geometric imaging model; S2: linear scanning: determining the equivalent projection angle θ of linear scanning and calculating the step distance of the X-ray source and the flat panel detector, and the system magnification ratio K; S3: circular scanning: determining the circular scanning offset angle and calculating the movement distance of the X-ray source and the movement distance of the flat panel detector; S4: respectively collecting a group of projection data in a group of linear scanning and circular scanning processes, and using the SIRT iterative algorithm to reconstruct the collected projection data.

[0004] With the wide application of the CL system in industrial detection, material analysis and other scenes, higher requirements are put forward for the accuracy and adaptability of the rotating shaft positioning, especially when there is no calibration phantom or the system has assembly errors, a more robust solution needs to be sought. SUMMARY

[0005] The application aims at the problems in the background art and provides a CL system rotating shaft projection position calculation method.

[0006] The technical scheme of the application is a CL system rotating shaft projection position calculation method, which comprises the following specific implementation steps:

[0007] S1, placing a sample on a CL system sample stage, adjusting the scanning platform height to ensure that the projection of the sample at all rotation angles is completely captured by the flat panel detector;

[0008] S2. Control the stage to rotate one revolution, and acquire I projection images at equal intervals within the range of 0° to 360°. Each image is labeled with a projection angle number and pixel coordinates, where the projection image at the i-th angle is denoted as P. i (x,y), where x is the detector's X-axis coordinate variable, y is the detector's Y-axis coordinate variable, and i = 1, 2, 3, ..., I;

[0009] S3. Superimpose all I projected images to generate a composite image P(x,y): After being superimposed, the fixed feature points in the sample will form an elliptical trajectory;

[0010] S4. Let the direction of the minor axis of the ellipse in the superimposed image be denoted as row N, and the other direction of the superimposed image be denoted as column M. Select the row coordinate interval [N]. chushi N jieshu [N] serves as the search range for candidate symmetry axes; chushi N represents the starting row coordinates of the candidate symmetry axis search interval; jieshu The coordinates of the final row of the candidate symmetry axis search interval;

[0011] S5. Traverse each row K within the candidate interval, taking the Kth row as the assumed axis of symmetry, and within the interval [N... chushi N jieshu Construct T symmetric row pairs within the [construction context]; where a row pair is defined as (x, 2K-x), and T = KN. chushi ;

[0012] S6. For each candidate axis of symmetry K, calculate the sum of pixel value differences for all symmetrical row pairs:

[0013]

[0014] Among them, D K This represents the sum of pixel value differences for all symmetrical row pairs when the Kth row is the axis of symmetry.

[0015] S7. Compare D corresponding to all K within the candidate range. K Value, select one that makes D K Minimum row coordinate K opt The rotation axis is the projected position coordinate of the detector in the Y direction, i.e., the rotation axis projection position coordinate.

[0016] Preferably, in step S4, the interval [N] chushi N jieshu There exists at least one row of pixels that contains a valid sample projection point at a certain projection angle.

[0017] Preferably, the candidate symmetry axis search interval [N] in S4 chushi N jieshu] represents the range [N / 4, N / 3] of the total number of rows N in the image row direction.

[0018] Preferably, the method for determining the rotation axis projection position coordinates in step S7 is as follows: traverse all integer row coordinates K within the candidate interval, and calculate the D corresponding to each K. K The value is determined by selecting the K coordinate corresponding to the global minimum as K. opt .

[0019] Compared with the prior art, the above-mentioned technical solution of the present invention has the following beneficial technical effects:

[0020] This invention designs a method for calculating the projection position of the rotation axis in a CL system. By analyzing the symmetry of the superimposed projection images, the method dynamically locates the projection position of the rotation axis, effectively eliminating the influence of system assembly errors. Furthermore, it achieves high-precision positioning without relying on a calibration model, significantly reducing implementation costs and improving adaptability. Based on the calculation principle of minimizing pixel differences in symmetrical rows, the stability of the calculated rotation axis projection position is greatly improved. When testing different samples under the same system parameters, the fluctuation range of the projection position is significantly reduced. Finally, image reconstruction using this position can significantly reduce motion artifacts, improve the resolution and clarity of the reconstructed image, and provide a reliable guarantee for the imaging quality of the CL system. Attached Figure Description

[0021] Figure 1 This is a flowchart of a method for calculating the projection position of the rotation axis of a CL system proposed in this invention;

[0022] Figure 2 To reconstruct the image by projecting it onto the detector center as the rotation axis;

[0023] Figure 3 The image is reconstructed based on the calculated rotation axis projection position. Detailed Implementation

[0024] Example 1, as Figure 1 As shown, the present invention proposes a method for calculating the projection position of the rotation axis of a CL system, which includes the following specific implementation steps:

[0025] S1. Place the sample on the CL system stage and adjust the height of the scanning platform to ensure that the projection of the sample at all rotation angles is completely captured by the flat panel detector.

[0026] S2. Control the stage to rotate one revolution, and acquire multiple projected images of the sample at equal angular intervals. Each image is labeled with a projection angle number and pixel coordinates, i.e.:

[0027] The CL system is controlled to rotate the stage and acquire projected images of the sample at equal intervals within the range of 0° to 360°.

[0028] Let the total number of projection angles be I (e.g., I = 360), and let the projected image at the i-th angle be denoted as P. i (x,y); where x is the detector's X-axis coordinate variable (column direction); y is the detector's Y-axis coordinate variable (row direction); i is the projection number, i.e., i = 1, 2, 3, ..., I;

[0029] S3. Superimpose all I projected images to generate a composite image P(x,y):

[0030]

[0031] After superposition, the fixed feature points in the sample will form an elliptical trajectory, with its minor axis direction corresponding to the axis of symmetry of the projection of the rotation axis.

[0032] S4. Define the row coordinate center region in the synthesized image as the candidate symmetry axis search range, requiring that at least one row of pixels within this range contains the projection of a valid sample at a certain projection angle:

[0033] Let the total number of rows in the composite image be N (Y direction) and the total number of columns be M (X direction). That is, let the direction of the minor axis of the ellipse in the superimposed image be denoted as the row N of the superimposed image, and let the other direction of the superimposed image be denoted as the column M.

[0034] Select the row coordinate range [N] chushi N jieshu (In this embodiment, [N / 4, N / 3, i.e., the image center region]) is used as the candidate symmetry axis search range;

[0035] Where, N chushi N represents the starting row coordinates of the candidate symmetry axis search interval; jieshu The coordinates of the final row of the candidate symmetry axis search interval;

[0036] It should be noted that N chushi and N jieshu The value of is required to ensure that, at least at one projection angle, one row of projected pixels contains valid projected points of the sample, i.e., N. chushi and N jieshu The corresponding row cannot be a projection point without attenuation at all projection angles;

[0037] S5. Traverse the position of each hypothetical axis of symmetry row within the candidate range, and construct a pair of symmetric rows based on that row (each pair consists of the current row and its mirror image):

[0038] Iterate through each row K (K∈[N]) within the candidate interval chushi N jieshu Let the Kth row be the axis of symmetry;

[0039] For each K, in the interval [N]chushi Constructing symmetric row pairs within [K]: A row pair is defined as (x, 2K-x), where x is the coordinate of the current row, generating a total of T = KN. chushi T is the total number of symmetric row pairs constructed for candidate symmetry axis K;

[0040] S6. For each hypothetical axis of symmetry, calculate the sum of pixel value differences between all symmetrical row pairs. This sum reflects the degree of symmetry deviation. That is, for each candidate axis of symmetry K, calculate and sum the differences between all symmetrical row pairs:

[0041]

[0042] Among them, D K This represents the sum of pixel value differences for all symmetrical row pairs when the Kth row is the axis of symmetry.

[0043] S7. Compare the sum of differences corresponding to all hypothetical axes of symmetry within the candidate range, and select the row position with the smallest sum of differences as the final rotation axis projection coordinates:

[0044] Iterate through all K values ​​within the candidate interval and calculate the difference D for all symmetric row pairs. K ;

[0045] Select D K Minimum row coordinate K opt As the final axis of symmetry, the coordinate K of this row... opt This refers to the coordinates of the projection of the rotation axis onto the detector's Y-axis.

[0046] Example 2: The present invention proposes a method for calculating the projection position of the rotation axis of a CL system. The method for verifying the effectiveness of this method in certain specific scenarios is as follows:

[0047] Using the method for calculating the projection position of the rotation axis of a CL system proposed in Example 1, no special calibration model is required. The projection position of the rotation axis is calculated by the symmetry of the superimposed projection image. Under the condition that the system parameters remain unchanged, projection data of multiple circuit boards are collected and the projection position of the rotation axis is calculated. The results are shown in Table 1.

[0048] Table 1 shows the results of calculating the projection position of the rotation axis while keeping system parameters constant and collecting data from different circuit boards.

[0049]

[0050] Compared with existing technologies, this method calculates a smaller deviation in the rotation axis projection position and yields more stable results. Image reconstruction is performed on the resolution card projection data using both the detector center as the rotation axis projection position and the calculated rotation axis projection position, respectively. Figure 2 and Figure 3As shown, the image is reconstructed by using the calculated rotation axis projection position, resulting in a clearer reconstructed image.

[0051] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. A method for calculating a projection position of a rotating shaft of a CL system, characterized by, The specific implementation steps include the following: S1, place the sample on the CL system stage, adjust the scanning platform height, and ensure that the projection of the sample at all rotation angles is completely captured by the flat panel detector; S2, control the rotation of the stage, in the range of 0° to 360°, at equal intervals of angle, collect I projection images, mark the projection angle number and pixel coordinates for each image, wherein the projection image of the i-th angle is denoted as P i (x,y), x is the X-axis coordinate variable of the detector, y is the Y-axis coordinate variable of the detector, i = 1, 2, 3,..., I; S3, superimpose all I projection images to generate a composite image P(x, y): After superimposition, the fixed feature points in the sample will form an elliptical trajectory; S4, in the composite image in which the short axis direction of the ellipse in the superimposed image is recorded as the row N of the superimposed image, and the other direction of the superimposed image is recorded as the column M of the image, the row coordinate interval [N chushi ,N jieshu ] is selected as the search range of the candidate symmetry axis; N chushi is the starting row coordinate of the search interval of the candidate symmetry axis; N jieshu is the ending row coordinate of the search interval of the candidate symmetry axis; S5. Traverse each row K within the candidate interval, taking the Kth row as the assumed axis of symmetry, and within the interval [N... chushi N jieshu Construct T symmetric row pairs within the [construction context]; where a row pair is defined as (x, 2K-x), and T = KN. chushi ; S6, for each candidate symmetry axis K, calculate the total sum of pixel value differences of all symmetric row pairs: wherein D K is the sum of the pixel value difference of all symmetric row pairs when the Kth row is the symmetry axis; S7, compare D corresponding to all K in the candidate range K value, select the K that makes D K the minimum row coordinate K opt the projection position coordinate of the rotation axis in the Y direction of the detector, i.e. the projection position coordinate of the rotation axis.

2. The method of claim 1, wherein, The interval [N chushi ,N jieshu ] in step S4 exists at least one row of pixels containing valid sample projection points at a certain projection angle.

3. The method of claim 1, wherein, S4 candidate axis search interval [N chushi ,N jieshu ] is the range of [N / 4, N / 3] of the total number of image row direction N.

4. The method of claim 1, wherein, The determination manner of the rotation axis projection position coordinate in step S7 is: traversing all integer row coordinates K in the candidate interval, calculating the D value corresponding to each K, and selecting the K coordinate corresponding to the global minimum value as the K K . opt ​

Images