Large-view CT (Computed Tomography) rotation center correction method

Abstract

The invention relates to the technical field of nondestructive testing, in particular to a large-view CT (Computed Tomography) rotation center correction method which comprises the following steps: S1, acquiring a rotation center search range of a CT image; s2, selecting a rotation center position in a rotation center searching range; s3, taking the rotation center position as a center line, obtaining a left-right symmetrical CT image area, and respectively calculating the sum of column values of gray values of pixels in the CT image area; s4, calculating a difference value of the sum of column values of the left and right symmetrical CT image areas, and then calculating an absolute value of the difference value; s5, repeatedly executing the steps S3-S4, obtaining absolute values of the multiple groups of difference values, and obtaining the sum of the absolute values; s6, repeatedly executing the steps S2-S5 until the sum of the set number of rotation center positions and the corresponding absolute values is obtained; and S7, performing polynomial fitting on the absolute value sum array, calculating the minimum value of a fitting polynomial in the specified area, and outputting the rotation center position corresponding to the minimum value. According to the invention, the problem of rotation center correction failure caused by incomplete projection drawing can be solved.

Description

Technical Field

[0001] The invention relates to the technical field of nondestructive testing, and in particular to a method for correcting a rotation center of a large-viewing-field CT. Background Art

[0002] Conventional linear array CT systems generally use the third generation CT scanning and reconstruction method, and its fan beam can cover the entire object. Its projection image has complete bilateral sinusoidal boundaries. The center of rotation can be automatically measured through the central symmetry characteristics. For example, the center of rotation is obtained by calculating the grayscale center of gravity of the image, and the center of rotation is obtained by image boundary extraction and symmetry operation. However, for large-field CT scanning, its fan beam can only cover most of the workpiece, such as Figure 1 Its projection diagram has only one-sided sine line, as shown in Figure 2 As shown in the figure, the traditional rotation center correction algorithm based on the projection image of the third generation CT scan fails due to the incomplete and asymmetric characteristics of the projection image. Therefore, it is necessary to design a dedicated rotation center correction method based on its data characteristics. Summary of the invention

[0003] The purpose of the present invention is to provide a large-viewing-field CT rotation center correction method to solve the problem that the traditional rotation center correction algorithm fails due to incomplete projection images.

[0004] In order to achieve the above object, a method for calibrating the rotation center of a large-viewing-field CT is provided, comprising the following steps:

[0005] S1, obtaining the rotation center search range of the CT image;

[0006] S2. Select the rotation center position within the rotation center search range;

[0007] S3, taking the rotation center position as the center line, obtaining a bilaterally symmetrical CT image area, and respectively calculating the sum of the column values ​​of the grayscale values ​​of each pixel in the CT image area;

[0008] S4, calculating the difference between the sum of column values ​​of the left-right symmetrical CT image regions, and then calculating the absolute value of the difference;

[0009] S5, repeatedly execute S3-S4, obtain the absolute values ​​of multiple groups of difference values, and sum them to obtain the sum of the absolute values;

[0010] S6, repeatedly executing S2-S5 until a set number of rotation center positions and the sum of the corresponding absolute values ​​are obtained;

[0011] S7. Perform polynomial fitting on the absolute value sum array, calculate the minimum value of the fitting polynomial in the specified area, and output the rotation center position corresponding to the minimum value.

[0012] Furthermore, before obtaining the rotation center search range of the CT image in step S1, the following steps are also included:

[0013] S0. Correcting basic parameters, specifically including the following steps:

[0014] S101, select the initial position C of the rotation center init ;

[0015] S102, set range width W c , determine the rotation center search range [C init -W c ,C init +W c ];

[0016] S003. Set the left data width W and the right data width W of the CT image area for difference calculation.

[0017] Further, selecting the rotation center position within the rotation center search range in step S2 includes the following steps:

[0018] Select the rotation center position C in turn within the rotation center search range i , C i =C i ±im;

[0019] Among them, i∈[0,n), n is a natural number, m is the pixel width, C i ∈[C init -W c ,C init +W c ].

[0020] Further, the step S3 specifically includes the following steps:

[0021] S301, construct a coordinate system with the length and width of the CT image as the Y and X axes respectively, and the rotation center position C i is the center line parallel to the Y axis; the CT image area that is symmetrical about the center line is a rectangular image area;

[0022] S302, respectively obtain the sum of the column values ​​of the left and right CT image regions LeftSum j 、RightSum j , where j represents the distance between the column and the center of rotation, and also represents the width of the CT image area, which ranges from (0, W].

[0023] Further, the calculation formula for calculating the absolute value of the difference in step S4 is as follows:

[0024] AbsSub j =abs(LeftSumj -RightSum j )

[0025] The calculation formula of the sum of absolute values ​​in step S5 is as follows:

[0026]

[0027] The calculation formula for performing polynomial fitting on the absolute value sum array in step S7 is as follows:

[0028] Sum(C i )=a4C i 4 +a3C i 3 +a2C i 2 +a1C i +a0

[0029] Calculate the fitted polynomial in the region [C init -W c ,C init +W c ], and obtain its corresponding rotation center position C min , which is the precise CT rotation center position.

[0030] Principles and advantages:

[0031] 1. The basic principle of this scheme comes from the geometric relationship of large field of view correction data compensation, that is, there is a certain degree of correlation between the data on both sides of the center line of the projection image. The correlation is that the sum of the two columns of data symmetrical to the rotation center position should be the same. Therefore, the more accurate the rotation center is, the smaller the difference in the sum of the two columns of data symmetrical to the left and right of the corresponding center line.

[0032] 2. The projection data of the wide-field CT scanning method is missing data compared to the projection data of the third-generation CT scanning. The missing data causes the projection image to not have rotation center symmetry. This makes the traditional method of automatically calculating the rotation center based on the symmetry of the projection data not applicable to the projection data of the wide-field CT scanning. In response to the above problems, this solution only uses the data in a certain area near the rotation center of the projection image to calculate the rotation center. It can solve the problem of missing data that the traditional method cannot handle, and can realize the automatic calculation of the rotation center of the wide-field CT scanning. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a schematic diagram of a scanning method of a large-viewing-field CT scan in the prior art;

[0034] Figure 2 for Figure 1 Projection images of medium and large field of view CT scan;

[0035] Figure 3 This is a flowchart of a method for calibrating a rotation center of a large-viewing-field CT according to an embodiment of the present invention;

[0036] Figure 4 Schematic diagram of the principle of large field of view rotation center correction;

[0037] Figure 5 It is a schematic diagram of a bilaterally symmetrical CT image region and the sum of its column values. DETAILED DESCRIPTION

[0038] The following is further described in detail through specific implementation methods:

[0039] Example

[0040] A method for calibrating the rotation center of a large field of view CT is basically as follows Figure 3 , Figure 4 , Figure 5 As shown, the following steps are included:

[0041] S1, obtaining the rotation center search range of the CT image;

[0042] When the rotation center search range of the CT image is obtained in step S1, the basic correction parameters are set, which specifically includes the following steps:

[0043] S101, select the initial position C of the rotation center init ;

[0044] S102, set range width W c , determine the rotation center search range [C init -W c ,C init +W c ];

[0045] Set the left data width W and right data width W of the CT image area for difference calculation.

[0046] S2, selecting the rotation center position within the rotation center search range; the step S2 of selecting the rotation center position within the rotation center search range comprises the following steps:

[0047] Select the rotation center position C in turn within the rotation center search range i , C i =C i ±im; indicates selecting the rotation center position C in the positive direction i Or select the rotation center position C in reverse order i ; Step S2 is executed once, and the rotation center position C is selected once i .

[0048] Among them, i∈[0,n), n is a natural number, and m is the pixel width (indicates that the rotation center position C is selected in turn i amplitude), C i ∈[C init -W c ,C init +W c ].

[0049] S3, taking the rotation center position as the center line, obtaining a bilaterally symmetrical CT image area, and respectively calculating the sum of the column values ​​of the grayscale values ​​of each pixel in the CT image area; the step S3 specifically includes the following steps:

[0050] S301, construct a coordinate system with the length and width of the CT image as the Y and X axes respectively, and the rotation center position C i is the center line parallel to the Y axis; the CT image area that is symmetrical about the center line is a rectangular image area;

[0051] S302, respectively obtain the sum of the column values ​​of the left and right CT image regions LeftSum j 、RightSum j , where j represents the distance between the column and the rotation center, and also represents the width value of the CT image area, and its range is (0, W]. When step S3 is executed once, the value of j changes once, indicating that the area size of the CT image area changes, that is, the number of pixels changes.

[0052] S4, calculating the difference between the sum of column values ​​of the left-right symmetrical CT image regions, and then calculating the absolute value of the difference; the calculation formula for calculating the absolute value of the difference in step S4 is as follows:

[0053] AbsSub j =abs(LeftSum j -RightSum j )

[0054] S5. Repeat S3-S4 to obtain the absolute values ​​of multiple groups of differences, and sum them to obtain the sum of the absolute values. The calculation formula for the sum of the absolute values ​​in step S5 is as follows:

[0055]

[0056] S6, repeatedly executing S2-S5 until a set number of rotation center positions and the sum of the corresponding absolute values ​​are obtained;

[0057] S7, perform polynomial fitting on the absolute value sum array, calculate the minimum value of the fitting polynomial in the specified area, and output the rotation center position corresponding to the minimum value. The calculation formula for performing polynomial fitting on the absolute value sum array in step S7 is as follows:

[0058] Sum(C i )=a n C i n +...+a4C i 4 +a3C i 3 +a2C i 2 +a1C i +a0

[0059] Among them, a n ,a n-1 ,...,a2,a1,a0 are constant coefficients, C i is a variable, n is the highest degree of the polynomial, and in this embodiment, n is 4.

[0060] Calculate the fitted polynomial in the region [C init -W c ,C init +W c ], and obtain its corresponding rotation center position C min , which is the precise CT rotation center position.

[0061] 1. The basic principle of this scheme comes from the geometric relationship of large field of view correction data compensation, that is, there is a certain degree of correlation between the data on both sides of the center line of the projection image. The correlation is that the sum of the two columns of data symmetrical to the rotation center position should be the same. Therefore, the more accurate the rotation center is, the smaller the difference in the sum of the two columns of data symmetrical to the left and right of the corresponding center line.

[0062] 2. This scheme can calculate the rotation center by only using the data of a certain area near the rotation center of the projection image, which can well solve the data missing problem that traditional methods cannot handle, and can realize the automatic calculation of the rotation center of large-field CT scanning.

[0063] The above is only an embodiment of the present invention. The common sense such as the known specific structure and characteristics in the scheme is not described in detail here. The ordinary technicians in the relevant field know all the common technical knowledge in the technical field of the invention before the application date or priority date, can obtain all the existing technologies in the field, and have the ability to apply the conventional experimental means before that date. The ordinary technicians in the relevant field can improve and implement this scheme in combination with their own abilities under the enlightenment given by this application. Some typical known structures or known methods should not become obstacles for ordinary technicians in the relevant field to implement this application. It should be pointed out that for those skilled in the art, without departing from the structure of the present invention, several deformations and improvements can be made, which should also be regarded as the protection scope of the present invention, which will not affect the effect of the implementation of the present invention and the practicality of the patent. The protection scope required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A method for calibrating the rotation center of a large field of view CT, characterized in that: The following steps are involved: S1, obtaining the rotation center search range of the CT image; S2. Select the rotation center position within the rotation center search range; S3, taking the rotation center position as the center line, obtaining a bilaterally symmetrical CT image area, and respectively calculating the sum of the column values ​​of the grayscale values ​​of each pixel in the CT image area; S4, calculating the difference between the sum of column values ​​of the left-right symmetrical CT image regions, and then calculating the absolute value of the difference; S5, repeatedly execute S3-S4, obtain the absolute values ​​of multiple groups of difference values, and sum them to obtain the sum of the absolute values; S6, repeatedly executing S2-S5 until a set number of rotation center positions and the sum of the corresponding absolute values ​​are obtained; S7. Perform polynomial fitting on the absolute value sum array, calculate the minimum value of the fitting polynomial in the specified area, and output the rotation center position corresponding to the minimum value.

2. The method for calibrating the rotation center of a large-viewing-field CT according to claim 1, characterized in that: When obtaining the rotation center search range of the CT image in step S1, the following steps are specifically included: S101, select the initial position C of the rotation center init ; S102, set range width W c , determine the rotation center search range [C init -W c ,C init +W c ]; Set the left data width W and right data width W of the CT image area for difference calculation.

3. The method for calibrating the rotation center of a large field of view CT according to claim 2, characterized in that: Selecting the rotation center position within the rotation center search range in step S2 includes the following steps: Select the rotation center position C in turn within the rotation center search range i , C i =C i ±im; Among them, i∈[0,n), n is a natural number, m is the pixel width, C i ∈[C init -W c ,C init +W c ].

4. The method for calibrating the rotation center of a large-viewing-field CT according to claim 3, characterized in that: The step S3 specifically includes the following steps: S301, construct a coordinate system with the length and width of the CT image as the Y and X axes respectively, and the rotation center position C i is the center line parallel to the Y axis; the CT image area that is symmetrical about the center line is a rectangular image area; S302, respectively obtain the sum of the column values ​​of the left and right CT image regions LeftSum j 、RightSum j , where j represents the distance between the column and the center of rotation, and also represents the width of the CT image area, which ranges from (0, W].

5. The method for calibrating the rotation center of a large field of view CT according to claim 4, characterized in that: The calculation formula for calculating the absolute value of the difference in step S4 is as follows: AbsSub j =abs(LeftSum j -RightSum j ) The calculation formula of the sum of absolute values ​​in step S5 is as follows: The calculation formula for performing polynomial fitting on the absolute value sum array in step S7 is as follows: Sum(C i )=a4C i 4 +a3C i 3 +a2C i 2 +a1C i +a0 Calculate the fitted polynomial in the region [C init -W c ,C init +W c ], and obtain its corresponding rotation center position C min , which is the precise CT rotation center position.

Images