Blade wall thickness size sub-pixel level measurement method and system based on industrial CT image

Abstract

The application discloses a kind of based on industrial CT image blade wall thickness size subpixel level measurement method and system, method includes obtaining the scanning CT image of turbine blade part to be measured;The edge of turbine blade in scanning CT image is extracted using pixel level edge detection algorithm, and the coarse positioning edge position of blade is obtained;The edge point of coarse positioning edge position is interpolated using subpixel edge detection algorithm based on cubic spline interpolation, and the subpixel edge profile of blade is obtained;Least square method is used to curve fitting to subpixel edge profile, and the blade wall thickness size of blade part to be measured is measured based on fitting result.The application combines pixel level edge detection algorithm with subpixel edge detection algorithm based on cubic spline interpolation to process CT image, and the edge pixel point is promoted from pixel level to subpixel level, which not only guarantees the stability of measurement result, but also effectively realizes high-precision measurement of turbine blade size.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of turbine blade wall thickness measurement, and particularly to a blade wall thickness size sub-pixel level measurement method and system based on industrial CT images. BACKGROUND

[0002] Turbine blades are important components of an aero-engine power system, mainly composed of a crown, a blade body, a rim plate and a tenon. Since they need to work at a temperature higher than 1000 DEG C, in order to improve the temperature resistance of the turbine blade, advanced aero-engine turbine blades mostly adopt a thin-walled hollow structure design, and the inner cavity is used as a gas cooling channel to realize the composite gas cooling of the blade. The wall thickness is an important geometric size index of the turbine blade, which directly affects the safety and life of the engine, so the wall thickness detection has always been a hot spot in the research of turbine blade size detection methods.

[0003] In actual production process, the requirement of the blade wall thickness size is also particularly strict, and the thickness error of the inner and outer surface normal directions at different interfaces should not exceed 0.1mm. At present, the edge extraction algorithm used in the traditional image measurement method is mainly a pixel level algorithm, which has the disadvantages of inaccurate positioning and low measurement precision. In order to further improve the measurement precision of the blade wall thickness size in the image, the present application provides a blade wall thickness size sub-pixel level measurement method and system based on industrial CT images. SUMMARY

[0004] The purpose of the present application is to provide a blade wall thickness size sub-pixel level measurement method and system based on industrial CT images, which can effectively measure the wall thickness of the turbine blade with high precision.

[0005] In order to achieve the above purpose, the present application provides the following scheme:

[0006] A blade wall thickness size sub-pixel level measurement method based on industrial CT images, the method comprising:

[0007] Obtaining a two-dimensional slice of a three-dimensional voxel model of a turbine blade, and obtaining a scanning CT image of a to-be-measured part based on the two-dimensional slice;

[0008] Extracting the edge of the turbine blade in the scanning CT image by using a pixel level edge detection algorithm to obtain the coarse positioning edge position of the turbine blade;

[0009] Performing interpolation processing on the edge points of the coarse positioning edge position by using a sub-pixel edge detection algorithm based on cubic spline interpolation to obtain the sub-pixel edge contour of the turbine blade;

[0010] The sub-pixel edge profile is curve-fitted by using a least square method, and a blade wall thickness size of the to-be-measured part of the turbine blade is measured based on a fitting result.

[0011] Optionally, the edge of the turbine blade in the scan CT image is extracted by using a pixel-level edge detection algorithm to obtain a coarse positioning edge position of the turbine blade, and specifically includes:

[0012] determining a structural element of image processing;

[0013] the scan CT image is subjected to corrosion processing by using a corrosion algorithm based on the structural element to obtain an inner edge image of the turbine blade;

[0014] the scan CT image is subjected to inflation processing by using an inflation algorithm based on the structural element to obtain an outer edge image of the turbine blade;

[0015] the scan CT image is subjected to inflation-corrosion processing by using an inflation and corrosion algorithm based on the structural element to obtain a gradient edge image of the turbine blade;

[0016] the scan CT image is subjected to edge detection processing by using a morphological edge detection algorithm combining inflation and closing operation based on the structural element to obtain a first morphological edge image of the turbine blade;

[0017] the scan CT image is subjected to edge detection processing by using a morphological edge detection algorithm combining corrosion and opening operation based on the structural element to obtain a second morphological edge image of the turbine blade;

[0018] the inner edge image, the outer edge image, the gradient edge image, the first morphological edge image and the second morphological edge image are fused to obtain a fusion image containing the coarse positioning edge position.

[0019] Optionally, the inner edge image is represented as:

[0020] D1=A-(AΘS)

[0021] wherein D1 represents the inner edge image; A represents the scan CT image; S represents the structural element; and Θ represents the corrosion operation.

[0022] Optionally, the outer edge image is represented as:

[0023]

[0024] wherein D2 represents the outer edge image; A represents the scan CT image; S represents the structural element; and ⊕ represents the inflation operation.

[0025] ​Optionally, the gradient edge image is represented as:

[0026]

[0027] wherein D3 represents the gradient edge image; represents the dilation operation; and represents the erosion operation.

[0028] Optionally, the first morphological edge image is represented as

[0029]

[0030] wherein D4 represents the first morphological edge image; A represents the scan CT image; and S represents the structural element; represents the dilation operation; and represents the closing operation.

[0031] Optionally, the second morphological edge image is represented as

[0032]

[0033] wherein D5 represents the second morphological edge image; and represents the erosion operation; represents the opening operation.

[0034] Optionally, the expression for calculating the fusion image is

[0035]

[0036] wherein R represents the fusion image; A represents the scan CT image; and S represents the structural element; represents the opening operation; represents the dilation operation; represents the closing operation; and represents the erosion operation.

[0037] The application further provides a turbine blade wall thickness size sub-pixel level measurement system based on an industrial CT image, and the system comprises:

[0038] an image acquisition module, configured to acquire a two-dimensional slice of a three-dimensional voxel model of a turbine blade, and obtain a scan CT image of a to-be-measured part based on the two-dimensional slice;

[0039] an edge coarse positioning module, configured to extract an edge of the turbine blade in the scan CT image by using a pixel-level edge detection algorithm, and obtain a coarse positioning edge position of the turbine blade;

[0040] an edge fine positioning module, configured to perform interpolation processing on edge points of the coarse positioning edge position by using a cubic spline interpolation-based sub-pixel edge detection algorithm, and obtain a sub-pixel edge contour of the turbine blade.

[0041] The blade wall thickness measurement module is configured to perform curve fitting on the sub-pixel edge contour by using a least square method, and measure the blade wall thickness size of the to-be-measured part of the turbine blade based on the fitting result.

[0042] Optionally, the edge coarse positioning module comprises:

[0043] A structure element determination unit is configured to determine a structure element for image processing.

[0044] An erosion processing unit is configured to perform erosion processing on the scanned CT image by using an erosion algorithm based on the structure element, and obtain an inner edge image of the turbine blade.

[0045] An expansion processing unit is configured to perform expansion processing on the scanned CT image by using an expansion algorithm based on the structure element, and obtain an outer edge image of the turbine blade.

[0046] An expansion-erosion processing unit is configured to perform expansion-erosion processing on the scanned CT image by using an expansion and erosion algorithm based on the structure element, and obtain a gradient edge image of the turbine blade.

[0047] An expansion-closing operation processing unit is configured to perform edge detection processing on the scanned CT image by using a morphological edge detection algorithm combined with expansion and closing operation based on the structure element, and obtain a first morphological edge image of the turbine blade.

[0048] An erosion-opening operation processing unit is configured to perform edge detection processing on the scanned CT image by using a morphological edge detection algorithm combined with erosion and opening operation based on the structure element, and obtain a second morphological edge image of the turbine blade.

[0049] An image fusion unit is configured to fuse the inner edge image, the outer edge image, the gradient edge image, the first morphological edge image and the second morphological edge image, and obtain a fusion image containing the coarse positioning edge position.

[0050] According to the embodiments of the present application, the following technical effects are provided.

[0051] The application provides a turbine blade wall thickness size sub-pixel level measurement method and system based on an industrial CT image. BRIEF DESCRIPTION OF DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0053] Figure 1 A turbine blade wall thickness size sub-pixel level measurement method flow chart based on an industrial CT image is provided for the embodiment 1 of the present application. DETAILED DESCRIPTION

[0054] The technical solutions in the embodiments of the present application will be described clearly and completely with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without any creative effort are within the protection scope of the present application.

[0055] The purpose of the present application is to provide a turbine blade wall thickness size sub-pixel level measurement method and system based on an industrial CT image, which selects a sub-pixel level edge detection method to improve the measurement accuracy, and can effectively measure the wall thickness of the turbine blade with high precision. And the image size measurement has high precision and good stability.

[0056] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application will be further described in detail with reference to the drawings and specific embodiments.

[0057] Embodiment 1

[0058] As Figure 1 shown, the embodiment provides a blade wall thickness size sub-pixel level measurement method based on industrial CT image, the method comprises:

[0059] S1: obtaining a two-dimensional slice of a three-dimensional voxel model of a turbine blade, and obtaining a scanned CT image of a to-be-measured part based on the two-dimensional slice.

[0060] The step-by-step scanning method is adopted by using an X-ray tomography imaging device. The test sample is fixed on a rotating sample stage, and then the ray tube is turned on to transmit the to-be-tested sample. After the X-ray is attenuated by the to-be-tested sample, the information collected by the detector is converted into real-time two-dimensional projection images. After the projection images are reconstructed, the sample tomography images are obtained. After three-dimensional reconstruction of a plurality of continuous tomography images, a three-dimensional voxel model of the sample is obtained, which is then exported to a two-dimensional slice, so as to realize scanning of the to-be-measured part and obtain the CT image of the to-be-measured part.

[0061] The model of the X-ray tomography imaging device is industrial CT-X5000.

[0062] The device parameters corresponding to the scanning of the X-ray tomography imaging device are as follows: the acceleration voltage used is 190kV, the current is 120μA, the exposure time is 1000ms, 1440 projection images are taken by rotating 360°, and the voxel size after reconstruction is 16.2μm.

[0063] S2: using a pixel-level edge detection algorithm to extract the edge of the turbine blade in the scanned CT image to obtain the coarse positioning edge position of the turbine blade.

[0064] The pixel-level edge detection algorithm is used to extract the image edge, and the edge position of the image is roughly positioned.

[0065] Specifically, step S2 comprises:

[0066] S21: determining the structure element of image processing.

[0067] The "structure element" is a professional term, which is a unit structure for processing an image, such as a distribution of some 0, 1 in a gray-scale image.

[0068] S22: using an erosion algorithm based on the structure element to perform erosion processing on the scanned CT image to obtain an inner edge image of the turbine blade.

[0069] The to-be-processed industrial CT image is denoted as A, and S is a suitable structure element. First, set A to be eroded by S (here, the definition of the erosion algorithm), and then let the image A be subtracted from it to obtain the inner edge image D1.

[0070] The inner edge image is represented as:

[0071] D1=A-(AΘS)

[0072] Wherein, D1 represents the inner edge image; A represents the scanning CT image; S represents the structure element; and Θ represents the erosion operation.

[0073] S23: based on the structure element, the scanning CT image is dilated by using a dilation algorithm to obtain an outer edge image of the turbine blade.

[0074] The outer edge image D2 can be obtained by using the dilation algorithm on the scanning CT image.

[0075] The outer edge image is represented as:

[0076]

[0077] Wherein, D2 represents the outer edge image; A represents the scanning CT image; S represents the structure element; represents the dilation operation.

[0078] S24: based on the structure element, the scanning CT image is dilated and eroded by using a dilation and erosion algorithm to obtain a gradient edge image of the turbine blade.

[0079] The gradient edge image D3 can be obtained by using the dilation and erosion algorithm on the scanning CT image.

[0080] The gradient edge image is represented as:

[0081]

[0082] Wherein, D3 represents the gradient edge image.

[0083] S25: based on the structure element, the scanning CT image is edge detected by using a morphological edge detection algorithm combined with dilation and closing operation to obtain a first morphological edge image of the turbine blade.

[0084] The morphological edge detection algorithm combined with dilation and closing operation can suppress the low valley noise, and the edge image D4 can be obtained by using the morphological edge detection algorithm on the scanning CT image.

[0085] The first morphological edge image is represented as:

[0086]

[0087] Wherein, D4 represents the first morphological edge image; A represents the scanning CT image; S represents the structure element; and '·' represents the closing operation.

[0088] S26: performing edge detection processing on the scanning CT image based on the structural element using a morphological edge detection algorithm combining erosion and opening operation to obtain a second morphological edge image of the turbine blade.

[0089] The morphological edge detection algorithm combining erosion and opening operation can suppress peak noise to obtain an edge image D5.

[0090] The second morphological edge image is represented as:

[0091]

[0092] D5 represents the second morphological edge image; represents an opening operation.

[0093] The algorithms of steps S22 to S26 are all for obtaining the edge contour line of the blade, but the accuracy is different.

[0094] S27: fusing the inner edge image, the outer edge image, the gradient edge image, the first morphological edge image and the second morphological edge image to obtain a fused image containing the coarse positioning edge position.

[0095] The morphological operations are combined according to the suppression characteristics of different noises to obtain a pixel-level edge detection algorithm.

[0096] The expression for calculating the fused image is:

[0097]

[0098] R represents the fused image, which is a fusion of the first five algorithms, and the previous algorithm introduction is to obtain R.

[0099] S3: performing interpolation processing on the edge points of the coarse positioning edge position using a sub-pixel edge detection algorithm based on cubic spline interpolation to obtain a sub-pixel edge contour of the turbine blade.

[0100] The algorithm of the previous image R obtains the specific position information of the pixel-level edge contour point of the image, and the cubic spline interpolation algorithm is used to obtain the sub-pixel level accuracy, which is higher in the position information accuracy of the edge contour point.

[0101] The following is a theoretical introduction of the cubic spline interpolation method:

[0102] (1) Definition of cubic spline function

[0103] The cubic spline function S(x j() refers to a function defined on the interval [a, b], given n+1 nodes and a set of corresponding function values, where the function satisfies:

[0104] (I) S(x) is satisfied at each node j )=f(x j (j = 0, 1, ..., n-1);

[0105] (II) It has continuous second derivatives on [a, b];

[0106] (III) In each interval [x j x j+1 If S(x) is a cubic polynomial on (j = 0, 1, ..., n-1), then S(x) is called a cubic polynomial. j ) cubic spline interpolation function.

[0107] (2) Cubic spline interpolation polynomial

[0108] The cubic spline interpolation function S(x) is a piecewise cubic polynomial. The goal is to find S(x) in each subinterval [x...]. j x j+1 The above requires determining four undetermined parameters. If S is used... j (x) represents that it is in the i-th subinterval [x j x j+1 The expression on ], then

[0109] S j (x)=a j0 +a j1 x+a j2 x 2 +a j3 x 3 (j = 0, 1, ..., n-1)

[0110] The cubic polynomial S(x) and its first derivative S′(x) and second derivative S″″(x) are continuous over the entire interpolation interval. Boundary conditions are added, and the coefficients are calculated.

[0111] (3) Cubic spline interpolation method

[0112] Given the second derivative values ​​at both endpoints:

[0113] S″(x0)=f″(x0)=y″0, S″(x n )=f″(x n )=f″ n

[0114] The special case is as follows:

[0115] S″(x0)=S″(x n ) = 0

[0116] The above equation is called natural boundary condition. The cubic spline interpolation function satisfying the natural boundary condition is called natural spline interpolation function.

[0117] Using the second derivative value S"(x) = M j (j = 0, 1,..., n) to express S(x), since S(x) is a cubic polynomial on the interval [x j , x j+1 ], S"(x) is a linear function on [x j , x j+1 ], which can be expressed as:

[0118]

[0119] Integrating S"(x) twice and using S(x j ) = y j and S(x j+1 ) = y j+1 , the integral constant can be determined, and the cubic spline expression is obtained:

[0120]

[0121] j = 0, 1,..., n-1

[0122] Here M j (j = 0, 1,..., n) is unknown, in order to determine M j (j = 0, 1,..., n), the derivative of S(x) is obtained:

[0123]

[0124] From which M

[0125]

[0126] Similarly, the expression of S(x) on the interval [x j-1 , x j ] can be obtained, and then:

[0127]

[0128] Using S'(x j +0) = S'(x j -0) can be obtained:

[0129] μ j M j-1 +2M j + λ j M j+1 = d j , j = 1, 2,..., n-1

[0130] wherein, j = 0, 1, …, n-1

[0131] For the natural spline interpolation function, we can let M0= y"0, M n = y" n Then a linear equation group about M1, M2, … M n-1 n-1 can be obtained, that is, a sub-pixel edge detection algorithm based on cubic spline interpolation:

[0132]

[0133] S4: curve fitting is performed on the sub-pixel edge profile by using the least square method, and the blade wall thickness size of the to-be-measured part of the turbine blade is measured based on the fitting result.

[0134] The least square method formula is as follows:

[0135]

[0136] The above formula can be simplified as:

[0137] XB = Y

[0138] Therefore, the coefficient matrix B is as follows:

[0139] B = (X' x X)-1x X' x Y.

[0140] In this embodiment, the original picture data is obtained by using the step scanning method of the X-ray tomography imaging device, and the obtained data is processed by combining the pixel-level edge detection algorithm and the sub-pixel edge detection algorithm based on cubic spline interpolation. On the basis of greatly reducing the instability of the measurement result, the high-precision measurement of the blade size is effectively realized.

[0141] In order to verify the conclusion, Table 1 shows the turbine blade wall thickness data measured by the traditional image measurement method, and Table 2 shows the turbine blade wall thickness data measured by the sub-pixel image measurement method. According to the comparison, it can be found that the measurement result obtained by the sub-pixel precision measurement of the blade wall thickness size based on the industrial CT image is more accurate.

[0142] Table 1: Turbine blade wall thickness values measured by the traditional image measurement method

[0143]

[0144] Table 2: Turbine blade wall thickness values measured by the sub-pixel image measurement method

[0145]

[0146]

[0147] In this embodiment, after image preprocessing of the original image, a pixel-level algorithm is used to extract the edge by using mathematical morphology method, and then a sub-pixel edge detection algorithm based on cubic spline interpolation is used to obtain the sub-pixel edge contour. Finally, the least square method is used to fit the edge curve. Compared with the traditional image measurement method and the pixel-level image measurement method, the results show that the method used in this application can effectively measure the wall thickness of the blade with high precision, and has high precision and good stability in image size measurement.

[0148] Embodiment 2

[0149] The embodiment provides a blade wall thickness size sub-pixel level measurement system based on an industrial CT image, and the system comprises:

[0150] An image acquisition module T1 is configured to acquire a two-dimensional slice of a three-dimensional voxel model of a turbine blade, and obtain a scanning CT image of a to-be-measured part based on the two-dimensional slice.

[0151] An edge coarse positioning module T2 is configured to extract an edge of the turbine blade in the scanning CT image by using a pixel-level edge detection algorithm, and obtain a coarse positioning edge position of the turbine blade.

[0152] Specifically, the edge coarse positioning module T2 comprises:

[0153] A structure element determination unit is configured to determine a structure element for image processing.

[0154] A corrosion processing unit is configured to perform corrosion processing on the scanning CT image by using a corrosion algorithm based on the structure element, and obtain an inner edge image of the turbine blade.

[0155] An expansion processing unit is configured to perform expansion processing on the scanning CT image by using an expansion algorithm based on the structure element, and obtain an outer edge image of the turbine blade.

[0156] An expansion combined with corrosion processing unit is configured to perform expansion and corrosion processing on the scanning CT image by using an expansion and corrosion algorithm based on the structure element, and obtain a gradient edge image of the turbine blade.

[0157] An expansion combined with closed operation processing unit is configured to perform edge detection processing on the scanning CT image by using a morphological edge detection algorithm combined with expansion and closed operation based on the structure element, and obtain a first morphological edge image of the turbine blade.

[0158] An erosion and opening operation processing unit is configured to perform edge detection processing on the scan CT image based on a morphological edge detection algorithm using erosion and opening operation combination of the structural element, to obtain a second morphological edge image of the turbine blade.

[0159] An image fusion unit is configured to fuse the inner edge image, the outer edge image, the gradient edge image, the first morphological edge image and the second morphological edge image to obtain a fusion image containing the coarse positioning edge position.

[0160] An edge fine positioning module T3 is configured to perform interpolation processing on the edge points of the coarse positioning edge position by using a sub-pixel edge detection algorithm based on cubic spline interpolation, to obtain a sub-pixel edge contour of the turbine blade.

[0161] A blade wall thickness measurement module T4 is configured to perform curve fitting on the sub-pixel edge contour by using a least square method, and measure the blade wall thickness size of the to-be-measured part of the turbine blade based on the fitting result.

[0162] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments, and the same or similar parts between various embodiments can be referred to each other. For the system disclosed by the embodiments, since it corresponds to the method disclosed by the embodiments, the description is relatively simple, and the relevant part can be referred to the method part.

[0163] The principles and implementation manners of the present application are described by using specific examples in the specification, and the above embodiment description is only used to help understand the method of the present application and its core idea; meanwhile, for the general technical personnel in the art, according to the idea of the present application, the specific implementation manner and application range can be changed. In conclusion, the content of the specification should not be understood as the limitation of the present application.

Claims

1. A method for sub-pixel level measurement of blade wall thickness based on industrial CT images, characterized in that, The method includes: Two-dimensional slices of a three-dimensional voxel model of a turbine blade are obtained, and CT scan images of the part to be tested are obtained based on the two-dimensional slices. The edges of the turbine blades in the scanned CT image are extracted using a pixel-level edge detection algorithm to obtain the coarse positioning edge position of the turbine blades; A subpixel edge detection algorithm based on cubic spline interpolation is used to interpolate the edge points at the coarsely located edge position to obtain the subpixel edge contour of the turbine blade. The least squares method is used to perform curve fitting on the sub-pixel edge contour, and the blade wall thickness of the test part of the turbine blade is measured based on the fitting result. Specifically, a pixel-level edge detection algorithm is used to extract the edges of the turbine blades in the scanned CT image to obtain the coarse positioning edge location of the turbine blades, including: Determine the structural elements of image processing; Based on the structural elements, the scanning CT image is eroded using an erosion algorithm to obtain the inner edge image of the turbine blade; Based on the structural elements, the scanned CT image is dilated using a dilation algorithm to obtain the outer edge image of the turbine blade; Based on the structural elements, the scanned CT image is subjected to dilation and erosion processing using dilation and erosion algorithms to obtain the gradient edge image of the turbine blade; Based on the structuring elements, an edge detection algorithm combining dilation and closure operations is used to perform edge detection processing on the scanned CT image to obtain the first morphological edge image of the turbine blade. Based on the structural elements, an edge detection algorithm combining erosion and opening operations is used to perform edge detection processing on the scanned CT image to obtain the second morphological edge image of the turbine blade. The inner edge image, the outer edge image, the gradient edge image, the first morphological edge image, and the second morphological edge image are fused to obtain a fused image that includes the coarse positioning edge position.

2. The method according to claim 1, characterized in that, The inner edge image is represented as follows: Where D1 represents the inner edge image; A represents the scanned CT image; and S represents the structural element. This represents the erosion operation.

3. The method according to claim 1 or 2, characterized in that, The outer edge image is represented as follows: Where D2 represents the outer edge image; A represents the scanned CT image; and S represents the structural element. This indicates the expansion operation.

4. The method according to claim 3, characterized in that, The gradient edge image is represented as follows: Where D3 represents the gradient edge image; This represents the expansion operation; This represents the erosion operation.

5. The method according to claim 1 or 4, characterized in that, The first morphological edge image represents Where D4 represents the first-morphological edge image; A represents the scanned CT image; and S represents the structural element. This indicates the expansion operation; ' indicates the closing operation.

6. The method according to claim 5, characterized in that, The second morphological edge image is represented as: Wherein, D5 represents the second-morphological edge image; This represents the erosion operation; ' indicates the opening operation.

7. The method according to claim 1 or 6, characterized in that, The expression for calculating the fused image is: Where R represents the fused image; A represents the scanned CT image; S represents the structuring element; ' indicates the opening operation; This indicates the expansion operation; ' indicates the closing operation; This represents the erosion operation.

8. A sub-pixel-level measurement system for blade wall thickness based on industrial CT images, as described in any one of claims 1 to 7, characterized in that, The system includes: The image acquisition module is used to acquire two-dimensional slices of the three-dimensional voxel model of the turbine blade, and obtain scanned CT images of the part to be tested based on the two-dimensional slices. The edge coarse localization module is used to extract the edges of the turbine blades in the scanned CT image using a pixel-level edge detection algorithm, and to obtain the coarse localization edge position of the turbine blades. The edge fine positioning module is used to perform interpolation processing on the edge points at the coarse positioning edge position using a sub-pixel edge detection algorithm based on cubic spline interpolation to obtain the sub-pixel edge contour of the turbine blade. The blade wall thickness measurement module is used to perform curve fitting on the sub-pixel edge contour using the least squares method, and to measure the blade wall thickness of the part to be measured of the turbine blade based on the fitting result. The edge coarse positioning module includes: Structural element determination unit, used to determine the structural elements of image processing; The corrosion processing unit is used to perform corrosion processing on the scanned CT image based on the structural elements using a corrosion algorithm to obtain the inner edge image of the turbine blade. An expansion processing unit is used to perform expansion processing on the scanned CT image based on the structural element using an expansion algorithm to obtain the outer edge image of the turbine blade. An expansion and corrosion processing unit is used to perform expansion and corrosion processing on the scanned CT image based on the structural elements using expansion and corrosion algorithms to obtain the gradient edge image of the turbine blade. An expansion-closure operation processing unit is used to perform edge detection processing on the scanned CT image based on the structuring element using a morphological edge detection algorithm that combines expansion and closure operations, to obtain the first morphological edge image of the turbine blade. The erosion-opening operation processing unit is used to perform edge detection processing on the scanned CT image based on the structural element using a morphological edge detection algorithm that combines erosion and opening operations, so as to obtain the second morphological edge image of the turbine blade. An image fusion unit is used to fuse the inner edge image, the outer edge image, the gradient edge image, the first morphological edge image, and the second morphological edge image to obtain a fused image containing the coarse positioning edge position.

Images