Registration method of core CT scan image and spliced ​​SEM image

Abstract

The present invention discloses a registration method for a core CT scanning image and a spliced ​​SEM image, which belongs to the field of core imaging and image processing technology, and includes: S1, making a positioning mark; S2, obtaining an X-CT three-dimensional image; S3, obtaining a spliced ​​SEM image of a section; S4, based on the positioning mark, obtaining a section image of a height corresponding to the spliced ​​SEM image on the X-CT three-dimensional image; S5, registering the spliced ​​SEM image and the section image to obtain multiple registration points; S6, determining the tilt axis based on the registration points; S7, rotating the X-CT three-dimensional image around the tilt axis to obtain a rotated image; S8, registering the rotated image with the spliced ​​SEM image, and the section image with the largest number of matching points is the one with the most suitable rotation angle. The present invention solves the problem of difficulty in determining the position of a two-dimensional image in a three-dimensional image, and further solves the problem of imaging plane tilt caused by difficulty in maintaining the same plane in different imaging methods.

Description

Technical Field

[0001] The invention relates to a registration method for a core CT scanning image and a spliced ​​SEM image, and belongs to the technical field of core imaging and image processing. Background Art

[0002] Core imaging technology has always been a hot topic in the field of rock analysis and has been widely used. Among them, X-ray CT imaging technology can obtain three-dimensional images of cores, which has been developed and mature, and is of great help for the analysis of the pore structure of cores. The stitched SEM image, also known as the MAPS image, is to scan thousands of small images of the same size with ultra-high resolution in the selected area, and use the small images to stitch into an ultra-high resolution, ultra-large area two-dimensional backscattered electron image, which has the characteristics of high resolution and wide field of view. By combining the advantages of two different imaging methods and matching the two images one by one in position, a more comprehensive analysis can be provided for the core, and the spatial distribution characteristics of the core pores and components can be described more accurately. However, the two images have the characteristics of different imaging methods, different resolutions, and different imaging dimensions, and it is difficult for general image registration algorithms to complete the registration of the two images. At the same time, the slight tilt of the instrument during imaging will cause the MAPS imaging plane to be not parallel to the XY plane of CT imaging, so it is necessary to obtain the oblique section of the CT three-dimensional image. Moreover, when performing MAPS imaging, the core section needs to be cut and polished to obtain a smooth section. There will be a certain tilt during the cutting process, so the registration process is more complicated.

[0003] In order to determine the position and tilt angle of the MAPS two-dimensional image in the CT three-dimensional image, the present invention provides a registration method for the core CT scanning image and the spliced ​​SEM image to solve the above-mentioned problems.

[0004] It should be noted that the above contents belong to the technical knowledge of the inventor and do not necessarily constitute prior art. Summary of the invention

[0005] The purpose of the present invention is to solve the problems existing in the prior art and provide a registration method for core CT scan images and spliced ​​SEM images, which solves the problem of imaging plane tilt caused by the difficulty in maintaining the same plane during experiments with different imaging methods and improves the registration accuracy.

[0006] The present invention achieves the above-mentioned purpose by adopting the following technical solutions:

[0007] The registration method of the core CT scanning image and the spliced ​​SEM image includes the following steps:

[0008] S1. Obtain a cylindrical core sample and make a positioning mark on the surface of the core sample;

[0009] S2, performing X-ray CT scanning on the core sample to obtain an X-CT three-dimensional image of the core sample;

[0010] S3. Cut the core sample to obtain a smooth section, perform SEM backscattered electron scanning imaging on the section, and obtain an ultra-high resolution and ultra-large area spliced ​​SEM image of the section;

[0011] S4, based on the positioning mark, obtaining a cross-sectional image at a height h0 position corresponding to the spliced ​​SEM image on the X-CT three-dimensional image;

[0012] S5, registering the spliced ​​SEM image obtained in step S3 and the cross-section image obtained in step S4 to obtain a plurality of registration points meeting the registration accuracy;

[0013] S6, determining the tilt axis of the core sample based on the position of the registration point obtained in step S5;

[0014] S7, rotating the X-CT three-dimensional image of the core sample around the tilt axis, with a rotation angle range of -β to β and a rotation step length of Δβ, to obtain a total of N rotation images;

[0015] S8. Use the SIFT image registration algorithm to register the N rotated images obtained in step S7 with the stitched SEM image, set a suitable matching threshold and perform manual screening. The rotated image with the largest number of matching points is the section image with the most suitable rotation angle.

[0016] Optionally, the method for making positioning marks on the core sample in step S1 is:

[0017] Three non-intersecting linear notches are made on the outer surface of the core sample, one of which is perpendicular to the upper end surface of the core sample, and the angles between the other two notches and the upper end surface of the core sample are α and -α respectively, and α≠90°.

[0018] Optionally, the specific method for obtaining the cross-sectional image in step S4 is: first, determine the height h0 of the imaging position of the spliced ​​SEM image in the X-CT three-dimensional image through the traces of the scratches on the spliced ​​SEM image, and then obtain the cross-sectional image at the height h0 on the X-CT three-dimensional image.

[0019] Optionally, in step S5, the images acquired in steps S3 and S4 are registered using a SIFT image registration algorithm.

[0020] Optionally, the specific method for determining the tilt axis of the lithology sample in step S6 is: establish a coordinate system with the lower left corner of the horizontal plane of the section image as the coordinate origin, perform a function fitting on all the registration points, obtain the distribution equation y=kx+b, and rotate the section image around the center point (m 1 ,m 2) is rotated clockwise by an angle θ to obtain a horizontal straight line y′, where 0 = arctan k, and y′ is:

[0021]

[0022] Take y=y′ and z=h0 as the tilt axis of the X-CT three-dimensional image.

[0023] Optionally, in step S8, the size of the rotated image is A×B, where A=2m 1 , The pixel value of the rotated image point (i, j) is calculated using the nearest neighbor interpolation method.

[0024] Optionally, when the nearest neighbor interpolation method is used to calculate the pixel value of the rotated image point (i, j), six adjacent pixel points on the target point plane are selected for interpolation. Then the pixel value of the rotated image point (i, j) is:

[0025] N(i 0 ,j 0 )=(N 0 (i,j,h-1)+N 0 (i,j,h+1)

[0026] +N 0 (i,j+1,h)+N 0 (i,j-1,h)

[0027] +N 0 (i,j,h+1)+N 0 (i,j,h-1)) / 6

[0028] Among them, N 0 (i,j,h-1) is the gray value of point (i,j,h-1) in the CT image before rotation, y is the intercept of the inclined axis, h 0 It is the height of the selected section in the X-CT three-dimensional image.

[0029] The beneficial effects of the present invention include but are not limited to:

[0030] The registration method of the core CT scan image and the spliced ​​SEM image provided by the present invention has the following advantages: 1) the approximate position of the spliced ​​SEM image section in the core sample can be determined according to the positioning mark before imaging the core sample, which solves the problem of the difficulty in determining the position of the two-dimensional image in the three-dimensional image in the traditional core image registration technology. 2) the most matching corresponding section in the X-CT three-dimensional image is obtained by image transformation and SIFT image registration method, which solves the problem of imaging plane tilt caused by the difficulty of maintaining the same plane during experiments of different imaging methods, and improves the registration accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0032] Figure 1 The X-CT three-dimensional image is obtained by scanning the core sample using X-ray CT three-dimensional scanning imaging technology;

[0033] Figure 2 To scan the core profile using SEM backscattered electron scanning imaging technology to obtain ultra-high resolution, ultra-large area two-dimensional backscattered splicing SEM images;

[0034] Figure 3 The cross-sectional image of the core section at the height h0 in the X-CT three-dimensional image is obtained by stitching the SEM image obtained by marking on the core sample before imaging;

[0035] Figure 4 The result of registering the preliminarily determined cross-section image and the spliced ​​SEM image using the SIFT image registration algorithm;

[0036] Figure 5 The SIFT registration results of 9 images with rotation angles ranging from -0.1° to -0.9°, with different numbers of registration points;

[0037] Figure 6 is the number of registration points obtained after SIFT registration at different tilt and rotation angles. When the tilt angle is -0.6°, the number of registration points is the largest, with 21. DETAILED DESCRIPTION

[0038] The present invention will be further described in detail in the following content. However, it should be pointed out that the following specific embodiments only provide specific operation examples of the present invention in an exemplary manner, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention is limited only by the claims. It is obvious to those skilled in the art that various other improvements and replacements can be made to the embodiments of the present invention within the protection scope defined by the claims of the present invention, and the same technical effects can still be achieved, thereby achieving the ultimate technical purpose of the present invention.

[0039] Embodiment 1:

[0040] The registration method of the core CT scan image and the spliced ​​SEM image provided in this embodiment has the following specific steps:

[0041] S1. Make positioning marks on the surface of the core sample. The specific method is: make three non-intersecting straight lines on the outer surface of the core sample, one of which is perpendicular to the upper end of the core sample, and the angles between the other two lines and the upper end of the core sample are 45° and -45° respectively. The two lines are etched in opposite directions. Figure 1 The positioning marks can be seen in the

[0042] S2, performing low-resolution X-ray CT three-dimensional scanning imaging on the core sample to obtain an X-CT three-dimensional image of the core sample; Figure 1 The acquired X-CT three-dimensional image contains 1650×1650×1200 pixels, with a resolution of 15.3μm and a 16-bit grayscale image format; Figure 1 The area with the smallest gray value is the pores, and the rest is the rock skeleton. Different gray levels represent different mineral components.

[0043] S3. Cut the core sample to obtain a smooth section, perform SEM backscattered electron scanning imaging on the obtained section, and obtain an ultra-high resolution and ultra-large area spliced ​​SEM image of the section, ensuring that the X-CT three-dimensional image and the section image are derived from the same core sample, and image registration can be performed; Figure 2 The obtained stitched SEM image contains 10500×10500 pixels with a resolution of 3μm. The area with the smallest gray value is the pore, and the other different gray values ​​represent different mineral components.

[0044] S4. Preliminary judgment is made on the corresponding positions of the X-CT three-dimensional image and the spliced ​​SEM image based on the positioning marks before imaging, and the height h0 of the spliced ​​SEM image in the X-CT three-dimensional image is determined; specifically, preliminary judgment is made based on the marks made on the core before imaging. Figure 3 It is a CT section with a height of 648. By judging the position of the groove line, it can be determined that this section is close to the position of the MAPS imaging section;

[0045] S5. Use the SIFT image registration algorithm to register the preliminarily determined corresponding spliced ​​SEM image and the cross-section image to obtain corresponding feature points that meet the registration accuracy; the registration result is as follows: Figure 4 As shown, the corresponding registration points are roughly distributed in a straight line;

[0046] S6. determining the tilt axis of the core sample based on the position characteristics of the corresponding registration points;

[0047] During CT imaging and MAPS imaging, the core sample will have a certain inclination. The two images have a high matching degree in the area close to the inclination axis. Therefore, the corresponding registration points are distributed near the inclination axis in a straight line. The coordinate system is established with the lower left corner of the horizontal plane of the section image as the origin of the coordinate, and a function fitting is performed on the distribution of the registration points. Figure 4 It can be seen that the registration points are basically distributed linearly. The fitting curve obtained by a fitting is y=0.0091x+695.31. A horizontal straight line can be obtained by rotating the cross-section image clockwise by an angle of θ, where θ=arctan k. The horizontal straight line obtained after rotation is:

[0048]

[0049] It is calculated that the image needs to be rotated 0.5° clockwise, and the tilted horizontal line after rotation is y=702.8. Therefore, y=703 and z=648 are taken as the tilt axis of the X-CT image in the three-dimensional coordinate system;

[0050] S7, rotating the X-CT three-dimensional image along the tilt axis, with the rotation angle ranging from -3° to 3° and a step length of 0.1°, and obtaining a total of 60 rotation images;

[0051] The rotation process will cause the image width to change. It is necessary to determine the size of the generated image two-dimensional matrix. The size of the rotated image is set to A×B, where A=2m 1 , The width is rounded down to an integer and then subtracted by 2 to facilitate subsequent interpolation operations. At the same time, a small reduction in edge pixels has no effect on the subsequent registration effect. Select six adjacent pixel points on the target point plane for interpolation, and the pixel value of the rotated image point (i, j) is:

[0052] N(i 0 ,j 0 )=(N 0 (i,j,h-1)+N 0 (i,j,h+1)

[0053] +N 0 (i,j+1,h)+N 0 (i,j-1,h)

[0054] +N 0 (i,j,h+1)+N 0 (i,j,h-1)) / 6

[0055] Among them, N 0 (i, j, h-1) is the grayscale value of point (i, j, h-1) in the CT image before rotation. y is the intercept of the inclined axis, h0 It is the height of the selected section in the X-CT three-dimensional image.

[0056] S8, registering the rotated image with the spliced ​​SEM image using the SIFT image registration algorithm, and selecting the best rotation angle based on the registration result; Figure 5 The SIFT registration results of 9 images with rotation angles ranging from -0.1° to -0.9°, with different numbers of registration points;

[0057] Figure 6 The figure shows the number of registration points obtained after SIFT registration at different tilt angles. When the tilt angle is -0.6°, the number of registration points is the largest, with 21. Therefore, -0.6° is taken as the optimal tilt angle to obtain the cross-sectional image in the X-CT three-dimensional image that is accurately registered with the spliced ​​SEM image.

[0058] The above specific implementation manner cannot be used as a limitation on the protection scope of the present invention. For those skilled in the art, any substitution, improvement or change made to the implementation manner of the present invention falls within the protection scope of the present invention.

[0059] The matters not described in detail in the present invention are all known technologies to those skilled in the art.

Claims

1. Registration method of core CT scan image and spliced ​​SEM image, It is characterized in that The steps include: S1. Obtain a cylindrical core sample and make a positioning mark on the surface of the core sample; S2, performing X-ray CT scanning on the core sample to obtain an X-CT three-dimensional image of the core sample; S3. Cut the core sample to obtain a smooth section, perform SEM backscattered electron scanning imaging on the section, and obtain an ultra-high resolution and ultra-large area spliced ​​SEM image of the section; S4, based on the positioning mark, obtaining a cross-sectional image at a height h0 position corresponding to the spliced ​​SEM image on the X-CT three-dimensional image; S5, registering the spliced ​​SEM image obtained in step S3 and the cross-section image obtained in step S4 to obtain a plurality of registration points meeting the registration accuracy; S6, determining the tilt axis of the core sample based on the position of the registration point obtained in step S5; S7, rotating the X-CT three-dimensional image of the core sample around the tilt axis, with a rotation angle range of -β to β and a rotation step length of Δβ, to obtain a total of N rotation images; S8. Use the SIFT image registration algorithm to register the N rotated images obtained in step S7 with the stitched SEM image, set a suitable matching threshold and perform manual screening. The rotated image with the largest number of matching points is the section image with the most suitable rotation angle.

2. The registration method of the core CT scan image and the spliced ​​SEM image according to claim 1, It is characterized in that The method for making positioning marks on the core sample in step S1 is: Three non-intersecting linear notches are made on the outer surface of the core sample, one of which is perpendicular to the upper end surface of the core sample, and the angles between the other two notches and the upper end surface of the core sample are α and -α respectively, and α≠90°.

3. The registration method of the core CT scan image and the spliced ​​SEM image according to claim 1, It is characterized in that The specific method of obtaining the cross-sectional image in step S4 is as follows: first, the height h0 of the imaging position of the spliced ​​SEM image in the X-CT three-dimensional image is determined by the traces of the scratches on the spliced ​​SEM image, and then the cross-sectional image at the height h0 is obtained on the X-CT three-dimensional image.

4. The registration method of the core CT scan image and the spliced ​​SEM image according to claim 1, It is characterized in that In step S5, the images acquired in steps S3 and S4 are registered using the SIFT image registration algorithm.

5. The registration method of the core CT scan image and the spliced ​​SEM image according to claim 1, It is characterized in that The specific method for determining the tilt axis of the lithology sample in step S6 is: establish a coordinate system with the lower left corner of the horizontal plane of the section image as the coordinate origin, perform a function fitting on all the registration points, and obtain the distribution equation y=kx+b, and divide the section image around the center point (m 1 ,m 2 ) is rotated clockwise by an angle θ to obtain a horizontal straight line y′, where θ = arctank, and y′ is: Take y=y′ and z=h0 as the tilt axis of the X-CT three-dimensional image.

6. The registration method of the core CT scan image and the spliced ​​SEM image according to claim 1, It is characterized in that In step S8, the size of the rotated image is A×B, where A=2m 1 , The pixel value of the rotated image point (i, j) is calculated using the nearest neighbor interpolation method.

7. The registration method of the core CT scan image and the spliced ​​SEM image according to claim 6, It is characterized in that When the nearest neighbor interpolation method is used to calculate the pixel value of the rotated image point (i, j), six adjacent pixel points on the target point plane are selected for interpolation. The pixel value of the rotated image point (i, j) is: N(i 0 ,j 0 )=(N 0 (i,j,h-1)+N 0 (i,j,h+1)+N 0 (i,j+1,h)+N 0 (i,j-1,h)+N 0 (i,j,h+1)+N 0 (i,j,h-1)) / 6 Among them, N 0 (i,j,h-1) is the gray value of point (i,j,h-1) in the CT image before rotation, y is the intercept of the inclined axis, h 0 It is the height of the selected section in the X-CT three-dimensional image.

Images