Dual-phase steel microstructure image generation method, image generation device, program

Abstract

To enable generation of an image with two phases that are more clearly separated even when training data is not available.SOLUTION: An image generation method is provided, comprising binarizing a grayscale image showing a dual-phase microstructure to generate a first image, and performing smoothing and binarizing processing for smoothing and binarizing the first image to generate a second image.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a method, an apparatus, and a program for generating a structural image of a dual-phase steel. [Background technology]

[0002] Dual-phase stainless steel is a type of microstructural form that can be used for a wide range of applications, including thin plates, steel pipes, and structural materials. Dual-phase stainless steel is a steel that contains two different microstructural phases, such as ferrite-martensite steel, ferrite-bainite steel, and ferrite-austenite steel. Because dual-phase stainless steel contains two phases with different strengths and chemical properties, it can enjoy various benefits, such as achieving both strength and elongation and reducing cracking in specific directions by controlling the microstructural morphology.

[0003] The type and form of dual-phase stainless steel respond to alloy composition and heat treatment. For example, when the microstructure fraction of a dual-phase steel changes, the properties of the dual-phase steel change. Therefore, it is important to quantitatively understand the microstructure fraction of dual-phase stainless steel when developing and producing dual-phase steels.

[0004] Measurement methods for obtaining microstructural information on dual-phase stainless steels include electron backscattering diffraction (EBSD) and electron probe microanalyzer (EPMA), but these methods have the disadvantage of requiring long sample preparation and measurement times.

[0005] On the other hand, measurement methods that use images taken with an optical microscope require a short measurement time. When taking images with an optical microscope, applying a corrosive solution such as nital creates unevenness corresponding to the differences in the reactivity of each structure within the dual-phase steel. This allows the two phases of the dual-phase steel to be imaged as areas of different brightness. However, the image taken is affected by factors such as the intensity of the light source, making it difficult to obtain data with uniform image quality like EBSD or EPMA.

[0006] Techniques for processing images of metal structures and the like are disclosed in, for example, Patent Documents 1 to 5. [Prior art documents] [Patent documents]

[0007] [Patent Document 1] International Publication No. 2021 / 153633 [Patent Document 2] Patent Publication No. 2021-166002 [Patent Document 3] Japanese Patent Application Publication No. 143769 / 1983 [Patent Document 4] Japanese Patent Application Publication No. 3-150447 [Patent Document 5] Japanese Patent Application Laid-Open No. 2015-210648 Summary of the Invention [Problem to be solved by the invention]

[0008] However, the techniques disclosed in Patent Documents 1 and 2 require learning using training data, and the techniques disclosed in Patent Documents 3 and 5 may not be sufficient for processing images of dual-phase stainless steel. An object of the present invention is to provide a method, an image generating device, and a program for generating a structural image of a dual-phase steel, which can generate an image in which the two phases are more clearly separated, even in the absence of training data. [Means for solving the problem]

[0009] One aspect of the present invention is a method for generating a structural image of a dual-phase steel, which comprises binarizing a grayscale image showing the structure of the dual-phase steel to generate a first image, smoothing the first image, and further binarizing the first image by performing a smoothing binarization process to generate a second image. [Effects of the Invention]

[0010] According to the present invention, even in the absence of training data, an image in which two phases are more clearly separated can be generated. [Brief explanation of the drawings]

[0011] [Figure 1] 1 is a diagram showing the configuration of an image generation system according to a first embodiment. [Figure 2] 1 is a diagram illustrating a configuration of an image generating apparatus according to a first embodiment. [Figure 3] 4 is a flowchart showing the operation of the image generating device according to the first embodiment. [Figure 4] FIG. 2 is a diagram showing an example in which image processing according to the first embodiment is applied to an image of ferrite-bainite steel. [Figure 5] 10 is a flowchart showing the operation of the image generating device according to the second embodiment. [Figure 6] 1A is a diagram showing an example of a first image (a), a smoothed image (b) obtained by smoothing the first image, a second image (c), and a third image (d) when the dual-phase steel is a ferrite-austenite dual-phase steel. [Figure 7] FIG. 10 is a diagram showing an example in which the second embodiment is applied to ferrite-austenite steel. [Figure 8] FIG. 10 is a diagram showing an example in which the second embodiment is applied to ferrite-bainite steel. [Figure 9] FIG. 10 is a diagram showing a region for calculating complexity C. [Figure 10] 10 is a flowchart showing the operation of the image generating device according to the third embodiment. [Figure 11] FIG. 10 is a diagram showing an example in which the third embodiment is applied to ferritic-martensitic steel. DETAILED DESCRIPTION OF THE INVENTION

[0012] First Embodiment Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. 1 is a diagram showing the configuration of an image generation system 1 according to the first embodiment. The image generation system 1 includes an image generation device 10, an imaging device 20, and a storage device 30.

[0013] The image generating device 10 generates an image based on the image acquired from the imaging device 20, and outputs the image to the storage device 30. The storage device 30 stores the image input from the image generating device 10.

[0014] The imaging device 20 captures a grayscale image showing the structure of the dual-phase steel (hereinafter referred to as a dual-phase steel image). The imaging device 20 includes, for example, an optical microscope 21 and a camera 22, and captures the dual-phase steel image by having the camera 22 capture an image of the dual-phase steel magnified by the optical microscope 21 in a mode for capturing a grayscale image. The imaging device 20 may include, for example, the optical microscope 21, the camera 22, and a processing device for converting an image into a grayscale image, and the dual-phase steel image may be captured by having the camera 22 capture an image of the dual-phase steel magnified by the optical microscope 21 and converting the captured image into a grayscale image. Note that the imaging device in this embodiment is not limited to an optical microscope, and may also be an SEM or the like as long as it is possible to identify each phase and grain boundary based on the brightness of the phases.

[0015] 2 is a diagram showing the configuration of an image generating device 10 according to the first embodiment. The image generating device 10 includes an image acquiring unit 11, an image processing unit 12, and an image output unit 13.

[0016] The image acquisition unit 11 acquires the dual-phase steel image from the imaging device 20 .

[0017] The image processing unit 12 performs a binarization process on the dual-phase steel image to generate a first image. The image processing unit 12 performs a smoothing process and a binarization process on the first image to generate a second image. The process of smoothing an image and then binarizing the smoothed image is hereinafter referred to as smoothing binarization. The smoothing process is, for example, a process of applying a Gaussian filter to the image.

[0018] The second image is an image in which a specific region has been removed, for example, a linear region corresponding to a grain boundary in a dual-phase stainless steel.

[0019] The binarization process converts pixels with a brightness equal to or greater than a predetermined threshold to white, and pixels with a brightness less than the predetermined threshold to black. Hereinafter, numerical processing may be performed with white represented as 1 and black represented as 0. The conversion by the binarization process may be reversed. The threshold in the binarization process may be determined arbitrarily. For example, the threshold in the binarization process may be set to the brightness most frequently occurring in a dual-phase stainless steel image.

[0020] The binarization process may be adaptive binarization, which is a method of performing binarization by determining a binarization threshold for each pixel. The threshold for each pixel is calculated by subtracting a constant from the average brightness of the area surrounding the pixel. For example, the threshold T for the pixel at coordinate (x, y) is x、y is expressed by equation (1).

number

[0021] In equation (1), H is a constant, I x、y is the brightness of the pixel at coordinate (x, y), W dx、dy is the weight, i.e., T x、y is the value obtained by subtracting a constant H from the average luminance in a square of width 2L+1 around the pixel being calculated.

[0022] By setting the threshold value in the binarization process to the brightness that occurs most frequently in the dual-phase steel image, the shading can be made uniform in the binarization process, and the regions representing each phase can be more clearly separated by binary values.

[0023] Furthermore, adaptive binarization allows for better binarization even when the brightness of the dual-phase steel image varies from place to place.

[0024] By smoothing the first image, thin black lines in the first image become light grey and then become white through a subsequent binarization process. By smoothing the first image, thin white lines in the first image become dark grey and then become black through a subsequent binarization process.

[0025] The image output unit 13 outputs the first image and the second image to the storage device 30.

[0026] 3 is a flowchart showing the operation of the image generating device 10 according to the first embodiment. The image acquisition unit 11 acquires a dual-phase steel image from the imaging device 20 (step S11). The image processing unit 12 processes the dual-phase steel image to perform binarization processing and generate a first image (step S12). The image processing unit 12 performs smoothing binarization processing on the first image to generate a second image (step S13). The image output unit 13 outputs the second image (step S14). The image processing unit 12 may perform the smoothing binarization process multiple times to generate the second image. A user of the image generating device 10 may cause the image processing unit 12 to perform the smoothing binarization process until there are no thin lines in the second image. In the image obtained by repeating the smoothing and binarization processes, the thin white linear regions in the first image become black pixels and disappear. For example, when the duplex stainless steel is a ferrite-austenite dual-phase steel, the image generation method according to the first embodiment results in a second image in which the white linear secondary austenite disappears and only the primary austenite is extracted as a white region. In this case, the second image is an image in which the secondary austenite has been removed as a specific region from the first image.

[0027] FIG. 4 shows an example in which the image processing according to the first embodiment is applied to an image of ferrite-bainite steel. In the first image of the ferrite-bainite steel shown here, the parent phase, ferrite, is observed as white, and bainite as black. The ferrite grain boundaries appear as black lines. While each of these black grain boundary regions is not large, their sum overwhelms the entire image. Therefore, if the area fraction is calculated assuming the white regions as ferrite and the black regions as bainite, the bainite area fraction will be overestimated. FIG. 4 shows a first image (FIG. 4(b)) obtained by binarizing the original image (grayscale) (FIG. 4(a)), and a second image (FIG. 4(c)) obtained by smoothing and binarizing the first image. The thin black lines corresponding to the grain boundaries have disappeared in the second image. Based on this second image in which the ferrite grain boundaries have been removed, the area fractions of the bainite and ferrite phases can be accurately calculated.

[0028] The image processing unit 12 may remove noise from the first image before the smoothing binarization process by performing a noise reduction process on the first image. The noise reduction process converts small black areas within large white areas and small white areas within large black areas to black and white, respectively. This allows the image generation device 10 to generate an image that more accurately reflects the microstructure distribution of the dual-phase stainless steel.

[0029] The image processing unit 12 may perform principal component analysis on the second image to calculate the orientation of the components. The image processing unit 12 may also evaluate the properties of the dual-phase stainless steel based on the calculated orientation of the components. This allows the image generating device 10 to predict the properties of the dual-phase stainless steel. Second Embodiment

[0030] In addition to the functions of the image processing unit 12 in the first embodiment, the image processing unit 12 in the second embodiment generates a third image by performing a logical sum operation on two pixels at the same position in each of the first and second images. Hereinafter, the logical sum operation will also be referred to as an OR operation. Specifically, a pixel that is white in either the first image or the second image is set to be white in the corresponding pixel in the third image, or a pixel that is black in either the first image or the second image is set to be black in the corresponding pixel in the third image.

[0031] The image output unit 13 in the second embodiment outputs a third image.

[0032] The image processing unit 12 may perform principal component analysis on the third image to calculate the directions of the components, and may evaluate the properties of the dual-phase stainless steel based on the calculated directions of the components.

[0033] 5 is a flowchart showing the operation of the image generating device 10 according to the second embodiment. The operations from steps S21 to S23 are the same as the operations from steps S11 to S13 in the first embodiment. The image processing unit 12 generates a third image by performing a logical OR operation on each pixel of the first image and the second image (step S24). The image output unit 13 outputs the third image (step S25).

[0034] In the smoothing-binarization process, relatively thin linear regions tend to disappear. For example, a thin linear region that was white in the first image becomes partially black through re-binarization. By repeating this process, the thin linear region will completely disappear. Therefore, the image output unit 13 determines the color (white or black) to be output to the third image as a result of an OR operation between pixels, depending on the color (white or black) of the thin linear region that should be retained in the first image. In the first embodiment, when processing an image of a ferrite-austenite dual-phase stainless steel, thin white lines of secondary austenite can be eliminated by repeating the process of generating a second image from a first image. However, if thin dips (black pixel areas) exist in the primary austenite, these dips also disappear, the original shape of the primary austenite is lost, and this may affect the calculation of the area fraction. In the second embodiment, an OR operation is performed to return black pixels that were black in either the first or second image, thereby obtaining a third image in which the thin dips present in the primary austenite are not eliminated. FIG. 6 shows an example of a first image (a), a smoothed image (b) obtained by smoothing the first image, a second image (c), and a third image (d) when the duplex steel is a ferrite-austenite dual-phase steel. In the second image (FIG. 6(c)) obtained by applying the first embodiment to the ferrite-austenite steel, some areas of primary austenite that were separate in the first image (FIG. 6(a)) appear to be merged. Here, by applying the second embodiment and performing an OR operation to change black pixels in either the first or second image to black, the thin black area (ferrite) that existed between two adjacent primary austenite areas remains in the third image. This prevents the area from being mistakenly recognized as a single primary austenite area.

[0035] FIG. 7 is a diagram showing another example in which the second embodiment is applied to ferrite-austenite steel. In the example of FIG. 7, the white areas in the first image are austenite, and the black areas are ferrite. A second image is obtained by performing smoothing and binarization on the first image (FIG. 7(b)). Here, an OR operation is applied to set black pixels that were black in either the first image or the second image, resulting in a third image (FIG. 7(c)). The area of ​​white pixels in FIG. 7(c) is primary austenite. Next, a process is performed to set black pixels in the first image (FIG. 7(b)) that correspond to white pixels in the third image (FIG. 7(c)), resulting in FIG. 7(d). The area of ​​white pixels in FIG. 7(d) is secondary austenite. In this way, in this embodiment, the area ratios of the primary austenite and secondary austenite phases can be determined.

[0036] As a third example of the second embodiment, the case where it is applied to ferrite-bainite steel will be described below. As described in the first embodiment with reference to FIG. 4, in the first image of a ferrite-bainite steel, the parent phase, ferrite, is observed as white, and the bainite is observed as black. The ferrite grain boundaries appear as black lines. In the second image, the grain boundaries disappear, but any thin white linear regions within the bainite also disappear. Therefore, by applying an OR operation to generate a third image, corresponding pixels that were white in either the first or second image are converted to white in the corresponding pixels in the third image, the ferrite grain boundaries are eliminated, while the thin white linear regions within the bainite are retained. FIG. 8 shows an example in which the second embodiment is applied to a ferrite-bainite steel. In the third image on the far right, the black linear grain boundaries have been eliminated. It can be seen that the thin white linear regions that existed within the black region in the upper right corner of the third image are retained. Similar image generation can also be applied to ferritic-martensitic steels. Third Embodiment

[0037] In addition to the image processing unit 12 in the second embodiment, the image processing unit 12 in the third embodiment inverts the values ​​of the pixels included in an area where the area is made up of pixels that take one value in the first image and is surrounded by pixels that take the other value, based on the complexity of the shape of the area. The complexity C of a region's shape is calculated based on the perimeter and area of ​​the region. For example, the complexity C is calculated using the number of pixels P that are adjacent to pixels that take the other value among the pixels included in the region as the perimeter, and the number of pixels N included in the region as the area, as shown in formula (2).

number

[0038] Fig. 9 is a diagram showing an area for calculating complexity C. In Fig. 9, the area for which complexity C is calculated is the white area surrounded by black. In Fig. 9, the number of pixels N is the number of pixels in the white area, and the number of pixels P is the number of pixels that the line passes through.

[0039] The image processing unit 12 inverts the values ​​of the pixels in the region when the complexity C is equal to or greater than a predetermined value. For example, when the value of P is large, such as when the shape of the region is sawtooth, the complexity P becomes large.

[0040] FIG. 10 is a flowchart showing the operation of the image generating device 10 according to the third embodiment. The operations from steps S41 to S42 are the same as the operations from steps S21 to S22 in the second embodiment. The image processing unit 12 inverts a predetermined pixel in the first image based on complexity (step S43). The operations from steps S44 to S46 are the same as the operations from steps S23 to S25 in the second embodiment. Note that step S43 may be placed between steps S44 and S45. That is, after the first image is subjected to smoothing binarization processing to generate a second image, a predetermined pixel in the second image may be inverted.

[0041] For example, when the dual-phase steel is a ferrite-martensite steel, the contrast within the martensite structure is so large that some parts appear white, making it difficult to distinguish from the ferrite structure. In this case, the white martensite region has a complex shape, such as a sawtooth shape, at the boundary with the black region. As shown in Figure 11, in the third embodiment, the image processing unit 12 inverts the complex-shaped white region to a black region, thereby generating an image that more accurately reflects the martensite structure region.

[0042] In the first to third embodiments, the image processing unit 12 may calculate the phase fractions of the structure that constitutes the dual-phase steel by using a point counting method. The image output unit 13 may output the calculated phase fractions.

[0043] The method for calculating the phase fraction using the point counting method is as follows. First, lines are drawn in a grid pattern on the image. Then, the points on the lines where the phase changes are counted. In dual-phase steel, the phase change points may be counted by distinguishing between, for example, primary austenite and secondary austenite as different phases, rather than being limited to two phases. The phase fraction can be calculated by tallying the counts. Furthermore, the two-dimensional extent of the phase can be calculated by comparing the counts in the vertical and horizontal directions.

[0044] Other Embodiments One embodiment of the present invention has been described in detail above with reference to the drawings, but the specific configuration is not limited to that described above, and various design changes and the like are possible within the scope that does not deviate from the gist of the present invention.

[0045] A computer may implement part or all of the image generation device 10 in the above-described embodiment. In this case, a program for implementing the functions may be recorded on a computer-readable recording medium, and the program may be loaded into a computer system and executed. Note that the term "computer system" as used herein includes an operating system and peripheral hardware. Furthermore, the term "computer-readable recording medium" refers to portable media such as flexible disks, optical magnetic disks, ROMs, and CD-ROMs, as well as storage devices such as hard disks built into computer systems. Furthermore, the term "computer-readable recording medium" may also include media that dynamically store programs for a short period of time, such as communication lines used when transmitting programs via networks such as the Internet or telephone lines, or media that store programs for a fixed period of time, such as volatile memory within a computer system serving as a server or client. The program may be designed to implement part of the functions described above, or may be capable of implementing the functions in combination with a program already stored in the computer system. Furthermore, part or all of the image generation device 10 may be implemented using a programmable logic device such as an FPGA (Field Programmable Gate Array). [Explanation of symbols]

[0046] 10 Image generating device, 11 Image acquiring unit, 12 Image processing unit, 13 Image output unit, 20 Imaging device, 21 Optical microscope, 22 Camera, 30 Storage device

Claims

1. binarizing the grayscale image showing the structure of the dual-phase stainless steel to generate a first image; A method for generating a structural image of a dual-phase steel, which performs a smoothing process and a binarization process on the first image to generate a second image in which specific regions have been removed.

2. The second image is an image in which at least one phase of the dual-phase steel is extracted. The method for generating a structural image of a dual-phase steel according to claim 1.

3. The specific region is a linear region corresponding to a grain boundary. The method for generating a structural image of a dual-phase steel according to claim 1.

4. performing a smoothing binarization process on the first image multiple times; The method for generating a structural image of a dual-phase steel according to claim 1.

5. performing a noise removal process on the first image; The method for generating a structural image of a dual-phase steel according to claim 1.

6. When a region consisting of pixels taking one value in the first image or the second image is surrounded by pixels taking the other value, the values ​​taken by the pixels included in the region are inverted based on the complexity of the shape of the region. The method for generating a structural image of a dual-phase steel according to claim 1.

7. the binarization process is an adaptive binarization process; The method for generating a structural image of a dual-phase steel according to claim 1.

8. generating a third image by performing a logical OR operation on two pixels at the same position in each of the first image and the second image; The method for generating a structural image of a dual-phase steel according to claim 1.

9. The dual-phase steel is a ferrite-bainite steel, a ferrite-martensite steel, or a ferrite-austenite steel. A method for generating a structural image of a dual-phase steel according to any one of claims 1 to 8.

10. A method for generating a structural image of a dual-phase steel, comprising performing principal component analysis on the second image according to any one of claims 1 to 7 or the third image according to claim 8, calculating the direction of the components, and evaluating the properties of the dual-phase steel based on the direction of the components.

11. an image processing unit that performs a binarization process on a grayscale image showing the structure of the dual-phase steel to generate a first image, smooths the first image, and further binarizes the first image to generate a second image from which specific regions have been removed; An image generating device comprising:

12. A program for causing a computer to execute the tissue image generating method according to claim 1.

Images