JPH0340351A - Analyzing method and device for educed object - Google Patents

Abstract

PURPOSE:To make analysis of educed substance with high accuracy and high efficiency by subjecting an electron microscopic image to a processing, and obtaining the logical sum of the two-value image in the region bounded by vivid boundaries and an image which has been turned into two values by the variable threshold value method. CONSTITUTION:A specimen for observation of educed substance of a surface dimensioned half the thickness of a Si steel annealed plate is prepared by means of extraction replica method, followed by photographing using an electron microscope, and the image is fed to an analyzer by the use of a TV camera 42. The analyzer converts 43 analog signals of electron microscopic photo 49 into digital signals and feeds to a central processing unit 45 via an image memory 44. The entered image of educed substance is subjected to image processing, and identification of the substance is made, followed by measurement of its size, dispersed condition, etc., and the result is output to a printer 48.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a method for analyzing precipitates and an apparatus therefor.

(Prior Art) Conventionally, analysis of precipitates using an electron microscope has generally been carried out by visual measurement.

However, this visual inspection has the following problems: 1) The measurement area is small and the entire sample cannot be grasped properly; 2) The recognition accuracy of precipitates is low.
Problems such as those have been pointed out (for example, Kama 1) edited by Jin:
“Recent Steel Condition Analysis” (Agne Publishing, 1979), 30
1 page).

In addition, methods for preparing observation samples for precipitates include l) thin film method, 2)
There is an extraction replica method, but the extraction replica method is suitable for observing a wide field of view. However, the contrast of precipitates in electron microscope images obtained by the extraction replica method with the replica film is small, and even when recognizing precipitates using image analysis technology that has advanced in recent years, it is difficult to recognize the precipitates using manual threshold specification. When recognizing precipitates using chemistries, the number and size of precipitates vary greatly due to slight differences in the threshold value, resulting in poor reliability.

On the other hand, precipitates in metal materials such as steel are factors that greatly affect mechanical and magnetic properties, and controlling precipitates is an important academic field in materials science. In addition to the influence of precipitates on product properties, control of precipitates plays an important role in the manufacturing technology of various materials. For example, grain-oriented electrical steel sheets are
jVN, Mn of several hundred λ to suppress grain growth of -order recrystallized grains
It is manufactured by generating secondary recrystallized grains using precipitates such as S and controlling the degree of accumulation of the crystal orientation.

Furthermore, it is important to control the amount and size of precipitates as a control factor for the recrystallized grain size, which affects the workability of materials. For example, Fe for aging control, and Mn as C precipitation nuclei.
The importance of precipitates such as S has been pointed out. On the other hand, the importance of high strain regions near precipitates as nucleation sites for recrystallization has also been pointed out, and the vicinity of precipitates is considered to be the preferential nucleation site for transformation. As described above, the importance of precipitates in materials science is too numerous to list, and the size and distribution state of precipitates are especially important.

However, although there are various analytical methods and identification methods such as electron beam diffraction for analyzing precipitates,
Conventionally, there have been insufficient means to quantitatively analyze the distribution state.

In addition, in recent years, the development of high-purity materials has pushed forward, and chemical analysis has reached its limit in measuring the amount of precipitates, and it is now necessary to consider measuring the amount of precipitates from electron microscope images. ing.

(Problems to be Solved by the Invention) The present invention provides a method and an apparatus for solving the problem that it is difficult to accurately analyze the size, distribution state, etc. of precipitates.

(Means for Solving the Problems) The present invention provides a method for analyzing precipitates using an electron microscope image of a sample created by an extraction replica method. This invention provides a method for recognizing precipitates by using the logical sum of a binary image of a surrounded area and an image binarized by a variable threshold method. In addition to the above method, the present invention also provides a method for recognizing sharp boundaries using an image obtained by binarizing and extracting a region of a grayscale image with a high two-dimensional differential value using a variable threshold method. Furthermore, we used an extraction replica creation device, an electron microscope, and created a binary image of a region surrounded by clear boundaries from the electron microscope image of the precipitate, and then created a binary image that was binarized using a variable threshold method. The object of the present invention is to provide a precipitate analysis device comprising an image analyzer that creates a binary image and calculates the logical sum of the two binary images.

The present invention will be explained in detail below.

The present invention uses image analysis to analyze the size and distribution of precipitates in steel and rattan materials, instead of the conventional visual measurement of electron microscope images. The aim is to do so with efficiency.

As a result of extensive studies on various methods for analyzing precipitates, the inventors have developed a variable threshold method for analyzing precipitates using electron microscope images of samples created by the extraction replica method. In order to achieve the objective, it is essential to solve the problem that when precipitates are recognized using a binary image binarized using the value method, there is a region near the center of the precipitate that is not extracted by binarization. I came to the conclusion that it is. The variable threshold method is a method in which, based on the average value of the gray level of the area near each pixel, a value with a gray level higher by a specified number than that value is used as the threshold for binarization. Although it is an excellent method for recognizing precipitates when analyzing precipitates using the extraction replica method, where the density of the replica film (background) usually varies, the density is almost equal near the center of the precipitate. However, when binarization is performed using the variable threshold method, there is a problem in that there are regions that cannot be extracted by binarization.

Therefore, the present inventors investigated various ways to solve this problem of missing extraction, and found that the boundary between the replica film and the precipitate is an area with large density changes (high two-dimensional differential value of the density image). The research focused on the following. Then, a region surrounded by a clear boundary (region with a large change in fi) is extracted by binarization, and the logical OR is performed with the binary image binarized by the variable threshold method. Through this method, we solved the above-mentioned problem of missing extraction and obtained new knowledge that precipitates can be recognized with high accuracy.

Next, the reasons for limiting the constituent elements in the present invention will be described.

In the present invention, in a method for analyzing precipitates using an electron microscope image of a sample created by the extraction replica method, a binary image of a region surrounded by a clear boundary is obtained from an electron microscope image of a precipitate. It is specified that precipitates are recognized by the logical sum of images binarized using the variable threshold method. In order to eliminate missing extraction in the vicinity, a region surrounded by a clear boundary (boundary between the precipitate and the replica film) is extracted by binarization, and the extracted binary image is combined with the binary image using the variable threshold method. This is because it is effective to perform a logical sum with the binary image obtained by digitization.

The method of extracting a region surrounded by clear boundaries by binarization is not particularly limited.

Differentiate the electron microscope image using an image analyzer (one-way differential,
5obel method, etc.), extract the region with high differential value by binarization (variable threshold method, manual threshold specification method, mode method, differential histogram method, discriminant analysis method, etc.), and extract the extracted region ( A method of extracting a region surrounded by a clear boundary by filling in the holes (with a clear boundary), or a method of binarizing an electron microscope image using an image analyzer (manual threshold specification method) , mode method, differential histogram method,
Any method may be used, such as a method of extracting a region surrounded by sharp boundaries (discriminant analysis method, etc.).

In the second invention, a two-dimensional differential value (D value) of a grayscale image
The reason why we specified that clear boundaries should be recognized using images binarized and extracted from regions with high values using the variable threshold method is because the density level of the replica film usually varies, so it is necessary to recognize the boundary between the precipitate and the replica film. It is effective to use the D value as an index and judge that the area with a high D value is the boundary (clear boundary) between the precipitate and the replica film. This is because it is effective to binarize and extract areas where the D value is higher than the average value in the vicinity of each pixel, where the value is locally increased, using the variable threshold method to recognize clear boundaries. .

There is no particular limitation on the method of performing two-dimensional differentiation of the grayscale image. 5obel method, Prewitt method, Robe method
Any method such as the rts method may be used.

In the third invention, an extraction replica creation device, an electron microscope, and an electron microscope image of the precipitate are used to create a binary image of a region surrounded by clear boundaries, and a variable threshold method is used to generate a binary image. The precipitate analysis device is defined as a precipitate analysis device consisting of an image analyzer that creates a digitalized binary image and calculates the logical OR of these two binary images.It is defined as a precipitate analysis device that has an extraction replica creation device, an electron microscope, and the above functions. This is because, without an image analyzer, precipitate analysis using the algorithm of the present invention is impossible. The method of inputting the electron microscope image to the image analyzer is not particularly limited. There are reliable methods such as transmitting images from an electron microscope to an image analyzer using electrical signals, or taking an electron microscope image and inputting the image in negative or positive film using a television camera. Also, there are no particular limitations on the specifications of the image analyzer other than those stipulated above.

Li! precipitate according to the algorithm of the present invention! In order to improve accuracy, it is preferable to use known methods for image smoothing, sharpening, and contrast enhancement as preprocessing. After that, thinning and connecting the caustic lines are performed sequentially, which has the effect of eliminating omissions in the extraction of sharp boundaries, which is more preferable in terms of improving accuracy.

Further, after the precipitates are recognized according to the algorithm of the present invention, the size (circular equivalent diameter, etc.) and distribution state (distance between nearest neighboring grain centroids, etc.) of the precipitates are usually quantified. The image analyzer can usually measure the equivalent circle diameter, circularity, circumference, etc. of the precipitate, and the specifications of the measurement parameters are not particularly limited.

Hereinafter, the present invention will be further explained based on the drawings.

FIG. 1 shows an example of an apparatus that performs etching, which is a preprocess for creating an extracted replica.

This consists of a potentiostatic device 3 having a saturated Amagi electrode 1 and a coulomb meter 2, and an electrolytic cell 4.

The sample 5, which has been previously finished into a mirror-like state by emery polishing, puff polishing, diamond polishing, etc., is electrolyzed at a predetermined potential between it and a platinum electrode 7 in a non-aqueous electrolyte 6. The sample that has undergone electrolysis is transferred into acetone 8 without drying.

FIG. 2 shows an example of an extraction replica creation device and creation procedure.

The etched sample was cleaned with clean methanol.
The surface of the sample is cleaned by spraying the sample with the liquid using the dropper 12. Furthermore, change the methanol 13 and repeat the washing. Next, an acetylcellulose film 15 soaked in methyl acetate 14 is attached to the etched surface of the sample so as to tightly adhere it.

After the acetylcellulose film is sufficiently dried, it is peeled off from the sample surface.

The peeled film is fixed on a slide glass 16 with the peeled side facing up. The slide glass 18 to which the film is fixed is set in the vacuum evaporation device 17 so as to be perpendicular to the evaporation source 19, and carbon 20 is evaporated thereon. The carbon film 21 has about 160 to 200 people. The vapor-deposited film is made of paraffin 2 dissolved on a slide glass 22.
The film 24 is attached so that the vapor-deposited side of the film 24 is in close contact with the material on which the film 3 is placed. A slide glass covered with a film is immersed in a container 26 containing methyl acetate 25. The paraffin and acetyl cellulose film membrane 27 are separated from each other by heating in a constant temperature bath for 60° C. Cl2O minutes.

After washing the carbon membrane with clean methyl acetate, it is transferred to a solution of 28 methanol. A carbon film in methanol is scooped onto a sheet mesh 29 for electron microscopy, dried and fixed on a filter paper 30.

FIG. 3 shows an example of an electron microscope for observing and photographing precipitates.

The electron microscope is a normal transmission electron microscope.

The mirror body 31 is maintained in a high vacuum, and an electron gun 32 is attached to the top of the mirror body 31. The electron beam emitted from the electron gun passes through an irradiation system lens 33, which is composed of an excitation lens, and a magnification system lens 34, and becomes a fluorescent light. It is projected onto the light plate 35. At this time, the replica sample 36 is set in the objective lens.

The image is recorded using film 3 in the camera chamber 37 below the fluorescent plate.
This is done by directly exposing 8 to an electron beam.

Figure 4 shows an example of an image analyzer for precipitate analysis.

The image analyzer converts an analog signal of an electron micrograph 49 (negative film or positive film) for observing precipitates inputted from a television camera 42 into a digital signal.
A-D converter 4 via D converter 43 and image memory 44
The central processing unit 45 is connected to the central processing unit 3. A main memory 41 , an image processor 47 , a CRT display 46 , and a printer 48 are connected to the central processing unit 45 . The image memory 44 functions as a buffer when sending the signal from the A-D converter 43 to the main memory 41. The image processor 47 performs image processing such as noise removal and various filtering processes.

For example, image processing is performed on the input precipitate image on a positive film according to the flow shown in FIG. The dispersion pattern B (distance M between nearest grain centroids) and the like are measured, and the results are output to the printer 48.

(Example) Example 1 By weight, C: 0.0015%, Si: 3.25%, Mn
: 0. 14%, S: 0. 0 0 7%, acid soluble M: 0, 0 2 6%, N: 0. O O
0.0% consisting of B 2%, balance Pa and unavoidable impurities.
A sample for precipitate observation on a half-thickness side of a 285 mm thick silicon steel annealed plate was prepared using the extraction replica method, a photograph was taken using an electron microscope, and the image was sent to an image analyzer using a television camera. A binary image of the precipitates was created manually using the image analysis flow shown in Figure 5. Figure 6 (a) is a metallographic photograph taken with an electron microscope showing the presence of precipitates in the sample.
(original image), and FIG. 6(b) is a metal structure photograph showing the binary image.

Example 2 By weight, C: 0. OO12%, Si:3.24%,
Mn: 0.075%, S: 0.024%, acid soluble N: 0.027%, N: 0.0081%, Sn
: 0. l 0%, Cu: 0.06%, balance Fe
A sample for observing the precipitates on the 115th plane of the thickness of a silicon steel annealed plate with a thickness of 0.220 mm, consisting of unavoidable impurities, was prepared using the extraction replica method, and a photograph was taken using an electron microscope.
The negative film is used as the original image and manually inputted into the image analysis machine.
The size (equivalent circle diameter) and number of precipitates were measured using the image analysis flow shown in the figure. In addition, to investigate the recognition accuracy of precipitates, the number of precipitates was visually measured. The measurement results are shown in the first
Shown in the table.

Table Example 3 By weight, C: 0.056%, Si: 3.24%, Mn
: 0.15%, S: O, OO6%, acid soluble A7:
0.025%, N: 0.0079%, balance Fe and unavoidable impurities. A hot rolled silicon steel plate with a thickness of 2,311 IIm was heated to 1200°C and then rapidly cooled to 1150°C.
It was treated under two conditions: held for 30 seconds, cooled down to 900°C for 2 minutes, and then rapidly cooled. Then, after cold rolling to a thickness of 0.285 mm, N, 15%, Hz: 75%, dew point: After annealing at 830"C for 70 seconds and then at 900"C for 20 seconds in a 59°C atmosphere,
Samples for precipitate observation on the 1/2 side of the plate thickness of these two types of annealed plates were prepared using the extraction replica method, 25 photographs of each sample were taken using an electron microscope, and the negative film was used as the original image. The data were input into an image analyzer and the size distribution (equivalent circle diameter) of the precipitates was measured using the image analysis flow shown in FIG.

The measurement results are shown in FIG. Note that here, the thickness of the observation region extracted on the replica film was approximated by the average size of the precipitates, and the number density of the precipitates was measured.

(Effects of the Invention) As described above, according to the present invention, it has a great influence on the mechanical properties and magnetic properties of materials such as steel, and is an important material science factor among the manufacturing technologies of various materials. Since it becomes possible to analyze the size, distribution state, etc. of a certain precipitate with high precision, it has great academic and industrial effects.

[Brief explanation of drawings]

Figure 1 is a diagram showing an example of a device that performs etching, which is a pretreatment for creating an extraction replica, Figure 2 is a diagram showing an example of the equipment and procedure for creating an extraction replica, and Figure 3 is a diagram showing the observation and photographing of precipitates. Figure 4 shows an example of an image analyzer for precipitate analysis. Figure 5 shows a binary image of the precipitate using an electron micrograph of the precipitate as the original image. Figure 6(a) is a metallographic photograph (original image) taken by an electron microscope showing the precipitate presence state of the sample, and Figure 6(c) is its binary image. Metal structure photograph, Figure 7 is a diagram showing the flow of image analysis in which a binary image of the precipitate is obtained using the negative film of the electron micrograph of the precipitate as the original image, and then the size and number of the precipitate are measured. The figure is a diagram showing an example of the classification results of precipitates measured using the flow shown in FIG. 7. 26:B Fig. 4 Master D's engraving (a) (b) 111← P To 2) tnt Fig. 8 Cultivated product circle equivalent sutra C, tL7 procedure correction group (method) Hei! 1 November 21, 2015, ``Affiliate Commissioner Yoshida Bunki 1, Indication of the case 1999 Patent Application No. 173181 2, Name of the invention Method and apparatus for analyzing precipitates 3, Person making the amendment Related Patent Yama Junjin 6° Subject of amendment (1) Specification page 15, lines 1 to 4 “Figure 6 (a) is...
......This is a photograph of the metal structure. ' shall be corrected as follows. ``Figure 6 (,) is a diagram schematically showing an electron microscope photograph (original image) showing the state of the precipitate in the sample, and Figure 6 (b) is a diagram schematically showing its binary image. (2) Page 18, lines 2 to 5, ``Figure 6 (a) is...''
......Metal structure photograph,'' is corrected as follows. ``Figure 6 (a) is a diagram schematically showing an electron microscope photograph (original image) showing the state of existence of precipitates, and Figure 6 (b) is a diagram schematically showing its binary image.'' (3 ) Correct Figure 6 as shown in the attached sheet. Procedural amendment (voluntary) January 21, 2008

Claims (3)

[Claims] (1) In a method of analyzing precipitates using an electron microscope image of a sample created by the extraction replica method, a binary image of a region surrounded by clear boundaries and a variable A precipitate analysis method characterized by recognizing precipitates by logical sum of images binarized using a threshold method. (2) The method according to claim 1, wherein a clear boundary is recognized by an image obtained by binarizing and extracting a region with a high two-dimensional differential value of a grayscale image using a variable threshold method. (3) Extraction replica creation device, electron microscope, creates a binary image of a region surrounded by clear boundaries from the electron microscope image of the precipitate, and binarizes the binary image using the variable threshold method. 1. A precipitate analysis device comprising an image analyzer that creates a binary image and calculates a logical sum of the two binary images.