JPH03150447A - Method and apparatus for analyzing crystal structure - Google Patents

Abstract

PURPOSE:To accurately and efficiency analyze a crystal structure by using the optical microscopic image for which a phase difference method is used as an original image and recognizing the grain boundary by binarization. CONSTITUTION:The optical microscopic image for which the phase difference method is used is used as the original image and the grain boundary is regulated to be recognized by the binarization. A variable threshold method is used in the binarization to extract the grain boundary in addition to technology and further, the grain boundaries cut here and there are restored by using the remov al of isolated points, graphic expansion, graphic contraction, and formation of finer lines after the grain boundary is extracted by the binarization. The crystal structure analyzing apparatus consisting of the optical microscope which can produce the phase difference image and an image analyzing machine which makes the calculation to extract the grain boundary of the optical microscopic phase difference image of the crystal structure by the binarization is provided. The crystal structure is analyzed with the high accuracy and the high efficiency in this way.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a crystal structure analysis method and an apparatus therefor.

[Conventional technology]

In general, to analyze the crystal structure, there are two methods: (1) cutting out each crystal grain in an optical microscope photograph and measuring its weight to determine the area per grain size of each crystal grain;
(or photo) A method of drawing a straight line and calculating the average grain size from the number of grain boundaries that cross the straight line. (3) Calculating the average area of grains from the number of grains within a unit area.

Methods include determining the average grain size, and (4) manually tracing grain boundaries in an optical microscope photograph and analyzing the traced image using an image analyzer.

However, in the case of (1) and (41), the efficiency is extremely low, and in the case of (2L (3), only the average information of the crystal structure is used, and the metallurgically important information such as the shape of the two grains in the grain size distribution is used. There are problems such as the inability to measure the amount.

On the other hand, the crystal structure in materials such as steel is a factor that greatly influences mechanical and magnetic properties, and controlling the crystal structure is one of the important academic fields of materials science. In addition to this influence on product properties, crystal structure control plays an important role in the manufacturing technology of each material. For example, grain-oriented electromagnetic copper sheets are manufactured by utilizing the secondary recrystallization phenomenon to obtain crystal oriented grains with a high degree of integration. Particle size 1 Particle size distribution indicates whether secondary recrystallization is good or bad.
It is known that the orientation of secondary recrystallized grains affects the properties of products. Furthermore, it is well known that the crystal grain size affects the workability of the product, and that the crystal grain size during the manufacturing process affects the texture after cold rolling and recrystallization.

Based on this background, there has been a demand in recent years for the development of a crystal structure analysis method with excellent accuracy and efficiency.

[Problem 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 and efficiently analyze a crystal structure.

[Means to solve the problem]

The present invention provides a method for analyzing a crystal structure using image analysis, in which an optical microscope image using a phase contrast method is used as an original image;
This provides a method for recognizing grain boundaries by binarizing them. In addition to the above techniques, the present invention also provides a method using a variable threshold method in binarization for extracting grain boundaries. Furthermore, in addition to the above-mentioned technique, the present invention provides a technique of extracting grain boundaries by binarization and then repairing broken grain boundaries using expansion of one figure with removal of isolated points and contraction and thinning of one figure. . Further, the present invention provides a crystal structure analysis apparatus comprising an optical microscope capable of expressing a phase contrast image and an image analyzer that performs calculations to extract grain boundaries in the optical microscope phase contrast image of a crystal structure by binarization.

When analyzing the crystal structure in materials such as steel, the present invention has been developed by (1) cutting out each crystal grain in an optical microscope photograph and measuring the weight, which has been generally used in the past to analyze the crystal structure of materials such as steel; (2) Draw a straight line on an optical microscope image (or photograph) and calculate the average grain size from the number of grain boundaries that cross the straight line. (3) Calculate the average grain size from the number of grains within a unit area. The method of determining the average area average grain size of the grains, and (4) the method of manually tracing the grain boundaries in an optical microscope photograph and analyzing the traced image using an image analyzer, are better in terms of accuracy, efficiency, amount of information, etc. Where there is a problem, the crystal structure can be determined with high precision by image analysis of the phase contrast image of an optical microscope.
The purpose is to analyze with high efficiency.

As a result of extensive studies on various methods for analyzing crystal structures, the present inventors used a phase contrast image of an optical microscope as an original image, which reveals the unevenness of a sample as a difference in brightness and darkness. We obtained new knowledge that extracting grain boundaries by binarization is extremely effective. In addition, since the brightness of the matrix of a phase contrast image usually differs depending on the location, new findings were obtained that the variable threshold method is more accurate than the fixed threshold method during binarization. Furthermore, after extracting the grain boundaries by converting them into negative values, the grain boundaries are usually cut off and also contain noise such as polishing scratches and inclusions.
This binary image is expanded,
We have obtained new knowledge that it is extremely effective to convert images containing only grain boundaries using line thinning.

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

In the present invention, in the method of analyzing crystal structures using image analysis, it is specified that an optical microscope image using a phase contrast method is used as the original image and grain boundaries are recognized by binarization.
The phase contrast method is a method that reveals the unevenness of the sample as a difference in brightness and darkness, and it is possible to create a difference in brightness and darkness from the matrix even at small angle grain boundaries that are difficult to appear in normal corrosion. This is because it is advantageous for extracting the field.

The method of inputting the optical microscope image into the image analyzer is not particularly limited. Any method may be used, such as inputting an optical microscope photograph using a television camera or directly inputting an electrical signal from an optical microscope to an image analyzer. In addition, there is no particular limitation regarding the image processing of the grayscale image for the input image, but
It is preferable to perform shading correction, smoothing, standardization of gray level, etc. from the viewpoint of improving accuracy.

In addition to the above invention, the variable threshold method is specified to be used in the binarization for extracting grain boundaries, because the brightness level of the matrix of the phase contrast image of the optical microscope changes continuously from place to place. This is because binarization using the variable threshold method is more effective in extracting grain boundaries than binarizing using the fixed threshold method. The variable threshold method is a method that usually changes the binarization threshold depending on the location of the image, but it calculates the density level of pack ground by smoothing etc., and extracts areas that are higher than hack ground by a predetermined density level. Method 1: Any method may be used, such as dividing the image into several regions and binarizing each region using a discriminant analysis method.

Furthermore, in addition to the above invention, it is specified that after grain boundaries are extracted by binarization, broken grain boundaries are repaired by expansion of isolated point removal 2 figures, 5 contraction and thinning of figures, and broken grain boundaries are repaired. After extraction by binarization, the grain boundaries are usually cut off and the image contains noise such as polishing scratches and inclusions, so this binarized image is extracted without noise.
In order to convert to an image in which the grain boundaries are completely connected, it is necessary to restore the width of the grain boundaries to the original size by connecting the isolated grain boundaries and subsequently shrinking the shape. This is because line thinning that minimizes the boundary width is effective.

Furthermore, when restoring a binarized image to an image containing only grain boundaries, the removal of isolated points, the expansion of the figure, and the contraction of the figure are performed as a series of processes. It is further preferable to repeat the above series of processes by changing the shrinkage width) from a small value to a large value in order to improve accuracy.

There are no particular limitations on the method of thinning.

Hilditch's method, Tsuruoka's method, Itaben's method.

Any method such as Dentsch's method or Enso's method may be used.

Further, after thinning, it is more preferable to perform isolated point removal, whisker removal, arrow connection, etc. in a timely manner from the viewpoint of improving accuracy.

In addition to the above invention, a crystal structure analysis device comprising an optical microscope capable of displaying a phase contrast image and an image analyzer that performs calculations to extract grain boundaries in the optical microscope phase contrast image of the crystal structure by binarization. This is specified because the above-mentioned hardware is required in order to perform crystal structure analysis using the algorithm of the present invention. Further, specifications of the optical microscope and image analyzer other than those stipulated above are not particularly limited.

Further, after grain boundaries are recognized according to the algorithm of the present invention, broken portions of the grain boundaries may be manually connected, if necessary, if extremely high precision is required. Thereafter, the crystal grain size (circular diameter), circularity 2 perimeter length, aspect ratio, etc. are usually measured.

〔Example〕

Example I C: 0.0010%, Si: 3.25%, Mn: 0.
14%, S: 0.006%, acid soluble A7: 0.02
6% N: Contains 0.0081%, the remainder being Fe and unavoidable impurities. A cross section of a silicon steel annealed plate with a thickness of 0.285 mm was polished for optical microscopic observation, then corroded with nital, and the remaining part was Fe and unavoidable impurities. A phase contrast image was developed, the image was input from a television camera into an image analyzer, and a grain boundary image was created using the image analysis flow shown in FIG. The original image (phase contrast image) ((a)) and grain boundary image ((b)) are shown in FIG.

Example 2 C: 0. OO15%, Si:3.28%, Mn:0
．． 15%, s:o, oo7%, acid soluble IV: 0.0
A cross-section of a silicon steel plate with a thickness of 0.285 mm containing 0.0083% of 27% N and the remainder Fe and unavoidable impurities was polished for observation with an optical microscope, then corroded with nital, and a phase contrast image was obtained using an optical microscope. After inputting the image from the TV camera into an image analyzer and creating a grain boundary image using the image analysis flow shown in Figure 1, the diameter of each crystal grain (
The average value and standard deviation were calculated. For comparison, the grain boundaries of the phase contrast image input into the image analyzer were manually traced, and the average value and standard deviation of the diameter of each crystal grain were similarly measured using the traced image. Also,
Grain boundaries were extracted from the above phase contrast image using a manual threshold specification binarization method, and the average value 0 and standard deviation of the diameter of each crystal grain were similarly measured.

show.

The above measurement results are shown in Table 1. Example 3 C: 0.0011%, Si: 3.21%, Mn: 0.1
4%, s: o, oo7%, acid soluble A1: o, 02
8%.

N: A cross section of a silicon steel plate with a thickness of 0.285mm containing 78% O, OO and the balance Fe and unavoidable impurities was polished for optical microscopic observation and then corroded with nital.
After developing a phase contrast image using an optical microscope and inputting the image from a TV camera into an image analyzer and creating a grain boundary image using the image analysis flow shown in Figure 1, the diameter of each crystal grain (equivalent to a circle) is calculated. diameter) and calculated the particle size distribution. For comparison, the grain boundaries of the phase contrast image input into the image analyzer were manually traced, the diameter of each crystal grain was similarly measured in the traced image, and the grain size distribution was calculated. The results obtained using the method of the present invention and the tracing method are shown in FIG.

[Effects of the Invention] As described above, according to the present invention, the crystal structure, which has a great influence on the mechanical properties and magnetic properties of materials such as steel, and is an important metallurgical factor in the manufacturing technology of each material. It has great academic and industrial effects because highly accurate analysis of crystal grain size, distribution state, etc. can be carried out efficiently.

[Brief explanation of the drawing]

Figure 1 is an image analysis flow for obtaining a grain boundary image using an optical microscope phase contrast image of a crystal structure as an original image, and Figure 2 shows (a) an optical microscope phase contrast image of a crystal structure, and ( b) An optical micrograph of a crystal structure showing an example of a grain boundary image obtained using the method of the present invention, and FIG. 3 is an example of crystal grain size distribution obtained by the method of the present invention and the tracing method. . 2 Figure 2 (b) 0IIm Procedural amendment (spontaneous) May 28, 1990 1゜2゜ Incident representation Y 1999 bond 5'"Previous section 289:36" Name of the invention Analysis of knotted silk fabric Method and device 3, relationship with the case of the person making the amendment Applicant: 6-3 Otemachi 2-chome, 11'1-ku, Chiyo, Tokyo (665)
Nippon Steel Corporation Representative Zen Yamamoto 4゜Embedded 2-4-1-5 Marunouchi, Chiyoda-ku, Tokyo 100 Date of amendment order Month Day 6, 1998 Invention of the amended lunar description Detailed explanation column 9 Brief explanation of the drawings. (2) Correct page 12, line 18, 1 - Optical microscope photograph - J to "photograph". (3) Correct each of Figures 1 and 2 as shown in the attached sheet. 27゛Σ゛(b)

Claims (1)

[Claims] (1) A method for analyzing a crystal structure using image analysis, characterized in that an optical microscope image using a phase contrast method is used as an original image, and grain boundaries are recognized by binarization. Method of analyzing crystal structure. (2) The crystal structure analysis method according to claim 1, wherein a variable threshold method is used in the binarization for extracting grain boundaries. (3) After the grain boundaries are extracted by binarization, broken grain boundaries are repaired by removing isolated points, expanding the figure, shrinking the figure, and thinning the line. A method for analyzing the crystal structure of. (4) A crystal structure analysis device characterized by comprising an optical microscope capable of displaying a phase contrast image, and an image analyzer that performs calculations to extract grain boundaries in the optical microscope phase contrast image of the crystal structure by binarization. .