JPS60143769A - Particle size and second phase fraction measuring apparatus - Google Patents

Abstract

PURPOSE:To enable the measurement of crystal particle size and the second phase fraction at a high accuracy with a better reproducibility by accentuating the gray tone of linear elements in an image obtained from a microscope in the measurement of ferrite particle size, austenite particle size and the second phase fraction. CONSTITUTION:After a nital etching process of metal, an optical image is taken on the surface of the sample with a camera 2 connected to a microscope 1 and is digitized withan A/D conversion section 3. Then, one picture is divided into numerous pixels with a linear element accentuation processing section 4 and the gray tone of each pixel is compared with the mean of neighbors thereof to accentuate linear elements. Then, the grain boundary portion and the second phase portion are extracted with a binary coding section 5. Then, small regions with linear images independent are eliminated as noise with a grain boundary information supplementary processing section 6 while lines yet to be closed are corrected by connection and finally, the results are displayed quantitatively with a particle size/second phase fraction measuring section 7. Thus, the removal of noise and others can be accomplished by accentuating the linear elements thereby enabling the measurement at a high accuracy in a short time with a better reproducibility making the metal texture free from effect of grinding flaws and rust.

Description

[Detailed description of the invention] Conventional, ferrite grain size, austenite grain size.

To measure the second phase fraction, etc., a comparison method has been widely used in which a standard sample with a known grain size number is compared with the target metal structure. However, in the above-mentioned comparative method, rough measurements can be made due to differences in shape and distribution between the sample to be measured and the standard sample, but accurate measurements are not possible.

In addition, accurate values can be obtained by taking a micrograph of the sample to be measured and measuring the area of individual grains and second phase components by manually tracing the grain boundaries or using a planimeter. This method was inefficient as it yielded few results for kneading, which required a lot of labor.

FIG. 1 is a diagram showing the configuration of a conventional automatic image analysis device capable of measuring particle size and second phase fraction.

In Figure 1, J3 is using a microscopic photograph of the sample or a microscopic image of the sample using a surface scanning image input device such as a television camera.
After the sample is loaded into an analyzer and subjected to binarization processing, the particle size and second phase fraction are automatically measured. However, when using the above-mentioned automatic analysis device to reveal the grain boundaries of a sample due to nital corrosion, it is difficult to
It was not possible to correctly distinguish polishing 1n, etc., and there was a problem in terms of measurement accuracy and reliability. The main reason for this is that the original image is binarized as is. That is,
This is because when inputting an image to an analysis device, a considerable amount of grain boundary information is already lost or noise is mixed in, so if the image is simply binarized and analyzed using IC, problems remain in the measured values.

The purpose of the present invention is to solve the above-mentioned problems and make it possible to measure crystal grain size and second phase fraction with high accuracy and reproducibility by performing image processing such as highlighting line elements and connecting two grain boundaries. It is an object of the present invention to provide a device for measuring particle size and second phase fraction.

The particle size and second phase fraction measuring device of the present invention includes a microscope for observing a metal structure, an imager connected to the microscope, and outputting analog image information by converting the metal structure image into an R image.
an analog/digital conversion unit that converts analog image information from the image pickup device into legal-level digital image information;
An image processing section that emphasizes grain boundary areas by applying line element extraction filter processing to the digital image information converted by the analog/digital conversion section and an image obtained from the image processing section are preset. A binarization processing unit that extracts the second phase portion and grain boundary portion from the metallographic image by comparing it with a threshold value, and a calculation that stores image information in which the extracted grain boundary portion is corrected to a perfect state. The processing unit is characterized by comprising a quantitative processing unit that measures the grain size, second phase fraction, etc. from the coordinates of the grain boundary position and second phase position in the image information obtained by the arithmetic processing unit. It is something to do.

FIG. 2 is a diagram showing an embodiment of the particle size and second phase fraction measuring device of the present invention. In FIG. 2, the structure of the measurement sample revealed with a nital corrosive solution or a picric acid corrosive solution is read out as optical image information by an imager 2 connected to a microscope 1, and sent to an A/D converter 3. Supplied. The optical image information supplied to the A/D converter 3 is 1
If the screen is made up of, for example, 512 x 512 pixels, each pixel is converted into digital image information having 64 or more gray tones, preferably 256 gray tones. Here, the legal level is the absolute value of the gray tone from black to white in the image, and 256 gray tone refers to the tone between black and white, where black is O1 and white is 255.
It is divided into 6 equal parts. The digital image information is stored in a modal image memory provided in the line element emphasis processing section 4 and consisting of a frame memory.

Next, the line element emphasizing processing section 4 performs line element emphasizing processing on the image stored in the legal image memory based on the principle shown below. First, for each pixel of the image stored in the modal image memory, create a line element emphasis filter shown in Figure 3 consisting of, for example, 5 x 5 pixels in the vicinity of that pixel, and set the center as shown below. Average A1- of gray tone level values of pixels located on straight lines in four directions passing through the pixels
84 and extract the total average value Aa of the pixels within the filter.

Note that aij in FIG. 3 represents the legal system level of each pixel.

Al-5 (a31+a32+a33+a84+a35
) A2-cho (a13+a284-a38+a48+a5
8) A8” 5 (a11+a22+a33+a44+
a55 ) A4-5 (a51+a42+a, +3+a
24"15) Next, extract the minimum value A min from the request l = A + ~ △5, compare this minimum value A min and the total average value Aa, and determine if the difference is greater than the preset threshold value. When it is large, the minimum value A min is given by 2, and when it is small, the total average value Aa is given as the new modal level of the target pixel.

If the above-described operation is performed on all pixels in the modal image memory, image information with line elements emphasized can be obtained. Note that this processing involves adding and subtracting digital values and comparing their magnitudes, and from the viewpoint of processing speed and the fact that similar processing is repeated, it is desirable to configure it with a dedicated electronic circuit.

The image information with line elements emphasized using the method described above is supplied to the binarization processing unit 5, where it is binarized using a preset threshold value and extracts only the grain boundary portion and the second phase portion. . ] The image information after the digitization process is stored in a binarized image memory consisting of a frame memory provided in the arithmetic processing section 6. Each pixel in the binarized image memory has a grain boundary portion and a second phase portion of 1. Other parts of the crystal are stored as O.

The binarized image information is subjected to grain boundary information supplementation processing in the arithmetic processing section 6. That is, for the image in the binarized image memory, 1) independent, non-linear small areas that are not connected to others are removed as signal noise; 2) For a linear region that is not closed, extend the line to a point where the evaluation function, which is inversely proportional to the square of the distance at points near its tip and proportional to the direction cosine in the direction of the line, is maximum. Extended J. By performing these two processes, all grain boundaries can be expressed as closed linear regions.

Hereinafter, the method for detecting connection points and connecting the connection points used in the above-described processing will be described in detail. First, in order to detect the end points of grain boundaries to be replenished, bonding points are defined for the two cases shown in FIG. 4(Δ) and (B). The white circles in the figure are the connection points. In the example shown in Figure 4 (A), the end point when tracing connected pixels is defined as a joining point. In the example shown in Figure 4 (B), connected pixels are connected in two directions or in one direction. In the case of branching, the protrusion when tracing connected pixels in each direction is defined as a joining point when O ABC in the figure is less than a certain angle.Next, the method for connecting joining points is as follows. Figure 5 (A
), as shown in (B), for each bond point Δ, an evaluation function F1 or F2 described below is calculated for a bond point B or a grain boundary point C existing in a certain region.

(1) Evaluation function between connection points A and B (Figure 5 (A)) Fl
=cO3θ1・COSθ2/L When the straight line segment Δ, B intersects with another grain boundary, it is set as Fl-0.

(2) Evaluation function of bonding point A and grain boundary point C (Fig. 5 (B))
4≦θ≦π/'4. a=1/2゜Here, find the bonding point B1 grain boundary point C where the evaluation functions Fl and F2 are maximum, and F+
If ≧F2, the maximum bonding point 13 is connected to the bonding point A, and if F+<12, the maximum grain boundary point C and the bonding point A are connected. By the method described above, an image can be expressed as a linear region in which the first region is closed.

Image 1 where grain boundary information has been supplemented using the method described above.
R: Information is supplied to the M constant processing unit 7, and the information is supplied to the constant M processing unit 7, and the grain boundary position n of
The particle size and second phase fraction are determined from the coordinates of the second phase position. At this time, line connection processing is not necessary to calculate the second phase fraction, and the judgment condition that the second phase portion extracted by binarization after line element emphasis is larger than a preset area is also used. The area of only the second phase portion is measured by covering the measurement area, and the ratio of the area to the total area is calculated.
The phase fraction is calculated.

The actual processing results will be explained below. FIG. 6 is a diagram showing a metallographic image of a ferrite-pearlite structure obtained by corroding a steel material with a nital corrosive solution. FIG. 7 is a diagram C showing the processing result of extracting grain boundaries from the metallographic image shown in FIG. 6 by only binarizing using the conventional method. In the example shown in FIG. 7, noise due to grain boundary discontinuities, scratches, rust, etc. can be seen. FIG. 8 is a J diagram showing the result of the metallographic image shown in FIG. 6 being subjected to line element enhancement, binarization processing and line connection processing using the apparatus of the present invention. It is clear that in the example shown in FIG. 8, the crystal grain size can be determined more accurately than in the example shown in FIG. 7 using the conventional apparatus.

In the embodiment described above, the image information was collected by capturing a microscope image with an imaging device, but it is also possible to use other image input devices, and the data processing device etc. can be adapted to the present invention. It is also possible to connect to a device.

As is clear from the detailed explanation above, according to the particle size and second phase fraction measuring device of the present invention, noise caused by polishing scratches, rust, etc. is removed by emphasizing line elements in image information. Furthermore, since grain boundary connection processing is performed on the image signal with emphasis on line elements, it is possible to measure crystal grain size more accurately than with conventional equipment. Measurements of the second phase fraction can be made.

[Brief explanation of the drawing]

Fig. 1 is a diagram showing the configuration of a conventional automatic double-image analyzer capable of measuring particle size and second phase fraction, and Fig. 2 is a diagram of the configuration of a conventional automatic double-image analyzer capable of measuring particle size and second phase fraction. Figure 3 is a diagram showing a line emphasis filter; Figures 4 (A) and (B) are diagrams for explaining connection points used in boundary information replenishment processing; Figures 5 (△) and (B) are diagrams for explaining the connection method of bonding points, Figure 6 is a diagram showing the metallic phase m image of the ferrite-pearlite structure, and Figure 7 is the same as Figure 6. Figure 8 shows the result of processing the metallographic image shown in Figure 6 by extracting grain boundaries only by binarizing it using the conventional method. FIG. 7 is a diagram showing the results of binarization processing and line connection processing after line element emphasis. 1...Microscope 2...Imaging device 3...A/D conversion unit 4...Image processing unit 5...Binarization processing unit 6...Essential processing unit 1...Quantitative processing Department. Patent applicant: Kawasaki Steel Corporation Applicant: Japan Regicolator Co., Ltd. No. 4 Figure 3 (A) (A) Figure (B) (B)

Claims (1)

[Claims] 1. A microscope for observing metal structures, an imaging device connected to this microscope that conveys images of metal structures and outputs analog image information, and converting the analog image information from this imaging device into legal-level digital image information. An analog/digital converter converts the data, an image processor performs line element extraction filter processing on the digital image information converted by the analog/digital converter to emphasize grain boundary areas, and an image processing unit performs line element extraction filter processing on the digital image information converted by the analog/digital converter. A binarization processing unit extracts the second phase part and grain boundary part from the metallographic image by comparing the image with a preset threshold value, and the extracted grain boundary part is corrected to a perfect state. Consists of an arithmetic processing section that stores image information, and a quantitative processing section that measures grain size, second phase fraction, etc. from the coordinates of grain boundary positions and second phase positions in the image information obtained by the arithmetic processing section. A measuring device for particle size and second phase fraction, characterized in that: