JPH02134541A - Inspection of metal structure - Google Patents

Abstract

PURPOSE:To improve accuracy by providing an outer lighting source which projects illuminating light on a specimen in the oblique direction, and taking only the overlapped part out of three kinds of images whose lighting conditions are different. CONSTITUTION:Inner illuminating light L1 is projected on the coarse surface part (Al eutectic layer) and a smooth surface (Si crystal part) of a specimen surface 2 through an objective lens 5 of a microscope 3. External illuminating light L2 is projected in the oblique direction at a specified angle with the specimen surface 2. The images of the surface structure of the specimen which are captured with the objective lens 5 are picked up with a camera 8 for the following cases: when both illuminating lights L1 and L2 are projected; when only the illuminating light L1 is projected; and when only the illuminating light L2 is projected. The images are binary-coded in bright and dark values. Regions P1 - P3 of the specified components of the surface structure are taken out. Then, only a part P4 where the regions P1 - P3 are overlapped is taken out as a true region P7 of the specified component.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of Industrial Application The present invention relates to a method for inspecting a metal structure using a microscope, and more specifically, for example, a method suitable for detecting the area ratio of a specific component contained in an alloy structure. Regarding inspection methods.

Conventional technology For example, as a method for detecting the precipitation area ratio of Si (silicon) crystal parts in the surface structure of an aluminum alloy, the ninth
There is a method as shown in the figure.

This irradiates the sample 2 set on the surface plate 1 with the illumination light from the internal illumination light source 7 reflected by the half mirror 6 through the lens barrel 4 and objective lens 5 of the epi-illuminated microscope 3, while the objective lens 5 The surface structure of sample 2 captured by
D, etc.) type camera 8 takes an image. Then, the image captured by the camera 8 is binarized at a predetermined level by the image processing device 9, and the area of the bright part corresponding to the aforementioned Si crystal part is calculated and determined. A precipitation area ratio of is obtained.

That is, as shown in FIG. 3, the surface of sample 2 was once smoothed and then etched to form a Si
The exposed surface of the crystal abnormality O remains a mirror-like smooth surface, whereas the AQ common layer 11 has been roughened by etching and has a rough surface.

Therefore, when internal illumination light is irradiated onto the surface of the sample 2, the AQ common layer 11 reflects it diffusely due to its rough surface, and only a small part of the reflected light enters the objective lens 5 side.
Since the crystal abnormal part O has a smooth surface, the irradiated light is specularly reflected, and most of the reflected light is incident on the objective lens 5 side. That is, in the above-mentioned binarized image, the portion where the specularly reflected light is strong is recognized as the Si crystal portion 10 and extracted.

Problems to be Solved by the Invention However, in the conventional method, the Si crystal part 10, which is a smooth surface, is extracted from the difference in the brightness of reflected light, so the part corroded by etching is not necessarily a rough surface. If the surface becomes close to a smooth surface by chance, it will not be possible to clearly distinguish between the Si crystal portion 10 and the single layer 11. Therefore, it is extremely difficult to set an optimal threshold for extracting the Si crystal portion 10, and detection errors are likely to occur. Moreover, the conventional method is susceptible to the influence of external light, which is one of the reasons for further increasing the above-mentioned detection error.

The present invention has been made in view of the above points, and its purpose is to extract only mutually overlapping parts from three types of images with different illumination conditions, for example, as a smooth surface of tissue. The present invention provides an inspection method with improved accuracy.

Means for Solving the Problems The present invention is a method for extracting and inspecting specific components of a surface structure based on the difference in light reflection characteristics between a rough surface and a smooth surface of a sample. When internal illumination light is irradiated through the objective lens and external illumination light is irradiated from an oblique direction forming a predetermined angle with the sample surface, cases in which both internal illumination light and external illumination light are irradiated and cases in which only internal illumination light is irradiated. , and when only external illumination light is irradiated, the surface tissue of the sample captured by the objective lens is imaged by an imaging device, and the image is binarized to extract a specific component area of the surface structure. The present invention is characterized in that only the portions where the regions extracted from the three types of binarized images having different illumination conditions mutually overlap are extracted as the true region of the specific component.

According to this method, if the illumination conditions differ, even if the same object is imaged, the region of the specific component extracted from each image will differ slightly due to variations in the amount of reflected light. Therefore, for example, for a region corresponding to a smooth surface of a metal structure, 1
By extracting only mutually overlapping portions on the three types of binarized images with different illumination conditions as true smooth surface regions, the extraction error is reduced.

Embodiment FIGS. 1 to 7 are diagrams showing an embodiment of the present invention,
As shown in FIG. 2, the specific configuration of the inspection device is such that the sample 2 is illuminated from an oblique direction.

It differs from the one shown in FIG. 9 in that it includes an external illumination light source 12 that illuminates the light.

First, as shown in FIGS. 2 and 3, only the internal illumination light source 7 is turned on, and the internal illumination light L1 is irradiated onto the surface of the sample 20 through the objective lens 5. At this time, the AQ eutectic layer 11
As described above, since the surface is roughened by etching, the internal illumination light Ll is diffusely reflected, and a small portion of the reflected light enters the objective lens 5 side. On the other hand, since the surface of the Si crystal part 10 is a smooth surface close to a mirror surface, internal illumination light is specularly reflected and most of the reflected light enters the objective lens 5 side as brightness.

The diffusely reflected light from the Aff common layer 11 and the specularly reflected light from the Si crystal portion 10 returning to the objective lens 5 side are captured by the camera 8 as the brightness of each substance, and are captured by the camera 8 as the brightness of each substance. Stored in image memory. The brightness of the image is determined by sampling one screen into, for example, 256×240 pixels using an A/D converter, and quantizing the brightness into 64 density levels.

Next, the lighting states of the illumination light sources are switched to turn on only the external illumination light source 12 as shown in FIGS.
Hold it and irradiate from an oblique direction. At this time, since the Al1 common layer 11 has a rough surface, external illumination light is diffusely reflected, and a small portion of the reflected light returns to the objective lens 5 side as brightness.

On the other hand, since the Si crystal part 10 has a smooth surface, the external illumination light is specularly reflected at an angle equal to the incident angle θ, so that the specularly reflected light hardly returns to the objective lens 5 side. The reflected light returning to the objective lens 5 side is the third
Similarly to the figure, the image is captured by the camera 8 and stored in the image memory of the image processing device 9.

Next, in order to change the illumination conditions again, both the internal illumination light source 7 and the external illumination light source 12 are turned on to illuminate the surface of the sample 2 as shown in FIG. 1(A). At this time, the Al common layer 11
The reflection state at the Si crystal part 10 is a combination of the states shown in FIG. 3 and FIG. The reflected light from the i-eutectic layer 11 and the Si crystal portion 10 returning to the objective lens 5 side is captured by the camera 8 and stored in the image memory of the image processing device 9 in the same manner as described above.

The three types of images stored in the image processing device 9 under different illumination conditions as described above are shown in FIGS. 5 to 7, for example.

That is, FIG. 5 shows an image when only internal illumination light 7 is irradiated, FIG. 6 shows an image when only external illumination light 12 is irradiated, and FIG. 7 shows an image when only internal illumination light 12 is irradiated. An image obtained when both light 7 and external illumination light 12 are irradiated is shown.

As is clear from comparing these three types of images, the fifth
In FIG. 5 and FIG. 7, the Si crystal portion 10 appears as a white portion in FIG. 5 and FIG. On the other hand, in FIG. 6, since almost no specularly reflected light from the Si crystal part 10 enters the objective lens 5 side, the Si crystal part 1
0 appears as a black area.

Therefore, if the white part recognized as the 81 crystal part 10 in FIGS. 5 and 7 and the black part recognized as the Si crystal part 10 in FIG. 6 completely match, there will be no problem. As is clear from the above figures, the shape of the region recognized as the Si crystal part 10 on each image differs slightly due to the difference in reflection characteristics.

Therefore, as shown in FIG. 1(B), an image P stored in the image processing device 9 when the internal illumination light L1 is irradiated,
, and the image P3 when both the internal illumination light L1 and the external illumination light are irradiated are compared, and the parts with no difference in brightness in both images P, P3 are determined using a method similar to the so-called pattern matching method. After recognizing it as the S crystal part 10, the black and white are inverted.

That is, by comparing both images Pl and P3, each image P,
, P, and only the mutually overlapping white portions are recognized as Si crystal portions 10.

In this case, if a flat surface that is not a rough surface happens to exist in the AI2 eutectic layer corroded by etching as described above, this flat surface will exhibit reflection characteristics similar to those of the Si crystal portion 10, which is a smooth surface. Si crystal part 1
It becomes indistinguishable from 0. As the result, image P,
Even on the image P4, which is a composite of 81 and P3, there is a possibility that a portion that is not actually the 81 crystal portion 10 is recognized as the Si crystal portion 10.

Therefore, the composite image P4 is binarized into light and dark, and the binarized image P5 is
At the same time, the external illumination light and the image P at the time of irradiation are converted into brightness and darkness, and both images P and P6 are compared to obtain each image P.
5. Of the black parts on P and P, only the mutually overlapping parts are extracted and recognized as the true region E of the Si crystal part 10 on the image P7.

That is, the external illumination light, the image P at the time of irradiation, and the above, S
Since the i-crystal part 10 is in a state close to a mirror surface, most of the illumination light is reflected specularly, and since the illumination light body is irradiated from an oblique direction, the Si crystal abnormal part O appears as a black part. On the other hand, if the A[eutectic layer 11 has a flat part that does not have a rough surface as described above, the planar part of the AQ eutectic layer 11 has a higher reflectance than the Si crystal part 10 because it is made of metal. , the rate of diffuse reflection of illumination light increases. Therefore 1. The plane portion of the l-eutectic layer 11 appears on the image Pl as a white portion that is brighter than the Si crystal portion 10.

Therefore, by using this property, the binarized images P and P
6 for the black portion, only the true region E of the Si crystal portion 10 is extracted as shown in image P7.

When the true region E of the Si crystal part 10 is extracted, the image processing device 9 calculates the area of the region E as the number of pixels, and calculates the area of the Si crystal abnormal part O from the ratio of this number of pixels to the total number of pixels of the screen. The precipitation area ratio will be calculated and output.

The embodiment shown in FIG. 8 shows an example in which a cylindrical inner circumferential surface 22a of a sample 22 is inspected, and the objective lens 25 is attached at right angles to the axis of the lens barrel 24, while an external illumination light source (not shown) The illumination light Li2 is emitted through the optical fiber 27 so as to intersect with the internal illumination light L11. 28 is a fiber holder, 26 is a reflecting mirror, and 29 is external illumination light L1. 30 is a set screw. The image processing procedure is the same as that described above, and the same effects as in the first embodiment can be obtained in this embodiment as well.

Effects of the Invention As described above, according to the method of the present invention, external illumination light is added to internal illumination light, and each binary value is obtained when both illumination lights are irradiated, when the internal illumination light is irradiated, and when the external illumination light is irradiated. By extracting only the parts where the regions extracted from the transformed image mutually overlap as the true regions of the specific component,
It is possible to extract specific components with high precision by reducing extraction errors due to the influence of external light, etc., and improve the reliability of test results.

[Brief explanation of the drawing]

FIG. 1(A) is a diagram showing one embodiment of the present invention, and FIG. 2 is an enlarged view of the main part when internal and external illumination light is irradiated, and FIG. 1(B) is a diagram showing an embodiment of the present invention.
is a schematic flowchart of the image processing process when extracting a specific component, FIG. 2 is a schematic explanatory diagram of the entire system showing an embodiment of the present invention, and FIG. 3 is an enlarged view of the main part when irradiating internal illumination light. FIG. 4 is an enlarged view of the main part when external illumination light is irradiated, FIG. 5 is an explanatory diagram showing an example of an image when internal illumination light is irradiated, FIG. Figure 7 is an explanatory diagram showing an example of an image when internal and external illumination light is irradiated;
The figure is an enlarged view of main parts showing another embodiment of the present invention, and FIG. 9 is a schematic explanatory diagram of the entire system for explaining a conventional inspection method. 2.22...sample, 4,24...lens barrel, 5,25...
- Objective lens, 7... Internal illumination light source, 8... Camera, 9... Image processing device, 10... Si crystal part (smooth surface), 11... Ag eutectic layer (rough surface) , 12... External illumination light source, L l+ L l 1... Internal illumination light,
L2+ll! ...External lighting light. Figure 1 (B) Sili6m11Bffl flesh image

Claims (1)

[Claims] (1) A method of extracting and inspecting specific components of the surface structure based on the difference in light reflection characteristics between rough and smooth surfaces of the sample, in which internal illumination light is applied to the sample through the objective lens of a microscope. At the same time, external illumination light is irradiated from an oblique direction forming a predetermined angle with the sample surface, and both internal and external illumination light, only internal illumination light, and only external illumination light are irradiated. In each case, the surface structure of the sample captured by the objective lens is imaged by an imaging device, and the image is binarized into brightness and darkness to extract a region of a specific component of the surface structure. A method for inspecting a metallographic structure, characterized in that only a portion where regions extracted from each binarized image mutually overlap is extracted as a true region of a specific component.