JP2005351910A - Defect inspection method and its apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To perform liquid penetrant inspection and magnetic particle inspection, that are non-destructive inspections that have been performed through visual observation, by use of image data. <P>SOLUTION: An image signal, acquired by imaging a sample to be inspected with a color video camera, is used to automatically detect defect candidates containing pseudodefects. Results are displayed on a screen to detect true defects among them. Image data are stored in a storage means to be repeatedly displayed on the screen. In liquid penetrant inspection, the chromaticity of each location in the screen is determined to extract defect candidates on the basis of color differences, and pseudodefects are distinguished from defects on the basis of differential values of color differences. In liquid penetrant inspection, a polarizing filter is used for removing regular reflection due to illumination. In magnetic particle inspection, an ultraviolet cut-off filter is mounted to the camera to prevent noise. By providing both a white illuminating lamp and an ultraviolet illuminating lamp, both inspections can be conducted with one probe. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

The present invention relates to a defect inspection method such as cracking of a metal surface, and more particularly to an inspection method and apparatus for performing nondestructive inspection called penetration inspection and magnetic particle inspection.

Penetration flaw detection and magnetic particle flaw inspection are for nondestructive inspection of defects such as cracks having openings on the metal surface. In penetrant flaw detection, a red liquid called a penetrating liquid is usually applied to the inspection surface, and after a predetermined time has passed, the penetrating liquid is wiped off and a white powder called a developer is applied. If there is a defect such as a crack, the permeate remaining in the crack appears on the surface by capillary action and gives a red defect indication.
On the other hand, in the case of magnetic particle flaw detection, a solution containing fluorescent magnetic particles is sprayed on a test body that is a magnetic body, and the test body is magnetized. If there is a defect such as a crack, the magnetic flux concentrates on the defective part, so that the fluorescent magnetic powder gathers, and when irradiated with ultraviolet rays, emits green light and presents a defect instruction.
Conventionally, these defect instructions are visually observed to inspect the defects.

When visual inspection is performed as in the prior art, defects may be overlooked due to fatigue of the inspector, inspection results may differ due to individual differences among the inspectors, and the inspection results may remain only in characters such as “pass”. There was a problem of inspection reliability that there was no.
As for magnetic particle flaw detection, an automatic inspection device has been developed for parts that are important and mass-produced. However, since it is a dedicated device, it has not been able to easily inspect various shaped parts.
Furthermore, for penetrant flaw detection, it is necessary to detect the surface color as a two-dimensional distribution with high accuracy, so even if there is a color meter that can measure chromaticity with high accuracy at a point, a two-dimensional sweep is required. Therefore, it has been difficult in terms of inspection time and cost to easily and automatically inspect various shaped parts.
Furthermore, when the specimen is large, it may be difficult to know which part of the specimen the image obtained by the automatic inspection is, and which part of the specimen is the detected defect.
Furthermore, if both the penetrant flaw detection and the magnetic particle flaw detection can be automatically inspected with a single device, the economic value is dramatically improved. However, there has never been such a device or technique.
An object of the present invention is to provide a defect inspection method, a defect inspection apparatus, and a defect inspection support method that solve the above-described problems and facilitate the determination of a true defect.
It is another object of the present invention to provide a defect inspection method, a defect inspection apparatus, and a defect inspection support method that make it possible to easily know a defect position even for a large specimen.

In the present invention, in order to achieve the above-mentioned object, the test specimen is imaged using a color video camera. However, if a color video camera is used as it is, in penetrant flaw detection, correct imaging cannot be performed due to specularly reflected light from the specimen under illumination. Also, in magnetic particle flaw detection, because of illumination light (ultraviolet light), foreign matter on the specimen emits blue light, making it difficult to distinguish from defects. For this reason, in the present invention, in order to remove specular reflection light, a polarizing filter is inserted in both the illumination and the camera. In addition, a filter was placed in front of the camera to cut off ultraviolet rays.
In the present invention, the color camera, the white illumination lamp, and the ultraviolet illumination lamp are configured as one probe so that they can be used for both penetration inspection and magnetic particle inspection.
For penetrant flaw detection, calculate the xy chromaticity of the specimen surface from the video signal from the color video camera, detect the red defect indicator, and for magnetic particle flaw detection, perform differential processing on the green video signal, It was made to detect after emphasizing the defect.
In order to prevent over-detection and over-detection of automatic inspection, in the present invention, the inspection result is displayed as a color image, the portion determined to be defective by the automatic inspection is surrounded by a rectangle, and the inspector uses the original image to identify the rectangular portion. I checked each one to determine if it was a real defect. The original image and the inspection result are stored as a record on a magneto-optical disk or the like.
Furthermore, in the present invention, when the test body cannot be fully covered in one field of view like a long object, the inspection position can be specified by placing a scale in the imaging field of view and simultaneously capturing the scale and the inspection image. I made it.
Further, according to the present invention, in order to perform color calibration of the color video camera used in the defect inspection method by the penetrant flaw detection method, R (red), G (green), B (blue), and W (white) standards. Using the signal obtained by imaging the color with the color video camera, create a conversion parameter for the xyY value unique to the color video camera, and once convert the image obtained by imaging the surface to be inspected with the color video camera The defect candidate is detected using the converted image.
Furthermore, in the present invention, the image data of the detected defect is stored in the memory and saved, so that it can be repeatedly displayed on the screen after the inspection. Moreover, the stored image data can be transmitted to a remote place using a communication line.

According to the present invention, since the image input is performed using the color video camera, and further, in the defect inspection by the magnetic particle flaw detection method, the ultraviolet ray reflected from the specimen can be cut by the ultraviolet cut filter. It is possible to easily check the defect inspection result. In addition, since defect candidates are automatically instructed and displayed on the screen, there is almost no oversight of inspection, and in order to save the inspection image, the image saved after inspection is displayed on the screen and the defect is again detected. This can be confirmed and the reliability of the inspection is improved.
Further, according to the present invention, since a color video camera is used, automatic defect inspection for magnetic particle inspection and penetration inspection can be performed with the same sensor probe, which greatly improves convenience.

  Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is an example of a defect inspected by the present invention.
(A) shows an example of a penetrant flaw detection image. A white developer is applied to the test body 1, and a defect 2 (high contrast) and a pseudo defect 3 (low contrast) are observed. In penetrant flaw detection, defect 2 is highlighted as a red pointing pattern. Pseudo defects occur when penetrating liquid accumulates on surface grinding lines, etc. and cannot be wiped cleanly, resulting in a light red indicating pattern.
(B) shows an example of the magnetic particle flaw detection image, and it is assumed that the defect 1 exists in the test body 1 and the fluorescent magnetic powder is already applied and magnetized. When this is irradiated with ultraviolet illumination, the fluorescent magnetic powder collected in the defect 2 due to magnetization emits green light. However, for example, if the test body 1 has a welded portion, the fluorescent magnetic powder collects along the weld bead, and thus the green pseudo defect 3 may be seen.

FIG. 2 is a block diagram of a defect inspection apparatus according to the present invention. The specimen 1 has a defect 2 and a pseudo defect 3. This is imaged by the color video camera 21. The white illumination lamp 24a is turned on when the penetrant inspection image is inspected, and the ultraviolet illumination lamp 24b is turned on during the magnetic particle inspection test. The white illumination lamp 24 a is connected to a white illumination lamp connector 25 a and is connected to the illumination power supply 8 by an illumination cable 26.
At the time of magnetic particle flaw detection, the illumination cable 26 is connected to the ultraviolet illumination lamp connector 25b. A hood 27 is attached to avoid the influence of outside light. In FIG. 2, the illuminating lamp is a ring-shaped lamp, but one or a plurality of bar-shaped lamps may be used.
The color video signal of the color video camera 21 includes a case where R, G, and B are independent signals and a case of a composite video signal. Stored as data. The color image data is analyzed by the computer 5 and the defect detection result is displayed on the color monitor 6.
The defect inspection result is stored in the data storage device 7. Furthermore, if necessary, an image displayed on the color monitor 6 can be printed out by a printer (not shown). Instead of the data storage device 7, it may be stored in a portable storage medium such as an optical disk or a flexible disk.
Furthermore, the inspection result can be transferred to a remote place via a communication line such as the Internet (not shown). As a result, the inspection data can be viewed immediately even at a location far from the inspection point. In addition, by transferring the inspection results via the communication line, data at a plurality of inspection points can be centrally managed at one place.
In front of the lens of the color video camera 21, a polarizing filter 22a and an ultraviolet cut filter 22b are mounted. A polarizing filter plate 23 is installed below the white illumination lamp 24a. The polarizing filter 22a and the polarizing filter plate 23 are used to prevent illumination reflection and regular reflection from the test body 1 during the penetrant inspection. The filter 22a is rotated and fixed at a place where there is no reflection or reflection. The polarization filter 22a can be automatically adjusted based on the video output signal of the color video camera 21.
The ultraviolet cut filter is for preventing unnecessary light emission from adhering foreign matter or the like by the ultraviolet illumination lamp 24b.

FIG. 3 is a diagram showing the effects of the polarizing filter 22a and the polarizing filter plate 23. As shown in FIG.
(A) shows a state in which there is no filter, and (b) shows a state in which each filter is displayed on the monitor screen when each filter is mounted and the rotation angle of the filter is adjusted. In (a), there is a reflection 30 of illumination, which makes defect detection difficult. The reason why the reflection is annular is that it is assumed that the white illumination lamp 24a is annular. In (b), this reflection disappears.

  FIG. 4 is a diagram showing the effect of the ultraviolet cut filter 22b. (A) shows a state in which there is no filter, and (b) shows a state in which it is displayed on a monitor screen when a filter is attached. In (a), light emission from the foreign matter 40 such as lint and regular reflection 41 from the test body 1 are imaged by the color video camera 21, which makes defect inspection difficult. In (b), these noises are cut off, and an image of only light emission by fluorescent magnetic powder is obtained, as in the case where a person visually observes the specimen 1.

First, a method for detecting a crack defect in a penetration inspection image will be described with reference to FIGS. FIG. 5 shows an automatic detection method of the defect 2 in the penetration inspection.
First, the test body imaging 50 is performed on the test body 1 coated with the developer using the white illumination lamp 24a. Next, chromaticity conversion 51 is performed to obtain the xy chromaticity value of each pixel from the obtained R, G, B color image data.
Next, reference white determination 52 for calculating the reference white chromaticity of the developer is performed, and hue / color difference calculation 53 is performed at each position on the image with respect to the reference white.
Thereafter, in order to perform defect candidate area extraction 54, an area having a specific range of hue and color difference is extracted by binarization.
The true defect 2 has a clear outline, and the pseudo defect often has an unclear outline. For this reason, differentiation 55 of the color difference image is performed to obtain the color difference change ratio of the contour portion of the extracted defect candidate area. Next, shape measurement 56 such as the area, aspect ratio, and length of the defect candidate region is performed. After that, in the defect detection 57, only a region having a large color difference change ratio and a length and area more than a specified value is detected as the true defect 2. Further, the inspection result is displayed on the color monitor 6 and after the defect is confirmed by the inspector, the image data, shape data, position information, etc. are printed out in the storage device 7 or stored as a hard copy (58).

In color inspection, it is necessary to quantitatively evaluate the color. Therefore, in the step of chromaticity conversion 51, RGB data of the captured color image is converted into chromaticity x, y and luminance Y defined by CIE (International Lighting Commission), and inspection is performed using these. A representation of chromaticity x, y in two-dimensional orthogonal coordinates is called a chromaticity diagram and is shown in FIG. In the chromaticity diagram, each color is arranged around white as a center, and each color becomes brighter as it goes away from white. Hereinafter, the hue is called hue, the vividness of each color is called saturation, and the distance between two chromaticity values on the chromaticity diagram is called color difference. FIG. 6 shows the chromaticity range 50 of the penetration flaw detection image.
In this method, in order to perform conversion from RGB data to chromaticity x, y, and luminance Y with high accuracy, color calibration is performed in advance using a camera calibration color chart 71 as shown in FIG. The processing flow is shown in FIG. The camera calibration color chart 71 is painted with three or more colors. This is imaged by the color video camera 21 (81), and the RGB value of each color is calculated (82). Further, the chromaticity x, y and luminance Y are measured by the colorimeter 72 (83). Here, the relationship between the RGB value and the xyY value is expressed by (Equation 1) and (Equation 2).

Here, X, Y, and Z are called tristimulus values.

Therefore, by calculating the xyY value by substituting the RGB value of each color captured from the camera into (Equation 1) and (Equation 2), and obtaining a11 to a33 such that this value matches the xyY value measured by the colorimeter. Therefore, a conversion parameter specific to the camera is obtained. Since there are nine unknown parameters, the parameters can be calculated from the RGB values (R1 G1 B1) to (R3 G3 B3) of at least three colors and the corresponding xyY values (x1 y1 Y1) to (x3 y3 Y3). .
From (Equation 2), XYZ can be calculated by the following (Equation 3) from the xyY value.

The xyY values of the three colors of the colorimeter are substituted into (Equation 3) to obtain XYZ, and are substituted into (Equation 1).

As a result, camera-specific conversion parameters a11 to a33 are obtained (84), and an xyY value equal to the value of the colorimeter can be obtained from the RGB values of the camera.
As the form of the camera calibration color chart 71 used for the penetrant flaw detection image inspection, the present inventors have used R, G, B, and W reference colors and several colors that change from white of the developer to red of the penetrant. Is suitable for color calibration.
From the chromaticity range 60 of the penetrant flaw detection image in FIG. 6, a white color close to a developing solution that changes from white to red, a pink color of a defect candidate, and a red color corresponding to a defect are selected in stages, and the xymeter 72 uses xyY. The reproducibility of the conversion parameter can be confirmed by measuring the value and calculating and comparing each xyY value from the value and the conversion parameter of the xyY value specific to the camera to be used. Periodically (preferably before inspection) during penetrant image inspection, color reproducibility is confirmed using the color chart 71 of FIG. 6 to realize reliable and highly accurate chromaticity measurement. There is an effect that can be done.
In addition, the present inventors have confirmed that the reference color of the color chart differs depending on the color temperature of the illumination light source. That is, there is a possibility that the chromaticity range of FIG. 6 is different between the case where the penetration inspection image inspection is performed using the white illumination lamp used in the embodiment of FIG. 2 as the light source and the case where the penetration inspection image inspection is performed using the halogen lamp as the light source. Therefore, it is necessary to select the chromaticity for calibration according to the light source.
Also, the chromaticity from white to red varies depending on the color of the surface condition of the object to be inspected (such as a metal surface such as stainless steel, an iron-based blackish surface, or a rusted brownish surface). For this reason, the number of interpolated colors in white, pink, and red is increased, and the color calibration can be performed with higher accuracy as the color calibration is performed.
Further, it is preferable to select a reference color of red, green, blue, and white and a chromaticity that changes from white to red according to the chromaticity of the developer or penetrant.

  After converting the RGB values obtained from the camera to xyY values using the conversion parameters specific to the camera calculated in advance by calibration and calculating the chromaticity distribution in the image, at 52, the developer is extracted from the image. , That is, the chromaticity of a portion that is not defective is calculated as a reference value. First, the chromaticity x and y of each pixel in the image is examined, and the number of pixels taking each x and y value is counted as shown in the graph of FIG. 9A to create a two-dimensional frequency distribution of chromaticity. Then, an x chromaticity value (FIG. 9B) and a y chromaticity value (FIG. 9C) having the largest number of pixels in the image are obtained. Since most of the image is not a defect, the x and y chromaticity values of the peak value of the two-dimensional frequency distribution are the xy chromaticity values of the reference white.

In 53, the hue and color difference at each position on the image with respect to the reference white color are calculated.
If the chromaticity of the reference white is (xc, yc) and the chromaticity at the position (i, j) on the image is (xij, yij), the hue at the position (i, j) is as shown in FIG. The calculation is performed in the direction with respect to the reference color on the chromaticity diagram. The calculation formula is shown in (Formula 5).

Further, as shown in FIG. 11, the color difference at the position (i, j) is calculated by the distance from the reference color on the chromaticity diagram. The calculation formula is shown in (Formula 6).

From the hue / color difference at each position of the image with respect to the reference white calculated as described above, the range to be detected as a defect in hue is limited as shown in FIG. 12 (in the figure, hue θ is θ1 ≦ θ ≦ θ2). The range of color vividness is limited by the color difference (in the figure, the color difference d is in the range of d1 ≦ d ≦ d2). Then, a portion within this range is extracted as a defect candidate region.

Thus, some defect candidates obtained by limiting the range by hue and color difference do not need to be detected as defects. For example, when the chromaticity gradually changes with respect to the reference white, not a defect but a region with a clear outline is a defect.
Thus, a gentle color change with respect to surrounding colors is regarded as a normal part or a pseudo defect 3, and only a rapid change is regarded as a defect 2. In 55, the amount of change in color difference from the reference white is obtained for the defect candidate area, and only those having a certain value or more are determined as defects.

  Referring to FIG. 13, (a) shows a defect candidate area 131 extracted by 54. (B) 133 is a color difference graph with reference white on 132 of (a). Further, the amount of change in the color difference 133 at each position on 132, that is, a product obtained by differentiating 133 is the color difference differential distribution 134 of (d). In this way, when the amount of change in color difference from the reference white is small, the differential value is also small. Here, as shown in (d), only those having a differential value larger than a certain value 135 are defined as defect areas. As a result, as shown in (c), only the defect area 136 having a large color difference and a large color difference change amount, that is, a sharp outline is detected.

  Next, a method for determining the threshold value 135 will be described with reference to FIG. In the graph of FIG. 12A, the vertical axis represents the maximum value of the color difference in each defect candidate area extracted by hue and color difference, and the horizontal axis represents the maximum value of the color difference differential value of the contour portion of each defect area. The value of defect 2 is plotted with x, and the value of pseudo defect 3 is plotted with ◯. Reference numeral 141a denotes a frequency distribution of each color difference differential value, and 142a denotes a frequency distribution of the color difference values. When the defect and the pseudo defect are clearly separated, the pass / fail judgment line 144a is a straight line 144a that passes through the peak value of the valley of the frequency distribution of 141a and 142a and is perpendicular to the principal axis of inertia 143a of the plotted point. When the defect and the pseudo defect are not separated, that is, when there is no peak of the valley of the frequency distribution, the determination line is 144b. That is, all defect candidate areas are detected as defects so as not to be overlooked and overlooked.

Next, a defect detection method in magnetic particle inspection will be described with reference to FIGS. 14 and 16.
FIG. 14 is an example of an image processing algorithm for analyzing the contents of the image memory 7 at the time of magnetic particle flaw detection. An RGB image is taken in (151), and then a G image differentiation process in which most of the emission information of the fluorescent magnetic powder is input is performed (152). As a result, a portion having a large linear luminance change such as a crack defect is emphasized, and a luminance is high such as a magnetic powder reservoir, but a portion having a small luminance change is not emphasized.
Next, a threshold value for binarization is determined from the average value of the G differential image, and binarization is performed (153). Image noise such as isolated points is removed from the binarized image (154), defect candidates are obtained, and the length, contrast, etc. of these defect candidates are calculated (155). If it is larger than the specified value, it is determined as a defect.

  FIG. 16 shows a method for distinguishing between defects and pseudo defects. For example, as shown in (a), when the luminance distribution of the defect 2 and the pseudo defect 3 is taken on a line 161, a luminance distribution 162 as shown in (b) is obtained. The luminance values of the defect 2 and the pseudo defect 3 are approximately the same. When the luminance distribution 162 is differentiated, a luminance differential distribution 163 as shown in (b) is obtained. The defect 2 has a sudden change in luminance, and the pseudo defect 3 changes gently. Therefore, by determining the result of the differential processing using the determination threshold 164 as shown in (d), Only the defect 2 can be extracted as shown in (c).

Now, defect confirmation and data storage will be described with reference to FIG. The defect should be separated from the pseudo defect by automatic inspection and only the defect should be extracted, but in order to prevent oversight and misjudgment, the defect is visually confirmed at the end whether it is penetration inspection or magnetic particle inspection. Do.
FIG. 17 is a flowchart showing a defect confirmation process. First, a defect candidate marker is displayed on a portion determined as a defect by automatic determination (171). Next, the computer 5 requests the inspector to determine each defect candidate one by one (172).
The inspector looks at the color original image to determine whether it is a true defect (173), and when it is recognized as a true defect, the position, length, contrast, etc. of the defect are registered in the data storage device 7. (174), the marker turns red (175).
If the inspector determines that the defect is a pseudo defect in the defect candidate confirmation, the marker is deleted (176). If a defect candidate still remains, a marker is displayed on the next defect candidate. When all the defect candidates have been confirmed (177), the color original image is stored in the data storage device 7 (178).
FIG. 18 shows an example of a defect candidate marker generation method. A center line 182 connecting the starting point P1 and the ending point P2 of the defect candidate 181 is obtained, and the long sides AB and CD of the defect candidate marker 183 are set at a position parallel to the center line 182 by a predetermined value m. The short sides AD and BC are determined in the same manner. The length of the defect is the distance between P1 and P2. At the time of magnetic particle flaw detection, the contrast related to the depth of the defect is scanned from the contrast calculation line 184 from P1 to P2, and the difference between the average luminance and the maximum luminance on this line is taken, and this difference is obtained from P1 to P2. The average value is used as the defect contrast. The defect candidate marker is not necessarily rectangular. A method of making the short sides AD and BC semicircular may be used, and an important point is to prevent the defect from being hidden by the marker.
FIG. 19 shows an example of a defect candidate display method in the color monitor 9. In order from the candidate with the longest defect length, the inspector is requested to confirm the original image. Initially, all the markers are displayed in white, the markers that are determined to be true defects are changed to another color, for example, red, and those that are determined to be pseudo are deleted.
FIG. 20 shows an example of a method for specifying the inspection position when the test body 1 is a long object. The scale 201 with the scale is fixed to the test body 1 and imaging is performed so that the scale 201 enters a part of the camera visual field 202. Regarding the scale of the scale, for example, a method of writing numbers for every centimeter can be considered. Moreover, the scale 201 for penetrant flaw detection and the scale 201 for magnetic particle flaw detection can be different in color. For example, when penetrating flaw detection, the scale 201 is, for example, red scales and numbers on a white background, and when magnetic particle flaw detection is performed, the scale numbers are green fluorescent color scales and numbers on a white background.
FIG. 21 shows an example of the captured screen. The scale 201 is simultaneously imaged at the bottom of the screen, and the camera position on the specimen 1 is calculated from the scale 201. That is, scale numbers 210 are written on the scale 201 and can be recognized by the pattern matching method using the computer 5. Further, the scale 201 includes a dividing line 211 for every centimeter, for example, so that a more detailed camera position can be calculated. A cross-sectional signal 213 is obtained as an image signal of the search lines 212 of C1 and C2 on the image. From this, the positions of the left end A and right end B of the image and the positions E1, E2, E3, E4, and E5 of the dividing line 211 are known. The imaging magnification can be calculated from the positions E1, E2, E3, E4, and E5, and together with the scale number 210, the exact position of the defect 2 on the specimen 1 can be known.
The above inspection results are stored in the storage device 7, and an example thereof is shown in FIG.
When the surface to be inspected of the test body 1 is large and the entire inspection surface cannot fit into one inspection screen, the image is divided into several images and inspected. At this time, the images to be divided are set so that the imaging ranges slightly overlap on the inspection surface. Reference numerals 221a, 221b, and 221c denote divided images of the test body 1, respectively. The inspection is performed on each divided image. As a result, as shown in 222, the entire image information is stored together for each specimen, and information such as position, length, area, chromaticity, and hue is also stored for each defect. Stored in
The inspector first displays and examines the data 222 for each test specimen stored in the storage device 7 on the screen of the monitor 6, and if he wants to see in detail the defective part, the test specimen name and A corresponding divided image can be called from the image NO and displayed on the screen of the monitor 6. At this time, information such as the position, length, area, chromaticity, and hue of the defect stored in association with the displayed image data can also be displayed on the screen of the monitor 6.
Further, by displaying the detected defect candidates with emphasis using a marker or the like on the screen, defects larger than 0.1 to 0.3 mm, which are the same as those conventionally performed by visual inspection, are displayed. , It can be prevented from being missed on the screen.
Furthermore, by increasing the detection magnification of the image, it becomes possible to detect defects that are finer than the size that can be visually observed. Then, by displaying on the screen a defect finer than the visible size, the position, length, area, chromaticity, hue, etc. of the finer defect than the visible size are also displayed. You can check it on the screen.
Instead of the data storage device 7, it may be stored in a portable storage medium such as an optical disk or a flexible disk. By storing in such a storage medium, the inspection result can be displayed on the display even at a location away from the inspection point.
Further, the inspection result can be transferred to a remote place via a communication line such as the Internet. As a result, the inspection data can be viewed immediately even at a location far from the inspection point. In addition, by sending the inspection result through the communication line in this way, it is possible to instruct immediately from a remote place such as re-inspection and addition / change of the inspection location, so that the efficiency of the inspection can be improved.

FIG. 1 is a perspective view showing an example of an inspection object handled in the present invention. FIG. 2 is a front view showing a schematic configuration of a defect inspection apparatus showing an embodiment according to the present invention. FIG. 3 is a front view of a monitor screen showing the effect of the polarizing filter in the apparatus configuration of FIG. FIG. 4 is a front view of a monitor screen showing the effect of the ultraviolet cut filter. FIG. 5 is a flowchart showing the flow of the automatic inspection method in the penetrant flaw detection according to the present invention. FIG. 6 shows an xy chromaticity diagram. FIG. 7 is a front view showing a configuration for camera calibration. FIG. 8 is a flowchart showing the flow of the camera calibration process. FIG. 9 is a diagram illustrating a method for obtaining the reference white chromaticity from the color difference image. FIG. 10 is a diagram for explaining a method of calculating a hue on a chromaticity diagram. FIG. 11 is a diagram illustrating a method for calculating a color difference on a chromaticity diagram. FIG. 12 is a diagram showing a method for obtaining a defect candidate area from the hue obtained in FIG. 10 and the color difference obtained in FIG. FIG. 13 is a diagram showing a method of determining a defect area by determining a pseudo defect from the defect candidate area obtained in FIG. FIG. 14 is a diagram for explaining a method for determining a threshold value 135 for determining a defect area by determining a pseudo defect from the defect candidate area shown in FIG. FIG. 15 is a flowchart showing an example of an image processing algorithm in magnetic particle flaw detection according to the present invention. FIG. 16 is a diagram for explaining a method for determining pseudo defects in magnetic particle flaw detection according to the present invention. FIG. 17 is a flowchart showing the process of defect confirmation and data storage in the present invention. FIG. 18 is a diagram illustrating an example of a defect candidate marker generation method according to the present invention. FIG. 19 is a diagram showing an example of a defect candidate display method according to the present invention. FIG. 20 is a diagram showing an example of a method for specifying an inspection position according to the present invention. FIG. 21 is a diagram showing an example of an inspection image containing information for specifying an inspection position according to the present invention. FIG. 22 is a diagram illustrating an example of a configuration of inspection result data stored in the storage device 7.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Test body 2 ... Crack defect 3 ... Pseudo defect
DESCRIPTION OF SYMBOLS 4 ... Color image memory 5 ... Computer 6 ... Color monitor 7 ... Data storage device 21 ... Color television camera 22a ... Polarization filter 22b ... Ultraviolet cut filter 23 ... Polarization Filter plate
24a ... White illumination lamp 24b ... UV illumination lamp 71 ... Color chart for camera calibration
183 ... Defect candidate marker 201 ... Scale 711 ... Color chart for camera calibration.

Claims (12)

  A defect inspection method using a penetrant flaw detection method, which uses information obtained by imaging a surface to be inspected using a color video camera and converting the chromaticity of the image obtained by imaging to a reference chromaticity And inspecting the surface to be inspected for defects.   A defect inspection method using a penetrating flaw detection method, in which an inspection surface of an inspection object is illuminated with polarized light, and the inspection surface illuminated with the polarized light is imaged with a color video camera via a polarizing filter, and the imaging A defect inspection method comprising: extracting defect candidates on the surface to be inspected from an image obtained by converting the chromaticity of an image obtained in this way into a reference chromaticity, and displaying the extracted defect candidate image.   The inspection target surface of the inspection object is imaged with a color video camera with the position information of the visual field in the visual field of the color video camera, and defect candidates in the inspection surface are extracted from the image obtained by the imaging The defect inspection method according to claim 1, wherein the extracted defect candidate image is displayed on a screen together with position information of the visual field.   The defect inspection method according to claim 3, wherein the position information of the visual field is based on a scale arranged in the visual field.    The defect inspection method according to claim 1, wherein the surface to be inspected in the previous period is imaged over a plurality of fields of view with the color video camera.   Conversion coefficients for imaging R, G, B, and white color samples (color charts) with the color video camera and converting the RGB color values specific to the captured color video camera into reference xy chromaticity values by calculation A color image obtained by imaging the surface to be inspected of the inspection object is converted into the xy chromaticity, and a defect of the surface to be inspected is detected using the converted color image. The defect inspection method according to claim 1 or 2.   R, G, B, white, developer (white or equivalent color) and penetration flaw detection solution (red or equivalent color), developer (white or equivalent color) and penetration flaw detection used in the color video camera An image of a color sample (color chart) using a mixed color due to a chemical change of the liquid (red or equivalent color) is taken, and the xy chromaticity based on the RGB chromaticity value specific to the color video camera obtained by the imaging A conversion coefficient to be converted into a value is obtained, a color image obtained by imaging the surface to be inspected of the inspection object is converted into the xy chromaticity based on the obtained conversion coefficient, and the converted color image is used. The defect inspection method according to claim 1, wherein a defect on the surface to be inspected is detected.   The defect inspection method according to claim 7, wherein the color samples (color charts) are formed on the same surface.   A defect inspection apparatus using a penetrating flaw detection method, wherein an illuminating unit illuminates a surface to be inspected of an inspection object, an imaging unit that images the surface to be inspected illuminated by the illuminating unit with a color video camera, and the imaging unit Defect candidate extracting means for extracting defect candidates on the inspection surface from the image of the inspection surface obtained by imaging in step, display means for displaying an image of defect candidates extracted by the defect candidate extraction means, and the defect Storage means for storing defect candidate images extracted by the candidate extraction means together with information on defects, and image selection for selecting a desired image from defect candidate images stored in the storage means and displaying the selected image again on the display means And a defect inspection apparatus.   A defect inspection apparatus using a penetrating flaw detection method, wherein an illuminating unit illuminates a surface to be inspected of an inspection object, an imaging unit that images the surface to be inspected illuminated by the illuminating unit with a color video camera, and the imaging unit Conversion coefficients for converting the RGB chromaticity values specific to the color video camera obtained by imaging R, G, B, and white color samples (color charts) with the color video camera to a reference xy chromaticity value are obtained. Penetration flaw detection for extracting defect candidates due to penetration flaw detection on the surface to be inspected from an image obtained by converting the image of the surface to be inspected obtained by imaging with the calculation means and the conversion coefficient obtained by the calculation means A defect inspection apparatus comprising at least defect candidate extraction means and display means for displaying an image of defect candidates extracted by the penetration flaw detection defect candidate extraction means.   11. The defect inspection apparatus according to claim 9, wherein position information display means for displaying position information of the visual field is disposed within a visual field of the color video camera that images the surface to be inspected. The defect inspection apparatus according to claim 11, wherein the position information display unit is a scale.

Images