JP2000230909A - Apparatus for inspecting flaw of heat insulating material - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a flaw inspecting apparatus capable of efficiently and accurately inspecting a flaw of a heat insulating material, which is formed by molding glass wool by using an adhesive, by image processing. SOLUTION: The flaw inspecting apparatus is constituted so that a test heat insulating material X is supported in parallel to a standard heat insulating material Y (support mechanism 1), the end surfaces of the respective heat insulating materials are caught within the same visual field to be photographed to form images, and the color data of the test heat insulating material is judged on the basis of the color data of the standard heat insulating material in the images to detect a flaw (a camera 2, an end surface image processing part 3, a flaw judging part 4). The peripheral surface images consisting of the blue and red light components of the surface of the test heat insulating material are individually obtained, and the images of respective color components are discriminated on the basis of a predetermined threshold value to respectively detect a discoloration flaw and an uneven flaw (cameras 5,6, peripheral surface image processing parts 8, 9).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a defect inspection apparatus capable of efficiently and accurately inspecting a surface defect of a heat insulating material formed by molding glass wool into a plate or a cylinder using an adhesive by image processing. .

[0002]

[Related Background Art] A heat insulating material (heat insulating material) formed by molding glass wool into a plate shape or a cylindrical shape using an adhesive is used as a heat insulating material for construction embedded in a wall or the like, or a coating material for covering an outer periphery of a pipe. Used as By the way, in this type of heat insulating material, glass wool is partially peeled off from its surface due to various manufacturing factors, and glass wool (cut pieces) peeled off from the surface adheres to other parts. Defects may occur. In the case of a cylindrical heat insulating material, wrinkles may be partially generated due to uneven winding of glass wool. Such a defect is called an irregularity defect.

[0003] In addition, the moisture mixed during the molding of the heat insulating material causes defects such as partial solidification of the glass wool as the material, and solidification of the adhesive used for the molding in a local lump. Sometimes. In particular, the partial solidification of glass wool caused by the water is called unburned,
Appears as white color unevenness on the surface of the heat insulating material that is colored yellow. The solidified portion of the adhesive appears as brown color unevenness.

Conventionally, a visual inspection of the surface of a heat insulating material has been performed exclusively by an inspector in order to exclude a heat insulating material having such unevenness defects and color unevenness defects as defective products. However, the inspection requires a great deal of labor and labor, and the inspection efficiency is extremely poor.

[0005]

Therefore, as a method of automating the above-mentioned defect inspection, it has been considered that a surface image of the heat insulating material is captured using a camera and the above-described defect is detected by using an image processing technique. I have. However, the surface of the heat insulating material has a dark yellow fiber pattern, and there is a problem that it is difficult to clearly distinguish the fiber pattern from color unevenness and light and dark due to unevenness on an image. In particular, when image processing is performed to enhance color unevenness, even the fiber pattern is emphasized, and it is difficult to distinguish between light and dark due to unevenness. Conversely, if the fiber pattern is erased by image processing in order to detect the irregularities on the surface of the heat insulating material, there is a problem that even the information on the color unevenness is lost.

The present invention has been made in view of such circumstances, and an object of the present invention is to easily and easily distinguish surface defects of a heat insulating material by image processing while clearly distinguishing the unevenness defects and color unevenness defects. It is an object of the present invention to provide a heat-insulating-material defect inspection apparatus capable of inspecting accurately and efficiently.

[0007]

SUMMARY OF THE INVENTION In order to achieve the above-mentioned object, a defect inspection apparatus for a heat insulating material according to the present invention is designed to detect a defect in a heat insulating material formed by molding glass wool into a plate or a cylinder using an adhesive. A support mechanism for inspecting by image processing, supporting a test insulation material (inspected product) to be inspected side by side with a standard insulation material (non-defective product) serving as a reference for defect inspection;
The cut end faces of the test heat insulating material and the standard heat insulating material supported by the support mechanism are simultaneously captured in the same field of view and imaged at the same time. An end face defect detecting means for judging color information of the test heat insulating material to detect a defect of a cut end face in the test heat insulating material, and capturing an image composed of a blue light component on the surface of the test heat insulating material to obtain a normal surface An image in which white and brown discolored portions (color unevenness portions) are emphasized with respect to yellow as a color is obtained, and the image of the blue light component is discriminated by a predetermined threshold value, and a discoloration defect (color unevenness) on the surface of the test heat insulating material is obtained. (Defect) detecting means for detecting discoloration defect, and an image in which the image composed of red light components on the surface of the test heat insulating material is imaged to emphasize the lightness and darkness due to irregularities that do not depend on the fiber pattern or color unevenness on the surface. Obtained, it is characterized by comprising a concavo-convex defect detecting means for detecting the irregular defect on the surface of the image of the red light component and discrimination at a predetermined threshold value the test insulation.

Preferably, in the discoloration defect detecting means, a monochromatic image (monochrome image) composed of a blue light component on the surface of the test heat insulating material is passed through a blue transmission filter mounted on a camera, for example. )
The unevenness defect detecting means is characterized in that a monochromatic image (monochrome image) composed of a red light component on the surface of the test heat insulating material is obtained through a red transmission filter. In this case, two cameras equipped with a blue transmission filter and a red transmission filter, respectively, may simultaneously obtain the images of the respective color components, or by exchanging the filters mounted on the cameras, Images may be obtained sequentially.

Alternatively, in the discoloration defect detecting means, the test heat insulating material is illuminated with blue light and an image of the surface is taken, so that a monochromatic image (monochrome image) composed of a blue light component is obtained. In the uneven defect detecting means, a monochromatic image (monochrome image) composed of a red light component is obtained by illuminating the surface of the test heat insulating material with red light and imaging the surface. . In this case, the images of the respective color components may be obtained while sequentially switching the light source for illuminating the test heat insulating material.

In a preferred aspect of the present invention, the support mechanism supports a peripheral surface of a heat insulating material formed into a cylindrical shape and rotates the heat insulating material, for example, It is desirable to provide two rollers and a V-shaped conveyor so as to be able to continuously take an image of the entire peripheral surface while rotating the heat insulating material. Further, as described in claim 5, each of the defect detection means includes means for identifying and highlighting a portion of the image of the test insulation material where a defect is detected, for example, by changing its display color. Is preferred.

[0011]

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a perspective view showing a defect inspection apparatus for a heat insulating material according to an embodiment of the present invention. FIG. 1 is a schematic configuration diagram of a defect inspection apparatus configured to inspect a surface defect of a cylindrical heat insulating material, and 1 is a support mechanism for placing a cylindrical test heat insulating material X and performing defect inspection. It is. The support mechanism 1 includes, for example, a pair of rollers 1a and 1b (rotating members) arranged in parallel, and is driven to rotate in one direction by a motor 1c to rotate a test heat insulating material X placed thereon. Provide for defect inspection.
At a position below the rollers 1a and 1b of the support mechanism 1, a non-defective standard heat insulating material Y1 used as a reference for defect inspection includes a test heat insulating material X placed on the rollers 1a and 1b and a shaft thereof. The directions are aligned and the cut end faces are aligned. In other words, the test heat insulating material X is formed by aligning the cut ends of the standard heat insulating material Y1 with the rollers 1a and 1b.
It is placed on the top and subjected to defect inspection.

The roller 1 in the support mechanism 1
As described above, one end side of each of the a and 1b is aligned with the standard heat insulating material Y1 and the cut end face thereof, and the test heat insulating material X placed on the rollers 1a and 1b is aligned with the axis thereof to perform the defect inspection. A non-defective standard heat insulating material Y2 used as a reference is arranged. Each of these standard heat insulating materials Y1 and Y2 is basically made of the same specification as the test heat insulating material X, but is used as each state sample (reference) of the cut end surface and the peripheral surface (surface). Therefore, it may be a short one cut to a predetermined length.

The cut end faces of the test heat insulating material X supported by the support mechanism 1 and the cut end portions of the standard heat insulating material Y1 arranged side by side with the test heat insulator X are located at the positions opposed to the cut end portions. A color camera 2 is provided which captures the respective end faces as one image at the same time by arranging them in the same visual field. Each of the cut end faces of the test heat insulating material X and the standard heat insulating material Y1 is uniformly illuminated by, for example, a white light source (not shown), and is imaged by the color camera 2 under this illumination condition. And the above color camera 2
Test insulation material X and standard insulation material Y imaged by
For example, as shown in FIG. 2A, an end face image E (E
(x, Ey) is subjected to predetermined image processing via an end face image processing unit 3, and then given to a defect determination unit 4 composed of a microprocessor or the like so as to be inspected for defects as described later. .

On the other hand, above the support mechanism 1, the upper surface (peripheral surface) of the heat insulating material X rotating on the rollers 1a and 1b, and the standard heat insulating material X disposed at the end of the shaft of the test heat insulating material X. Two line scanning type monochrome cameras 5 and 6 for scanning the material Y2 in the axial direction and capturing an image are provided side by side. These monochrome cameras 5 and 6 have a blue transmission filter (not shown) and a red transmission filter (not shown) mounted on their lens portions, respectively, and are provided beside the test heat insulating material X and the standard heat insulating material Y2. The peripheral surfaces of the test heat insulating material X and the standard heat insulating material Y2 illuminated from above the side by the arranged white light sources 7a and 7b are imaged through the filters, respectively, so that a monochromatic image of a blue light component is obtained. (Monochrome image) and a monochromatic image (monochrome image) of a red light component. In particular, each of the monochrome cameras 5 and 6 has the roller 1 as described above.
a, 1b, the peripheral surfaces of the test heat insulating material X and the standard heat insulating material Y2 are continuously imaged.
In addition, a peripheral image of the peripheral surface of the standard heat insulating material Y2 is obtained as an image developed in a plane.

Thus, the test insulation X and the standard insulation Y2 imaged by the monochrome cameras 5 and 6, respectively.
For example, the peripheral images Cb (Cbx, Cby) and Cr (Crx, Cry) as shown in FIGS. 2B and 2C, respectively, composed of the blue light component B and the red light component R on the peripheral surface of FIG. After being subjected to predetermined image processing (binarization) via the plane image processing units 8 and 9, the image data is given to the defect determination unit 4 and inspected for defects as described later.

Here, it is assumed that a blue transmission filter and a red transmission filter are attached to the two monochrome cameras 5 and 6, respectively, to obtain peripheral images Cb and Cr of a blue light component B and a red light component R. However, by using only one monochrome camera and alternately exchanging filters attached to the monochrome camera, the peripheral images Cb and Cr of the blue light component B and the red light component R are alternately captured and input. You may do it. Instead of selectively mounting the blue transmission filter and the red transmission filter on the monochrome camera, the light itself for illuminating the peripheral surfaces of the test insulation X and the standard insulation Y2 is changed to illumination light of a blue light wavelength and red light. The illumination light of the light wavelength may be alternately switched, and the peripheral images Cb and Cr of the blue light component B and the red light component R described above may be obtained under these illumination light conditions. In this case, it is of course possible to use the above-mentioned blue transmission filter and red transmission filter together.

Here, the images picked up as described above and defects in the test heat insulating material X will be described. The color unevenness that occurs on the surface of the heat insulating material is mainly yellow, which is a fibrous pattern, and the white part of the unburned part caused by the moisture contained in the production and the lump of the adhesive used in the molding And a brown part. When trying to detect such a white portion and a brown portion from the yellow region which is the main component, for example, under illumination conditions of white light in the entire wavelength region, the luminance level difference between white and yellow is small, and It is difficult to reliably detect these discolored (color unevenness) portions because the difference in luminance level between the color and the brown is small. Moreover, if there is uneven illumination or the illumination conditions change, it becomes more difficult to detect them.

Therefore, in the present apparatus, as described above, only the blue light component (blue wavelength region) on the peripheral surface of the heat insulating material is captured as a single-color image, so that the brightness level difference between the white portion and the brown portion is emphasized. This ensures that color unevenness consisting of a white portion and a brown portion can be detected. That is, white is a light component including the entire wavelength range, and even if only the blue component is selectively extracted through the blue transmission filter, the luminance level does not drop.
On the other hand, brown has a lower luminance level than yellow and contains a lot of red components. When only the blue component is selectively extracted through a blue transmission filter, the luminance level drops significantly (decreases). . In contrast to such white and brown colors, yellow, which is the main component, consists of intermediate wavelength components between red and green. Even if only the blue component is selectively extracted through a blue transmission filter, the luminance level of the yellow component is reduced. The dip is slight.

Therefore, when the blue component of each of the white, yellow, and brown color components is selectively extracted through the blue transmission filter, the luminance level changes in the order of blue [white]>yellow>(red)> brown, and The color change (color unevenness) of white and brown is emphasized with respect to the surface of the heat insulating material mainly composed of, and the white portion becomes bright (high luminance) and the brown portion becomes dark (low luminance). The monochrome camera 5 equipped with the blue transmission filter captures the peripheral image Cb of the blue light component (blue wavelength region) B having such properties to detect a white and brown discolored portion (color unevenness portion). We are trying to make it easier.

On the other hand, the unevenness existing on the surface of the heat insulating material appears as a shadow due to the illumination light from the side. However, the shadow caused by the unevenness is difficult to distinguish from the fiber pattern appearing on the surface of the heat insulating material. Therefore, in the present apparatus, as described above, by capturing only the red light component (red wavelength region) on the peripheral surface of the heat insulating material as a single-color image, information emphasizing only the shadow caused by the unevenness is obtained, and the unevenness defect is reliably detected. So that it can be detected.

That is, since the white color is a light component including the entire wavelength range, even if only the red component is selectively extracted through the red transmission filter, the luminance level does not drop. Also, since brown contains many red components, even if only the red components are selectively extracted through the red transmission filter, the luminance level does not drop. Further, yellow, which is the main component, has an intermediate wavelength component between red and green. Even if only the red component is selectively extracted through a red transmission filter, the drop in the luminance level is slight. The yellow fiber pattern has a small color difference between light yellow and dark yellow. If only the red component is selectively extracted through a red transmission filter, the difference is almost lost.

As a result, when the peripheral surface image Cr (Crx, Cry) of the heat insulating material is obtained through the red transmission filter, the brightness levels of the white portion, the brown portion, and the yellow portion forming the fiber pattern are substantially equal. , It becomes almost impossible to discern these differences. On the other hand, since the shadow caused by the unevenness appears as a difference in the brightness level, the image C in which the shadow caused by the unevenness on the surface of the heat insulating material is emphasized.
r can be obtained, and the unevenness defect can be easily detected without being affected by color unevenness or fiber pattern.

On the other hand, on the cut end surface of the heat insulating material, a yellow layered pattern appears due to the winding of glass wool.
A rarely unburned white portion may be exposed on the cut end surface. In addition, the color unevenness on the cut end surface often appears as a more subtle color change than the color unevenness on the heat insulating material surface described above. Therefore, in this apparatus, in order to reliably detect the subtle color change, an image E of the cut end face is obtained as a color image having a much larger amount of color information than a single-color image, and a white discolored portion (color) is obtained from the luminance information. (Uneven portion). In particular, since the color image tends to have a difference in color information depending on the lighting conditions, the cut end face image Ey of the standard heat insulating material Y1 and the cut end face image Ex of the test heat insulating material X are simultaneously obtained on one image. By determining the cut end face image Ex of the test heat insulating material X on the basis of the color information of the cut end face image Ey of the standard heat insulating material Y1, the discolored portion (color unevenness portion) can be surely obtained regardless of the fluctuation of the lighting condition. Is to be detected.

According to the apparatus thus constructed, the color camera 2 has an end face image obtained by arranging the cut end faces of the standard heat insulating material Y1 and the test heat insulating material X side by side as shown in FIG. An inspection image E including Ey and Ex is obtained. In addition, the monochrome cameras 5 and 6 are developed in a state where the peripheral surface is developed with the rotation of the test heat insulating material X, and for example, a peripheral image Cb of a blue light component as shown in FIG.
(Cbx, Cby) and the peripheral image Cr (Crx, Cry) of the red light component as shown in FIG.

As shown in FIG. 2 (a), the end face image processing section 3 converts the RGB primary color image including the end face images Ex and Ey of the standard heat insulating material Y1 and the test heat insulating material X into a luminance component V. , A color component S, and a hue component H. The defect determination unit 4 pays attention to the blue component image B in the color image of the three primary colors RGB, the converted color component image S and the hue component image H, and generates the edge image Ex in accordance with the reference obtained from the edge image Ey. Judgment is made and the defective portion (discolored portion to white) is detected.

More specifically, the defect judging section 4 converts the luminance component image V of the image E into a threshold Th1 which is temporarily set in advance.
, And the inspection target area in the end face image Ey indicating the standard heat insulating material Y1 is provisionally set in accordance with the binary image. And the average luminance value K1 in this inspection target area
Is calculated, and an optimal binarization threshold Th2 is set based on the average luminance value K1.

After such pre-processing, the luminance component image V of the image E is binarized again using the binarization threshold Th2, and an end face image Ey showing the standard heat insulating material Y1 is obtained from the binarized image. The inspection target area in the inside and the inspection target area in the end face image Ex showing the test heat insulating material X are set.
Thereafter, the average brightness of the inspection target area set as described above in the end face image Ey showing the standard heat insulating material Y1, particularly, for each of the blue component image B, the color component image S, and the hue component image H, is obtained. Values Kb, Ks, Kh
And the variances Db, Ds, and Dh of the luminance are obtained. Then, based on the average luminance values Kb, Ks, Kh and the variance values Db, Ds, Dh, the luminance value K of each part in the inspection target area in the end face image Ex showing the test heat insulating material X is represented by Kb−αb. Db ≦ K ≦ Kb−βb · Db (αb and βb are coefficients) Ks−αs · Ds ≦ K ≦ Ks−βs · Ds (αs and βs are coefficients) Kh−αh · Dh ≦ K ≦ Kh−βh · Dh (Αh, βh are coefficients) It is determined whether or not the following condition is satisfied. Then, the portions where the above conditions are not satisfied are detected as defective portion regions Nb, Ns, Nh. At this time, the defective portion is displayed in white in the end face image Ex or the like, and is distinguished and highlighted from other normal portions.

Further, each area (number of pixels) Ab, As, Ah of the regions Nb, Ns, Nh detected as defective portions in the blue component image B, the color component image S, and the hue component image H.
When the areas Ab, As, and Ah exceed predetermined determination thresholds, these are determined as defective. Furthermore, the defective areas Nb, Ns, Nh are logically processed as, for example, Nbsh = Nbor (Ns and Nh) to obtain the entire defective area Nbsh and the total area Absh. Even when the value exceeds a predetermined determination threshold, it is determined as a defective product. In accordance with the defective product information determined in this way, the disposal of the test heat insulating material X is performed.

On the other hand, the peripheral images Cb and Cr of the blue light component and the red light component of the standard heat insulating material Y2 and the test heat insulating material X obtained as shown in FIGS. The defect detection is performed by performing image processing as described above.
That is, with respect to the peripheral image Cb (Cr), first, the luminance component (monochrome image) is binarized using a threshold Th1b (Th1r) provisionally set in advance, and the standard heat insulating material Y2 is used in accordance with the binary image. Is temporarily set in the inspection target area in the peripheral image Cby (Cry). Then, an average luminance value K1b (K1r) in the inspection target area is calculated, and an optimum binarization threshold Th2 is calculated based on the average luminance value K1b (K1r).
Set b (Th2r).

After such pre-processing, the peripheral image Cb
The luminance component of (Cr) is converted to a newly set binarization threshold Th.
2b (Th2r), the inspection object area in the peripheral image Cby (Cry) showing the standard heat insulating material Y2, and the peripheral image Cbx showing the test insulating material X from the binarized image.
Each inspection target area in (Crx) is set. still,
When setting the inspection target area in the peripheral surface image Cbx (Crx), for example, a notch provided on the peripheral surface of the test heat insulating material X is used as a clue, for example, for one round of the test heat insulating material X. Cut out only the peripheral image.

Thereafter, for the inspection target area set as described above in the peripheral image Cby (Cry) showing the standard heat insulating material Y2, the average luminance value Kby (Kry) and its variance value Dby (Dry) are calculated. Ask for each. Then, the average luminance value Kby (Kry) and the respective variance values Dby (D
ry), the peripheral image Cbx showing the test insulation X
The luminance value K of each part in the inspection target area in (Crx) is Kbx−αbx · Dbx ≦ K ≦ Kbx−βbx · Dbx (αbx and βbx are coefficients) or Krx−αrx · Drx ≦ K ≦ Krx−βrx · Drx ( αrx and βrx are coefficients). Then, a portion where the above condition is not satisfied is detected as a defective portion region Mb (Mr), and the area (number of pixels) Sb (S) of the detected region Mb (Mr) is detected.
r), and if the area Sb (Sr) exceeds a predetermined determination threshold, this is determined as a defective product. On this occasion,
The defective portion is displayed in blue (red) in the peripheral surface image Cbx (Crx) or the like, and is distinguished and highlighted from other normal portions. Then, disposal of the test heat insulating material X determined to be defective is urged.

According to the present apparatus for performing the defect inspection as described above, the white or brown discoloration (color unevenness) defect and the irregularity defect appearing on the surface of the heat insulating material can be detected with high accuracy. . Moreover, the defect inspection can be performed by effectively utilizing the image processing without being affected by the color information of the peripheral surface forming the fiber pattern of the heat insulating material. Further, as described above, the image of the test heat insulating material X is captured simultaneously with the image of the standard heat insulating material Y1 (Y2), and the image of the test heat insulating material X is processed based on the image information of the standard heat insulating material Y1 (Y2). Even if the lighting conditions fluctuate,
The defect inspection can be executed under a certain determination criterion without being affected by the influence, and the inspection accuracy can be improved.
Therefore, the inspection processing efficiency can be improved, and further, the defect inspection can always be stably performed at a constant quality level without being affected by the test environment such as the lighting conditions, and a great effect in practical use. Is played.

Since the end face images Ey and Ex of the standard heat insulating material Y1 and the test heat insulating material X are taken in as one image as described above, for example, the thickness (known) is obtained from the end face image Ey of the standard heat insulating material Y1. As the image width Wy, and this image width Wy
It is also possible to obtain the thickness from the image width Wx in the end face image Ex of the test heat insulating material X on the basis of. Then, verify whether the thickness is within a predetermined allowable error range,
Thereby, the dimensional defect may be inspected.

In this thickness inspection, the luminance component image V in the above-described end face image is used, the component image V is binarized using a threshold value Th1 provisionally set in advance, and the standard thermal insulation is performed according to the binary image. An inspection target area in the end face image Ey showing the material Y1 is provisionally set. Then, an average luminance value K1 in the inspection target area is calculated, and an optimum binarization threshold Th2 is set based on the average luminance value K1. Then, the luminance component image V is binarized again using the newly set binarization threshold Th2, and the standard heat insulating material Y is obtained from the binarized image.
An inspection target area in the end face image Ey indicating 1 and an inspection target area in the end face image Ex indicating the test heat insulating material X are set. Note that these processes can be omitted by using the information obtained by the above-described end face inspection as it is.

Thereafter, the image width (thickness) Wy of the end face image Ey is obtained as the number of pixels, and the correspondence (ratio) to the actual thickness previously obtained is obtained. Then, similarly, the image width (thickness) Wx in the end face image Ex is determined as the number of pixels, and the thickness of the test heat insulating material X is calculated according to the above ratio. Then, it is sufficient to determine whether or not the thickness Wx is within a predetermined allowable error range, and determine whether or not the thickness dimension is good.

After the inspection of the surface defects by the above-mentioned image processing is completed, for example, pressurizing cylinders are placed at three places at both ends and the center of the test heat insulating material X placed on the rollers 1a and 1b. Apply a predetermined pressing force using
The amount of depression (the length pushed by the cylinder) of the peripheral surface of the test heat insulating material X which elastically bends by this is Ll, Lc, Lr.
And each is measured. And these dent amounts Ll, L
From c and Lr, for example, the deviations Δl and Δr are calculated as Δl = Ll−Lc and Δr = L2−Lc, and these deviations Δl and Δr are compared with a predetermined judgment threshold value to check for a deviation in hardness. You may do it. Furthermore, it is also possible to incorporate a weighing scale into the support mechanism 1, measure the weight of the test heat insulating material X placed on the rollers 1a and 1b, and execute the weight inspection.

In the apparatus configured as described above, the end image E of the test heat insulating material (test body) X and the image Cb of the blue light component on the peripheral surface thereof are obtained by using three cameras 2, 5, and 6.
And the image Cr of the red light component is obtained at the same time, but the image Cb of the blue light component and the image Cr of the red light component on the peripheral surface of the test insulation X are sequentially obtained while changing the color of the illumination light as described above. For example, the processing may be performed according to the processing procedure shown in FIG. That is, first, it is determined whether or not the test heat insulating material X has been placed on the rollers 1a and 1b of the support mechanism 1 [Step S1]. When the setting of the test heat insulating material X is confirmed, the motor 2 is rotationally driven to rotate the test heat insulating material X [Step S]
2].

In this state, the end face image E of the heat insulating material X, Y1 is input via the camera 2 [Step S3], and color information is detected from the end face image Ey in the input image [Step S3].
4], the end face image E of the test insulation X based on the color information
The color information of x is determined [Step S5]. Then, as described above, a portion having a different color (color changing portion) is detected as a defective region [Step S6]. Then, the defect detection site is identified and displayed on the end face image Ex [Step S7]. Thereafter, the thickness of the test heat insulating material X is measured from the end face images Ex and Ey as described above [Step S].
8].

Thereafter, first, the peripheral surfaces of the test heat insulating material X and the standard heat insulating material Y2 are irradiated with blue light (blue wavelength component) [Step S11], and the camera 5 extracts the blue light components of the heat insulating materials X and Y2. The peripheral image Cb (Cby, Cbx) is input [Step S12]. Then, after performing preprocessing such as setting the binarization threshold by obtaining the average luminance and the variance value from the peripheral image Cby in the input image, the peripheral image for one rotation from the peripheral image Cbx is obtained. The clipping is performed [Step S13], and the discoloration defect portion is detected by performing the threshold processing as described above [Step S14]. Then, the detected defective portion is identified and displayed [Step S15].

Similarly, the peripheral surfaces of the test insulation X and the standard insulation Y2 are irradiated with red light [Step S2].
1] The camera 6 captures and inputs a peripheral image Cr (Cry, Crx) composed of red light components of the heat insulating materials X and Y2 [Step S22]. Then, after performing a predetermined pre-processing in the same manner as in the case of the blue light component described above, a peripheral image Crx for one round is cut out from the input image [Step S23], and a predetermined threshold value processing or the like is performed to obtain the uneven defect portion. Is detected [Step S24]. Then, the detected defective portion is identified and displayed [Step S25].

When the above series of processing is executed, for example, the motor 2 is stopped [Step S31], and the weight inspection [Step S32] and the hardness inspection [Step S33] are executed. The weight inspection is performed by the support mechanism 1
This is performed by using, for example, a weight sensor incorporated in the device. Inspection of hardness is performed using a pressurized cylinder as described above. It goes without saying that the weight measurement and the hardness measurement are performed independently of the defect inspection by image processing.

Thus, according to the present apparatus configured as described above, a defect occurring in the heat insulating material X can be detected simply and reliably by simple image processing. In particular, it is possible to accurately detect a defect appearing at an end portion of the heat insulating material X formed by winding glass wool as a raw material to a predetermined thickness and a defect occurring on the peripheral surface thereof. Moreover, while simultaneously taking in the information of the standard heat insulating material under the same imaging condition, the defect inspection can be performed by appropriately setting the defect judgment standard based on the color (luminance) information required from the standard heat insulating material. There is an effect that there is no ambiguity as in the inspection and the defect inspection can always be performed at a constant quality evaluation level. As a result, the inspection efficiency can be improved, and a great effect in practical use such as stable quality control can be achieved.

Further, according to the apparatus configured as described above, the image of the end surface and the peripheral surface is taken in while rotating the cylindrical heat insulating material X on the rollers 1a and 1b. Image information over the circumference can be input. In addition, since it is only necessary to scan the upper surface of the heat insulating material rotating on the rollers 1a and 1b in the axial direction and input an image, the entire area of the peripheral surface of the heat insulating material is stably imaged under a constant imaging (illumination) condition. be able to. Furthermore, the image of the test insulation is processed based on the color (luminance) information required from the standard insulation while simultaneously taking in the information of the standard insulation under the same imaging conditions, so that the image is always irrespective of fluctuations in the lighting conditions. The effect is obtained that a stable image can be obtained and used for defect detection.

The present invention is not limited to the above embodiment. For example, the present invention can be similarly performed when inspecting a surface defect of a heat insulating material formed of glass wool into a flat plate shape. However, in this case, the surface of the flat heat insulating material may be two-dimensionally scanned to obtain an image of the entire surface. It is also possible to use an orange transmission filter that can be regarded as substantially red instead of the red transmission filter. In addition, the present invention can be variously modified and implemented without departing from the gist thereof.

[0045]

As described above, according to the present invention, it is possible to easily and accurately detect a defective portion accompanied by discoloration or an uneven defect occurring on the surface of a heat insulating material from an image obtained by imaging the surface of the heat insulating material. Can be detected. In particular, by obtaining an image composed of a blue light component and an image composed of a red light component, respectively, of the surface of the heat insulating material mainly having a yellow fiber pattern, the discoloration-enhanced image and the brightness difference due to unevenness are emphasized. The obtained images are obtained, and the discoloration defect and the unevenness defect are respectively detected from these images, so that the effect of effectively increasing the detection accuracy can be obtained.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram of a defect inspection apparatus for a heat insulating material according to an embodiment of the present invention.

FIG. 2 is a view schematically showing an example of an end face image and a peripheral face image of a heat insulating material obtained by the defect inspection apparatus shown in FIG.

FIG. 3 is a diagram illustrating an example of a schematic inspection processing procedure in the example apparatus.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Support mechanism 2 Color camera (edge image capturing means) 3 Edge image processing part 4 Defect determination part 5, 6 Monochrome camera (surround image capturing means) 7a, 7b Illumination light source 8 Perimeter image processing part 9 Perimeter image Processing unit

Continuation of the front page (72) Inventor Yasuo Katogi 2-2-1 Toranomon, Minato-ku, Tokyo Japan Tobacco Inc. Engineering Department F-term (reference) 2G020 AA08 DA02 DA03 DA04 DA05 DA13 DA34 DA35 DA52 2G051 AA90 AB06 AB07 AB20 BA04 CA04 CB01 DA05 EA11 EB01 EC03 FA10

Claims (5)

[Claims] 1. A supporting mechanism for supporting a heat insulating material formed by molding glass wool into a plate shape or a cylindrical shape by using an adhesive along with a standard heat insulating material used as a reference for defect inspection, and supported by the supporting mechanism. The test insulation material and the cut end surfaces of the standard insulation material are captured simultaneously in the same field of view and imaged simultaneously, and the color information of the test insulation material is determined based on the color information of the standard insulation material in the captured image. An end face defect detecting means for detecting a defect of a cut end face in the test heat insulating material; and taking an image of a blue light component on the surface of the test heat insulating material, discriminating the image of the blue light component by a predetermined threshold, and performing the test. A discoloration defect detecting means for detecting discoloration defects on the surface of the heat insulating material, an image of a red light component on the surface of the test heat insulating material is picked up, and the image of the red light component is discriminated by a predetermined threshold value; On the surface of A defect inspection device for a heat insulating material, comprising: a concave / convex defect detecting means for detecting a concave / convex defect in the material. 2. The discoloration defect detecting means obtains an image composed of a blue light component on the surface of the test heat insulating material via a blue transmission filter, and the unevenness defect detection means obtains the test heat insulating material via a red transmission filter. The defect inspection apparatus for a heat insulating material according to claim 1, wherein an image composed of a red light component on the surface of the material is obtained. 3. The discoloration defect detecting means images a surface of the test heat insulating material illuminated by a blue light source to obtain an image composed of a blue light component. The thermal insulation material defect inspection apparatus according to claim 1, wherein an image composed of a red light component is obtained by imaging the surface of the test thermal insulation material thus obtained. 4. The heat insulating material according to claim 1, wherein the support mechanism includes a rotating member that supports a peripheral surface of the heat insulating material formed into a cylindrical shape and rotates the heat insulating material. Defect inspection equipment. 5. The defect of the heat insulating material according to claim 1, wherein each of the defect detecting means includes means for identifying and highlighting a portion of the test heat insulating material where a defect is detected in the image. Inspection equipment.