JP3427248B2 - Foreign object detection method - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a foreign matter detecting method.

[0002]

2. Description of the Related Art A conventional object detecting method is disclosed in Japanese Patent Laid-Open No. 2-10.
There is a method described in Japanese Patent No. 204. In this method, as shown in FIG. 14, the two illumination devices 2 and 2 installed with the inspection object 1 sandwiched therebetween emit light to generate shadows A and B on the inspection object 1 and the image pickup camera 3 and the image are generated. A luminance histogram of shadows A and B is obtained by the recognition device 4. Then, the presence or absence of the inspection object 1 is accurately detected by comparing the brightness histograms when the two illumination devices 2 and 2 illuminate.

[0003]

In the above object detecting method, when the flat plate portion 6 on which the inspection object 1 is placed has an uneven shape, the uneven shape is generated even though the inspection object 1 does not exist. There is a risk that light and shade will occur and this light and shade will be erroneously detected as the inspection object 1. In other words, when foreign matter detection is performed using such a conventional object detection method, it is not possible to distinguish between the irregular shape existing as a normal shape on the normal surface and the irregular shape caused by the foreign matter, and all are erroneously detected as foreign matter. Therefore, it cannot be used as a foreign matter detection method.

The present invention has been made in view of the above points, and an object of the present invention is to prevent an irregular shape existing as a normal shape on a normal surface from being erroneously detected as a foreign matter. An object of the present invention is to provide a foreign matter detection method capable of detecting only a foreign matter with high accuracy. The object of the invention of claim 2 is that, in addition to the object of the invention of claim 1 , the product may be accumulated when the thickness of the object to be inspected is not constant or when foreign matter is mixed.
Even when the mounting height changes, an appropriate offset can be calculated using the formula.
An object of the present invention is to provide a foreign matter detection method capable of calculating the distance to be moved.

[0005]

[0006]

[0007]

[0008]

[0009]

In order to achieve the above object, in the invention of claim 1, an imaging camera with a focus lens is arranged so as to face the upper surface of the object to be inspected, and
The light source is installed diagonally above the object to be inspected, which has an uneven surface, and the imaging camera is focused on a point offset by a certain distance to the upper surface of the object to be inspected. In contrast, oblique illumination is performed, the upper surface of the object to be inspected is imaged by the imaging camera, a grayscale image is obtained from the captured image, the differential value of the grayscale image is obtained, and the obtained differential value and a preset threshold It is characterized in that the foreign matter which is not the above-mentioned uneven shape is detected by comparing with the value.

According to a second aspect of the invention, in the first aspect of the invention, one or a plurality of objects to be inspected are stacked, and an image pickup camera and
When inspecting the object to be inspected, the loading height is
If different for each inspection object, the top surface of the inspection object and the imaging
Image taken after imaging to determine the distance to the camera
Extends from the outer edge of the inspected object from the obtained grayscale image
Find the length of the shadow, and calculate the shadow length and the angle of oblique lighting.
After calculating the height of the inspection object, the height, the distance, and the
The focal length and effective diameter of the focus lens are covered as a function.
Determine the distance to offset the upper side of the inspection object
Characterized in that it constant.

[0011]

[0012]

[0013]

[0014]

[0015]

BEST MODE FOR CARRYING OUT THE INVENTION The method of the present invention will be described below with reference to embodiments. (Embodiment 1) FIG. 1 shows a schematic configuration diagram of an inspection apparatus using this embodiment. For example, an inspection object (not shown) having an inspection object 1 having an uneven shape such as a groove pattern on the upper surface like a roof tile is shown. No)
The image pickup camera 7 is composed of a CCD camera and is mounted on the upper surface of the object to be inspected 1 so as to be inspected for foreign matter. A strobe light source 8 arranged above and performing oblique illumination on the upper surface of the inspection object 1, a timing circuit 9 for driving the strobe light source 8, and an image recognition unit 10 for processing an image obtained by the image pickup camera 7. , The lens unit 7 of the imaging camera 7
It includes a focus mechanism provided in a, a lens control circuit 11 that controls an electric drive unit of the zoom mechanism, and a control unit 12 that integrally controls the image recognition unit 10, the lens control circuit 11, and the timing circuit 9. The image pickup camera 7 is provided with a drive unit (not shown) that is movable in the X and Y directions, and the position can be moved under the control of the control unit 12.

Thus, in this embodiment, in order to eliminate the influence of the uneven shape, the offset amount (a) for focusing the lens portion 7a of the image pickup camera 7 to a point above the upper surface of the inspection object 1. Set this offset amount (a)
The control unit 12 controls the lens unit 7a through the lens control circuit 11 so that the position determined by is focused. Next, the timing circuit 9 is controlled to perform oblique illumination by the strobe light source 8, and the image pickup camera 7 captures an image of the entire upper surface of the inspection object 1 under the oblique illumination.

Here, since the focus is on the position determined by the offset amount (a), the groove pattern which is a normal uneven shape is out of focus, so that the image is out of focus, and conversely the foreign matter X is Foreign object X if present
The focus is on and the image is clearly captured. This image signal is taken into the image recognition unit 10. Image recognition unit 10
Converts the captured image into a grayscale image, sets a predetermined inspection area (window) for foreign matter detection in this grayscale image, scans the inside of the inspection area to obtain a differential value, and obtains the obtained differential value in advance. A predetermined threshold value is compared, and a portion whose differential value exceeds the threshold value is detected as a foreign substance.

As described above, according to the present embodiment, even if the upper surface of the inspection object 1 has an irregular shape such as a normal groove pattern,
Only the foreign matter X can be detected with high accuracy.
Next, an example for setting the offset amount a in the present embodiment will be described below. First, while the inspection object 1 is loaded on the inspection table, the foreign matter of the inspection object 1 facing the imaging camera 7 is detected. However, when the inspection object 1 has a constant thickness (T1), The following control is performed.

That is, the lens unit 7a of the image pickup camera 7 is controlled through the lens control circuit 11 to focus on the upper surface of the object 1 to be inspected as shown in FIG. 2, and from this point of time, a function of f (T1) is used. The focus is shifted upward by the calculated offset amount a, and the image capturing for detecting the foreign matter is performed. Also when the stacking is repeated and the distance L between the imaging camera 7 and the inspection object 1 is changed, a = f (T1)
The offset amount is determined by.

In the example of FIG . 2, the object to be inspected 1 has a fixed camera position .
Is always constant, but when the camera position is fixed and the thickness T1 is not constant, the distance L between the inspection object 1 and the imaging camera 7 changes, so as shown in FIG. The offset amount a is not constant only by moving the focus upward.

Therefore, as shown in FIG. 3, the lens unit 7a of the image pickup camera 7 is controlled through the lens control circuit 11 under the control of the control unit 12 before the image pickup for foreign matter detection is performed, as shown in FIG. As shown in FIG. 3, the upper surface of the inspection object 1 is imaged by the imaging camera 7 in a state where the surface of the inspection object 1 is focused and the strobe light source 8 obliquely illuminates, and the image recognition unit 10 obtains a grayscale image from the captured image. The distance L between the upper surface of the inspection object 1 and the imaging camera 7 is obtained by calculating the grayscale image. After that, the focus is moved upward from the distance L by a predetermined offset amount a.

[0022]

Here, a method for obtaining the thickness T1 of the inspection object 1
As an example, as shown in FIG.
As shown in (a), the strobe light source 8 obliquely illuminates, and the length s of the shadow of the inspection object 1 obtained by the illumination is obtained from a grayscale image obtained, and this length and the tilt angle of the strobe light source 8 are obtained. From T, it is calculated by the equation T1 = s · tan θ. this
Based on the obtained thickness T1 and the fixed position of the imaging camera
And the distance L from the image pickup camera 1 to the inspection object 1 is obtained,
The focus is moved upward from this distance L by a predetermined offset amount a.

As a result, it is possible to secure a constant offset amount a even with the thin inspection object 1 shown in FIG. 3 (b) as compared with the inspection object 1 in FIG. 3 (a). According to the method shown in FIG. 3, even when the thickness T1 of the inspection object 1 changes as shown in FIG. for,
It is possible to capture an image that is out of focus with the normal uneven shape (groove pattern). By the way, the thickness (height) T1 of the object to be detected 1 and the object to be inspected 1 and the image pickup camera 7 are also measured by the above-described method not only when the thickness of the object to be inspected 1 changes but also when the stacking height is not constant due to foreign matter. Is calculated, and the distance L between the imaging camera 7 and the inspection object 1 is calculated.
Using the thickness T1, the focal length f of the lens portion 7a and the effective diameter A as a function f (T1, L, f, A) , the offset amount a is obtained as described above , and only the obtained offset amount a is obtained. The focus may be moved upward.

According to the above method, even when the height of the object 1 to be inspected is not constant or when the stacking height changes due to the inclusion of foreign matter, the appropriate offset amount a
Can be asked. On the other hand, when the thickness of the foreign matter X of the inspected object 1 is limited, in order to identify the shape such as the size, outer shape, and height of the foreign matter X and the state of the material, the type, etc. after the foreign matter is detected, After the foreign substance X is detected on the surface 1, the focus is set on the upper surface of the foreign substance X from the state where the focus is set corresponding to the offset amount a of the imaging camera 7 as shown in FIG. In this case, the focus setting distance La is obtained from the offset amount a, the thickness of the foreign matter X, and the distance L to the inspection object 1.

By focusing on the foreign matter X and picking up an image, the size and shape of the foreign matter X can be recognized and the surface condition of the foreign matter X can be grasped, and analysis of the cause of the foreign matter X can be performed. If the foreign matter X is magnified and imaged, it is possible to accurately identify the shape of the foreign matter X, the shape such as the outer shape, the height, and the state of the material type. Therefore, based on the detection information of the post-foreign matter detection image recognition unit 10, the imaging camera 7 is moved from the position (center of the imaging screen) of FIG. 6A to the position of FIG. 6B under the control of the control unit 12. The foreign matter X is moved to above the foreign matter X and the zoom mechanism of the lens portion 7a is operated to move the foreign matter X.
It is better to take an enlarged image of.

The foreign substance inspection area (window) is set in the image picked up by the image pickup camera 7 in order to detect the foreign substance X described above. However, in the case of sequentially inspecting the loaded inspected objects 1, depending on the loading position. When the distance between the imaging camera 7 and the inspection object 1 changes and the inspection object 1 is imaged at the same angle of view, the size of the image captured by the imaging camera 7 is the inspection object 1
It will change depending on the position of.

Therefore, a region is set in advance, and by this region setting, a foreign substance inspection region for the object 1 to be inspected which is actually the inspection object is secured.
Even if the size of the image of the object to be inspected 1 changes depending on the loading position, the foreign matter inspection area can always cover the entire object to be inspected 1. That is, as shown in FIG. 7, in setting the area, the image of the inspection object 1 at the loading position closest to the imaging camera 7 is captured,
The outer edge of the density image obtained from this image is detected by the differential processing to calculate the X and Y coordinates for determining the inspection area. Here, the inspection area is divided into four areas so that the largest image is divided into four.
Areas W _{1 to} W ₄ are set, and (X1 , Y1 , X2 , Y2 ) Of the coordinates of X1
, Y1 The inspection areas W ₁ and W ₂ are determined by
Two , Y2 The inspection areas W ₃ and W ₄ are set by.

After this setting, in the actual inspection, the image obtained by picking up the inspected object 1 is converted into a grayscale image by the image recognition section 10, and the grayscale image is differentiated to differentiate the grayscale image. Coordinates of the outer edge of (X1 , Y1 , X2
, Y2 ) Is calculated, and the inspection areas W _{1 to} W ₄ set as described above are set in the image of the object 1 to be inspected in such a manner that the outer two sides are arranged along the outer edge based on the calculated coordinates.

The coordinate calculation method of the outer edge will be described in detail. First, in the image captured as shown in FIG. 8A, for example, the upper left coordinate is (0, 0) and the lower right coordinate is (255, 232). ), The coordinates of the outer edge can be obtained within this range. The inspection area for recognizing the coordinates of the outer edge is detected as (X1, X2) and (X2, Y2) by providing an edge detection area for X coordinate detection and Y coordinate detection by absolute position designation. and, from the detected coordinates, can determine the coordinate points a and the upper left coordinate point B at the lower right, set the inspection area W ₁ to _W-4 of the object 1 as illustrated by this decision, as described above It is possible.

In this case, the inspection areas W _{1 to} W ₄ are overlapped as shown in FIG. 8A or 8B depending on the size of the picked-up image, and the size of the overlapped portion changes, The inspection area for detecting a foreign substance can cover the entire upper surface of the inspection object 1 even if the size of the image of the inspection object 1 becomes small as shown in FIG. 8B. By the way, since the inspection object 1 is loaded, the loading may deviate. Therefore, in order to detect the stacking deviation, in the present embodiment, the edge detection area of the inspected object 1 at this time is set based on the edge detection position set for the image of the immediately preceding inspected object 1, The stack deviation is detected by determining whether the edge can be detected.

That is, as shown in FIG. 9, an image of the inspection object 1 is picked up.
And the outer shape of the image of another object under inspection 1
The edge detection position is shown as reference coordinates, for example, in FIG.
X1 Edge inspection area with a width of ± M pixels based on
Set Wa to the current image as shown in Fig. 10 (b).
Set. Differentiate the grayscale image in the edge inspection area Wa.
It is determined whether or not the outer shape edge can be detected. Figure 1
0 (b) this time X2 The outer edge coordinates are
Since it exists in the inspection area Wa, the
It is not impossible to get out, and the edge inspection area W of the next inspection object 1
X2 as the reference coordinate for setting a Register
After the end of the object detection process, the inspection of all inspected objects 1 is completed.
If not completed, the above processing is performed again. Also above
When the external edge cannot be detected in the edge inspection area Wa
If it is determined that the stacking deviation amount of the inspection object 1 is M pixels or more,
It

In the case of FIGS. 9 and 10, the edge detection area Wa is set with reference to the edge detection coordinates of the immediately preceding inspection object 1 as the reference. When the area Wa is fixed and the object to be inspected 1 is imaged as shown in FIG. 12A, the edge detection area Wa is set to perform edge detection. When the edge detection is successful, the next object to be inspected is detected. The position coordinates Xn, Yn of the outer edge are registered to detect the foreign matter of the object 1, and if the edge cannot be detected, the position coordinates Xn-1 of the outer edge registered at the time of the inspection of the previous object 1 to be inspected, Based on Yn-1
As shown in FIG. 12B, the foreign matter detection is performed using the inspection area for foreign matter detection, and the position coordinate Xn− of the outer edge is detected for the next foreign matter detection of the inspection object 1.
Register 1, Yn-1.

By using such a method, even when the edge detection for determining the position of the inspection object 1 in the inspection image cannot be performed due to the presence of the foreign matter X in the edge detection area Wa, the previous inspection object is detected. The foreign substance can be detected by setting the inspection area for the foreign substance inspection using the edge detection position data of the image No. 1. By the way, in the above method, the outer edge detection position and the position coordinates after the second time are used, but for the first time, the edge detection area is set at the absolute position and the processing is performed.

( Reference Example ) In the case of the above-described first embodiment, the surface of the inspection object 1 is covered with water,
When there is a glossy surface due to oil, it may not be possible to distinguish it from the surface on which the foreign matter X is present, and it may not be possible to detect only the foreign matter. Wherein the light source (or of course strobe light source) which directly faces continuous emission to the upper surface of the specimen 1 as shown in FIG. 13 13 set up, and to extract only the foreign substance X in the following manner.

Next, this reference example will be described. First, ginseng
In the consideration example , the control unit 12 is set before the image capturing for detecting the foreign matter is performed.
Under the control of the above, the lens portion 7a of the image pickup camera 7 is controlled through the timing circuit 9 so that the focus is adjusted to the surface of the object to be inspected 1 and the upper surface of the object to be inspected 1 is taken up by the image pickup camera 7 while being obliquely illuminated by the strobe light source 8. An image is picked up, and the image recognition unit 10 obtains a grayscale image from the picked-up image and arithmetically processes the grayscale image to obtain a distance L between the upper surface of the inspection object 1 and the image pickup camera 7. Thereafter, the focus is moved upward from the distance La by a predetermined offset amount a.

After that, the timing circuit 9 is controlled to perform oblique light illumination by the strobe light source 8, and the image pickup camera 7 captures an image of the entire upper surface of the inspection object 1 under the oblique light illumination. In this case, the brightness of the surface when the foreign matter X exists on the upper surface of the inspection object 1 becomes dark. Further, when the glossy surface Y exists as shown in FIG. 13, the lightness also becomes dark. Therefore, when foreign matter is detected using this image, it cannot be distinguished from the glossy surface Y.

Then, the illumination of the strobe light source 8 is stopped, the illumination of the light source 13 is performed, and the image pickup by the image pickup camera 7 is performed. In this case, the brightness of the surface on which the foreign matter X is present does not change,
The brightness of the glossy surface Y becomes brighter. Therefore, the surface on which the foreign matter X is present and the glossy surface Y are identified from the image data obtained by previously capturing the image under the oblique illumination and the image data obtained by capturing the image under the facing illumination this time. As a result, when comparing the differential value of the grayscale image and the threshold value, the foreign material X can be detected with high accuracy without erroneously detecting the glossy surface Y as the foreign material X.

Needless to say, the method of determining the offset amount a, the method of recognizing enlargement of foreign matter, the method of setting the detection area, the method of setting the edge detection area, and the like used in the first embodiment can also be used in this reference example .

[0040]

According to the first aspect of the present invention, the imaging camera with the focus lens is arranged so as to face the upper surface of the object to be inspected, and the light source is installed obliquely above the object to be inspected having the uneven shape on the upper surface. Then, the imaging camera is focused on a point offset by a certain distance upward with respect to the upper surface of the inspection object, and the light source illuminates the inspection object obliquely, and the imaging camera images the upper surface of the inspection object. , To obtain a grayscale image from the captured image to obtain a differential value of the grayscale image, and to detect the foreign matter that is not the uneven shape by comparing the obtained differential value and a preset threshold value, Even if an uneven surface such as a pattern is present on the normal surface of the object to be inspected, it is possible to detect only the foreign material with high accuracy without being erroneously detected as the foreign material.

[0041]

[0042]

[0043]

[0044]

According to a second aspect of the present invention, in the first aspect of the present invention, one or a plurality of inspection objects are loaded, and when the inspection objects facing the imaging camera are inspected, the loading heights of the inspection objects are different. If the difference depends on the object, after imaging to determine the distance between the upper surface of the inspection object and the imaging camera, the length of the shadow extending from the outer edge of the inspection object is determined from the grayscale image obtained from the captured image, and the shadow is calculated. After the height of the object to be inspected is calculated from the length of the object and the angle of oblique illumination, the height, the distance, and the focal length and effective diameter of the focus lens are functions as functions above the upper surface of the object to be inspected. Since the offset distance is determined, it is possible to calculate an appropriate offset distance by a calculation formula even when the thickness of the object to be inspected is not constant or when the stacking height changes due to the inclusion of foreign matter. effective.

[0046]

[0047]

[Brief description of drawings]

FIG. 1 is a schematic configuration diagram of an apparatus using a first embodiment of the present invention.

FIG. 2 is an explanatory diagram of an offset amount determination method of the above.

FIG. 3 is an explanatory diagram of a focus distance setting method of the above.

FIG. 4 is an explanatory diagram of another offset amount determination method of the above.

FIG. 5 is an explanatory diagram of focus control for recognizing a foreign matter in the above.

FIG. 6 is an explanatory diagram of the foreign matter enlargement recognition method of the above.

FIG. 7 is a flowchart of the inspection area setting method of the above.

FIG. 8 is an explanatory diagram of an inspection area setting method of the above.

FIG. 9 is a flowchart of the edge inspection area setting method of the above.

FIG. 10 is an explanatory diagram of the edge inspection area setting method of the above.

FIG. 11 is a flowchart of the foreign matter inspection area setting method of the above.

FIG. 12 is an explanatory diagram of a foreign matter inspection area setting method of the above.

FIG. 13 is a schematic configuration diagram of an apparatus using a reference example .

FIG. 14 is a schematic configuration diagram of an apparatus using a conventional method.

[Explanation of symbols]

1 Inspected 7 Imaging camera 7a lens part 8 Strobe light source 9 Timing circuit 10 Image recognition unit 11 Lens control circuit 12 Control unit X foreign body

─────────────────────────────────────────────────── ─── Continuation of the front page (56) Reference JP-A-8-240535 (JP, A) JP-A-3-261807 (JP, A) JP-A-6-62407 (JP, A) JP-A-2- 10204 (JP, A) (58) Fields surveyed (Int.Cl. ⁷ , DB name) G01N 21/84-21/94 G01B 11/30

Claims (2)

(57) [Claims] 1. An upper surface of an object to be inspected, wherein an imaging camera with a focus lens is arranged so as to face the upper surface of the object to be inspected, and a light source is installed obliquely above the object to be inspected having an uneven shape. The image pickup camera is focused on a point offset by a certain distance to the upper side, and the light source illuminates the object to be inspected obliquely, and the image pickup camera images the upper surface of the object to be inspected. A foreign matter detecting method, characterized in that a grayscale image is obtained, a differential value of the grayscale image is obtained, and the obtained differential value is compared with a preset threshold value to detect the foreign matter having no uneven shape. 2. One or a plurality of objects to be inspected are loaded and an imaging camera is mounted.
Loading when inspecting the object to be inspected facing the camera
If the height is different for each inspected object,
After imaging to find the distance to the imaging camera,
From the grayscale image obtained from the image
Find the length of the shadow that extends and the angle of the oblique illumination with the length of the shadow
After calculating the height of the inspected object from the
And the focal length and effective diameter of the focus lens as a function
The distance to offset the upper surface of the object to be inspected.
Foreign matter detection <br/> how according to claim 1, wherein the determining the release.