JPH11160046A - Appearance inspection method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an appearance inspection method capable of detecting a ruggedness defect present on the surface for an object to be inspected whose surface is in a dome shape. SOLUTION: The surface of the object 1 to be inspected in the dome shape is linearly irradiated with a laser beam, the feature of a ruggedness defect part appearing on a feature line obtained by projecting the reflected light on a screen 3 is recognized and the normal/defective conditions of the object 1 to be inspected are judged. Images including the feature line projected on the screen 3 are picked up, the feature line is extracted as a curve in an almost arch shape by an image processing, an inspection line is prepared by slightly moving it and a defect candidate part is detected by obtaining the intersection of the inspection line and the original feature line.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a visual inspection method for detecting an uneven defect such as a dent or a projection on a surface of an inspection object having a dome shape.

[0002]

2. Description of the Related Art Conventionally, as a method for extracting defects by visual inspection, as disclosed in Japanese Unexamined Patent Publication No. Hei 3-175343, a predetermined inspection area is added to an image obtained by imaging the surface of an inspection object. Is known, a raster scan is performed in the area, a differential image, an edge image, and the like are obtained according to a predetermined algorithm, and a defect on the surface of the inspection object is extracted.

[0003]

In the above-mentioned prior art, when the surface shape of the inspection object is flat, the setting of the inspection optical system can be easily and surely inspected, but the surface of the inspection object has a dome shape. Therefore, when trying to recognize a concave / convex defect existing on the surface, setting of the inspection optical system is difficult, and a reliable inspection cannot be performed.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a visual inspection that can detect an uneven defect existing on the surface of an inspection object having a dome shape. It is to provide a method.

[0005]

According to the appearance inspection method of the present invention, in order to solve the above-mentioned problem, as shown in FIG. , And the quality of the object 1 to be inspected is determined by recognizing the features of the uneven defect portion appearing on the feature line L obtained by projecting the reflected light on the screen 3.

In order to detect a concave / convex defect portion from a characteristic line, for example, an image including a characteristic line projected on a screen is captured, and the characteristic line is extracted as a substantially arched curve by image processing. An inspection line having substantially the same shape is created by slightly shifting the line, and the intersection of the inspection line and the original feature line is determined to detect a defect candidate portion. Then, an evaluation window is set for the defect candidate portion, and a density difference between windows, a differential absolute value, and the like are comprehensively evaluated to determine whether the defect is a defect.

Alternatively, a differential direction value of the extracted characteristic line is obtained, and a change pattern of the differential direction value when the object to be inspected is a non-defective item is stored in advance as a non-defective item, and the inspected object to be measured is stored. The pattern of change in the differential direction value is compared with a non-defective pattern, and a difference is detected to detect an irregularity defect. Alternatively, the unevenness defect is detected by comparing the magnitude of the differential direction value at a position separated by a predetermined pixel on the change pattern. At this time, the entire length of the characteristic line may be used as a reference for judging unevenness defects.
Further, it is preferable to remove disturbance of the change pattern of about 1 to 2 pixels as noise.

In addition, the defect portion is emphasized by superimposing the feature lines extracted for each imaging screen while rotating the object to be inspected, and the unevenness defect is detected based on the multiple feature lines obtained by the superimposition. It is also very effective.

[0009]

(Embodiment 1) FIG. 1 is a perspective view showing the appearance of an inspection optical system for carrying out the appearance inspection method of the present invention. In the figure, reference numeral 1 denotes a dome-shaped inspection object, such as a fire detector. Reference numeral 2 denotes a line light source that irradiates the surface of the inspection object with line light using, for example, a semiconductor laser. Reference numeral 3 denotes a screen for projecting light reflected from the surface of the inspection object. Reference numeral 4 denotes a television camera that captures a projection image L on a screen, and includes, for example, CC
It is composed of a D camera. In this inspection optical system, the inspection object 1 can be rotated in the direction of the arrow, and imaging can be performed a plurality of times while rotating the inspection object 1. The rotation axis of the inspection object 1 is substantially coincident with the center axis of the dome shape, and the position of the projected image hardly changes even if the inspection object 1 rotates. Although the screen 3 is translucent and shoots transmitted light, the screen 3 may be replaced with an opaque screen to shoot reflected light. Alternatively, the line light source may be replaced with a beam light source, and the light beam may be scanned linearly during the charge accumulation time of the CCD camera, thereby equivalently forming a line light source.

FIG. 2 is a block diagram of the image processing system. The screen projection image captured by the television camera 4 is A /
The image is converted into a digital grayscale image by the D converter 5 and taken into the image memory 7 as an original image. The grayscale image taken into the image memory 7 is subjected to predetermined image processing by the preprocessing unit 6, converted into a differential image, a differential direction value image, and an edge image described later, and stored in the image memory 7 again. The images after these pre-processing are evaluated by the determination unit 8,
The presence or absence of the irregularity defect is determined. The system configuration of the inspection optical system shown in FIG. 1 and the image processing system shown in FIG. 2 is hereinafter referred to as “system configuration 1”.

Here, the image processing by the pre-processing unit 6 is disclosed in detail in Japanese Patent Laid-Open Publication No. 3-175343 of the Company, and the main points thereof will be described again as follows. First, the original image obtained by imaging the screen is, for example, a grayscale image as shown in FIG. 3, and the process of extracting the laser line light characteristic image from this grayscale image is as follows.
Based on the concept of detecting a portion where a change in density is large, that is, an edge, by differentiating the density. The differentiation process is performed by dividing the original image into local parallel windows of 3 × 3 pixels as shown in FIG. That is, a pixel E of interest and eight pixels (near eight) A around the pixel E
To D and F to I, a local parallel window is formed, and a vertical density change ΔV and a horizontal density change ΔH of the density of the pixels A to I in the local parallel window are obtained by the following equation: ΔV = ( A + B + C) − (G + H + I) ΔH = (A + D + G) − (C + F + I) Further, the differential absolute value | e | and the differential direction value ∠e are obtained by the following equations.

| E | = √ (ΔV ² + ΔH ² ) ∠e = tan ⁻¹ (ΔV / ΔH) + π / 2 where A to I indicate the density of the corresponding pixel. As is apparent from the above equation, the differential absolute value | e | represents the rate of change in density in the area near the pixel of interest in the original image,
The differential direction value Δe represents a direction orthogonal to the direction of the density change in the vicinity area.

By performing the above operation for all pixels of the original image, it is possible to extract a portion where the density change is large with respect to the background such as a laser line light characteristic image and the direction of the change. Here, the density of each pixel is
An image expressed by the differential absolute value | e | is called a differential absolute value image, and an image expressed by the differential direction value ∠e is called a differential direction value image.

Next, a thinning process is performed. The thinning process is performed while paying attention to the fact that the larger the absolute differential value, the larger the density change. That is, the differential absolute value of each pixel is compared with the differential absolute values of the surrounding pixels, and those that are larger than the surrounding pixels are connected, whereby 1
The edge having the width of the pixel is extracted.

As shown in FIG. 5, the position of each pixel is represented by XY.
Considering a differential absolute value pixel represented by coordinates and taking the differential absolute value on the Z axis, thinning corresponds to finding a ridge line on this curved surface. By the processing so far, all edges are extracted regardless of the magnitude of the differential absolute value.
Since the ridge lines obtained at this stage include unnecessary small peaks due to noise or the like, as shown in FIG. 6, a threshold value is appropriately set, and only the values above this threshold value are adopted. To remove noise components. The image obtained by this processing tends to be a discontinuous line when the contrast of the original image is insufficient or when there is much noise. Therefore, edge extension processing is performed. In the edge extension processing, starting from the end point of the discontinuous line, the pixel of interest is compared with its surrounding pixels, the line is extended in a direction in which the evaluation function represented by Keep doing this until you crash.

F (e _j ) = | e _j | · cos (∠e _j −∠e ₀ ) · cos ｛(j−1) π / 4−∠e ₀ ) where e ₀ is a central pixel (local a differential data corresponding to) the E parallel window, e _j is the differential data of the adjacent pixels (corresponding to 8 near the local parallel window), j = 1,2, ..., 8.

By the above processing, as shown in FIG. 6, an edge image which traces a portion where the density change is large in the original image is obtained. The edge image can be regarded as representing a laser line light characteristic image. By the pre-processing described above, the original image, the differential absolute value image, the differential direction value image, and the edge image
There are obtained various types of images, and each image is stored in the original image memory 7.
1, stored in a differential absolute value image memory 72, a differential direction value image memory 73, and an edge image memory 74. In the following description, it is assumed that the positions of the pixels in each image are represented by XY coordinates, and the density of the pixels in each image is f ₁ (x,
y), f ₂ (x, y), f ₃ (x, y), f ₄ (x, y
y).

The original image is a grayscale image, and the density is usually represented by 8 bits. Therefore, the density a (= f
₁ (x, y)) is 0 ≦ a ≦ 255. Further, the density of the differential absolute value pixel (that is, the differential absolute value) b (= f ₂
(X, y)) is represented by, for example, 6 bits, and 0 ≦ b ≦ 6.
The density of the differential direction value image (that is, the differential direction value) c (= f ₃ (x, y)) is represented, for example, in 16 directions, and 0 ≦ c ≦ 15. For an edge image, only the presence or absence of a line is represented, so that a pixel serving as a line is represented as "1" and the other pixels are represented as "0". That is, f ₄ (x,
The value range of y) is {0, 1}. In the following description, the term "density" indicates white density, and the higher the density value, the brighter the density.

Next, in the judgment section 8, as shown in FIG.
The inspection area is appropriately set for the edge image. When the inspection region is set, as indicated by the arrows in FIG. 3, by raster scanning the inspection area within the edge image, f ₄
Pixels satisfying (x, y) = 1 are detected and set as flag points. Starting from this flag point, f ₄ (x,
The contour line is traced while extracting pixels for which y) = 1. By tracking the contour in this way, the coordinates of all the pixels on the contour can be known, and the maximum and minimum values of the X coordinate and the Y coordinate are obtained, respectively. As a method of obtaining the maximum and minimum values of the X coordinate and the Y coordinate, a method of tracking the contour and storing all the coordinates in a buffer,
There is a method in which the magnitude relationship between the coordinates is determined while the contour is being tracked without storing all the coordinates. As described above, the maximum values Xmax, Xmax,
If Ymax and the minimum values Xmin and Ymin can be obtained, a rectangular defect candidate area circumscribing the laser line light characteristic image can be set as shown in FIG. After the defect candidate area is set in this way, a middle point o of the defect candidate area is set as shown in FIG. The middle point o is set on the concave side of the arched laser line light characteristic image.

The above image processing procedure is described in Japanese Unexamined Patent Publication No.
It is disclosed in detail in 175343. Using this image processing procedure, image processing such as differential absolute value, differential direction value, edge extraction, and edge extension is performed to obtain an image of a feature line. Then, a non-defective item is determined by detecting the irregularity defect characteristic portion generated on the characteristic line.

The projected image of the laser line light applied to the dome-shaped inspection object has the shape shown in FIG. 7 for a non-defective product, and FIGS.
The shape is as illustrated in FIG. Good quality reflection,
Although the projection pattern has a smooth arch shape (bow shape), the curvature changes abruptly in a part of the projection pattern because the reflection angle changes at the concave / convex defect portion. 8A to 8C show disturbances of the projection pattern corresponding to the concave / convex defects.

FIG. 9 shows the principle of forming an image as shown in FIG. 8 when an unevenness defect exists on the surface of the inspection object. Optical paths (2) and (3) in the figure correspond to FIGS. 8A and 8B, respectively. When there is no defect on the surface of the inspection object (FIG. 3
(In the case of the surface shape shown by the broken line in FIG. 7), the light passes through the optical path (1), so that the projection pattern has an arch shape as shown in FIG.

FIG. 10 is a flowchart showing a processing procedure of the appearance inspection method according to the first embodiment of the present invention. In this embodiment, while rotating the inspection object 180 degrees in the direction of the arrow shown in FIG.
The laser line light is continuously irradiated, and the projected pattern of the reflected light is imaged. The captured image is input to an image processing apparatus, and a differential absolute value, a differential direction value, and the like are obtained, and a process such as edge extension is performed to extract a feature. The image is enlarged or reduced based on the obtained characteristic line image, and an inspection line having substantially the same shape as the characteristic line image is created around the characteristic line image. By searching for the created inspection line or feature line, a change point such as a density value, a differential absolute value, or a differential direction value is obtained, and unevenness defects on the feature line are recognized based on the surrounding image data. Things. A more specific method of detecting the irregularity defect will be described in Examples 2 to 13 below.

(Embodiment 2) FIG. 11 is a flowchart showing a processing procedure of an appearance inspection method according to Embodiment 2 of the present invention. In this embodiment, in the optical system shown in the system configuration 1 described above,
A projection image of the reflected light is captured by a television camera while irradiating the inspection object with laser line light and rotating the inspection object. The pre-processing unit 6 of the image processing apparatus shown in FIG. 2 obtains an image of a differential absolute value and a differential direction value from the captured image, performs a process such as edge extension, and performs feature extraction.

Next, in the judgment unit 8 of the image processing apparatus, FIG.
As shown in (1), a pixel having a differential value equal to or more than a predetermined value and having an edge flag is obtained by performing raster scanning in a preset inspection area. This utilizes the feature that the laser line light has a higher luminance than the screen background, and the differential value becomes greater than or equal to a predetermined value when image processing is performed. With the obtained pixel as a starting point, a search window of n × m pixels is set and traced around the obtained pixel, and a pixel satisfying the same condition is searched. If pixels satisfying the condition exist continuously and the number of pixels is larger than a preset count value, the line is recognized as a laser line light characteristic image.

Next, the width of the characteristic image and the midpoint o of the width are determined, and the distance from the midpoint o to the characteristic image is radially measured. Using the distance measurement value, the distance is enlarged or reduced to create an inspection line. FIG. 12 shows an inspection line creation method. In the figure, the thick line is the feature image,
The two thin lines above and below are an enlargement line and a reduction line, respectively. When the feature of the defective portion occurs outside the feature line (see FIG. 9A), the distance between the midpoint o and the feature line is enlarged by the measured value plus p (p is arbitrary). When the feature of the defective portion occurs inside the feature line (see FIG. 9B), the distance between the midpoint o and the feature line is reduced by the measured value minus m (m is arbitrary). By shifting the phase of the created enlarged / reduced line in the ± direction, an intersection with the feature image is obtained. The presence / absence of an unevenness defect (or a candidate thereof) on the feature image is detected based on the presence / absence of the intersection.

Here, the reason for performing the phase shift is shown in FIG.
This is because, as shown in (1), the intersection with the inspection line (thin line) cannot be found only by enlarging the characteristic line (thick line) in which the defective portion exists. Therefore, FIG.
As shown in (a) and (b), the intersection is obtained by shifting the phase by a predetermined pixel in the ± direction.
FIG. 14A shows an intersection generated by movement in the negative direction, FIG.
4 (b) shows an intersection generated by the movement in the + direction.

When the intersection of the characteristic line and the inspection line is obtained according to the second embodiment, the range in which the sudden change in the curvature on the characteristic line is to be evaluated is limited to the portion surrounded by the rectangle in FIG. If the feature line is smooth around the intersection, it is unlikely that a large density difference is observed in the direction along the feature line with the intersection as the center. On the other hand, when the feature line is not smooth around the intersection, when the density difference is measured in the direction along the feature line with the intersection as the center,
It is considered that a sudden change occurs. With this in mind,
In the following Examples 3 to 8, by evaluating the magnitude of the density difference occurring around the intersection, the magnitude of the differential absolute value, and the like,
It is determined whether or not the detected intersection is a true irregularity defect.

(Embodiment 3) FIG. 16 is a flowchart showing a processing procedure of an appearance inspection method according to Embodiment 3 of the present invention. In the present embodiment, a density measurement window (stick mask) is installed at the intersection determined in the second embodiment. Here, the stick mask is a bar-shaped mask as shown in FIG. 17, and the center part indicated by a black circle in the figure is coincident with a point to be observed (here, the intersection of the characteristic line and the inspection line). In addition, as shown in FIG. 18, a plurality of pairs of substantially symmetrically arranged masks a1, b1, and a2 set along the direction in which the density change is to be evaluated (here, the direction of the characteristic line at the intersection). b2, a3 and b3, a4 and b4, ...
It is comprised including. The size of each mask is not limited, and may be any size. However, it is convenient if the size of the mask is set in consideration of the size of unevenness defects that occur relatively frequently.

In this embodiment, a stick mask is set along the direction of the characteristic line at the intersection with the inspection line, and the sum of density differences (of the original image) in the set stick mask is determined. The calculated sum of the density differences is compared with a preset threshold value, and if the sum of the density differences is larger than the threshold value, it is determined to be defective. When this is shown by a mathematical formula, the sum of the density differences can be obtained by the following mathematical formula. S = | a1-b1 | + | a2-b2 | + | a3-b3 |
+ | A4-b4 | + ...

In the above equation, a1, b1,..., Etc. mean the densities (light amounts) measured by the individual masks a1, b1,. By comparing the calculated density difference sum S with a preset threshold T1 of the density difference sum,
If S> T1, the result is determined to be defective. This utilizes the fact that characteristic lines of the original image have luminance.
In the case of a non-defective product, there is no intersection in the first place, so there is no erroneous determination.

(Embodiment 4) In Embodiment 4 of the present invention, when a plurality of intersections as defect candidate portions are obtained by the processing procedure of Embodiment 2 described above, the processing of Embodiment 3 is performed for each intersection. Do. That is, a plurality of pairs of windows that are arranged substantially symmetrically along the feature line around each intersection are set, and the sum of the density differences between the paired windows is determined for each intersection, and each intersection is determined. Are accumulated one after another to obtain an integrated value ΣS of the density difference sum S. The integrated value ΣS is compared with a preset threshold value T2, and if ΣS> T2 is satisfied, it is determined that the integrated value is defective. According to the present embodiment, as the number of intersections increases, the difference from non-defective data increases, so the probability of being determined to be defective increases accordingly. For example, the integrated value ΣS is divided by the number of intersections and averaged. It is possible to make a much more severe judgment than to evaluate. That is, the possibility of overlooking a defect can be minimized, and the inspection accuracy can be improved.

(Embodiment 5) In Embodiment 4 of the present invention, the stick mask shown in FIG. 17 is set as shown in FIG. 18 at the intersection determined as a candidate defect portion by the processing procedure of Embodiment 2 described above. . Then, each mask a1, b1, a
The sum E of the differential absolute values of 2, b2, a3, b3, a4, b4,...

E = a1 + b1 + a2 + b2 + a3 + b3
+ A4 + b4 +... In the above equation, a1, b1,... Mean the absolute differential values measured with the individual masks a1, b1,. The obtained sum E of the differential absolute values is compared with a preset threshold value T3 of the sum of the differential absolute values, and if S> T3 is satisfied, it is determined to be defective.

(Embodiment 6) In Embodiment 6 of the present invention, when a plurality of intersections as defect candidate portions are obtained by the processing procedure of Embodiment 2 described above, the processing of Embodiment 5 is performed for each intersection. Do. That is, a plurality of pairs of windows that are arranged substantially symmetrically along the characteristic line around each intersection are set, and the sum of the differential absolute values of the pixels included in each window is obtained for each intersection, and each intersection is determined. Are accumulated one after another to obtain an integrated value ΣE of the total sum E of the differential absolute values. This integrated value ΣE
Is compared with a predetermined threshold value T4, and ΔE
If> T4 is satisfied, it is determined to be defective. Also in the present embodiment, as in the fourth embodiment, as the number of intersections increases, the difference from non-defective data increases, so that the probability of being determined to be bad increases accordingly. It is possible to make a much more severe judgment than to evaluate by dividing by and averaging. In other words, the possibility of overlooking a defect can be minimized,
The inspection accuracy can be improved.

(Embodiment 7) Embodiment 7 of the present invention is a combination of Embodiment 3 and Embodiment 5 described above. The stick mask shown in FIG. And the sum E of the differential absolute values of the respective masks in the stick mask and the sum S of the density differences of the paired masks are obtained, and the sum (E + S) of the two is calculated as a predetermined threshold value T5. In comparison, if (E + S)> T5 is satisfied, it is determined to be defective. According to the present embodiment, even if one of the evaluation based on the sum E of the differential absolute values and the evaluation based on the sum S of the density differences is not effective for some reason, the presence or absence of the unevenness defect can be determined by the other evaluation. The inspection accuracy can be further improved as compared with the case where the third embodiment or the fifth embodiment is used alone.

(Eighth Embodiment) In the eighth embodiment of the present invention, when a plurality of intersections as defect candidate portions are obtained by the processing procedure of the second embodiment, the processing of the seventh embodiment is performed for each intersection ( That is, the processing of the third and fifth embodiments) is performed, and the sum of the integrated value of the sum E of the differential absolute values and the sum S of the density differences (Σ
E + ΣS) is obtained and compared with a preset threshold value T6. If (ΣE + ΣS)> T6 is satisfied, it is determined to be defective. According to this embodiment, the effects of the seventh embodiment and the effects of the fourth and sixth embodiments can be simultaneously obtained. The number of intersections, the density difference in the direction along the characteristic line around each intersection, and the differential absolute By comprehensively evaluating the values, it is possible to almost completely discriminate between a good product and a defective product.

(Embodiment 9) FIG. 19 is a flowchart showing a processing procedure of an appearance inspection method according to Embodiment 9 of the present invention. In the present embodiment, the pass / fail judgment is performed using the data of the differential direction value described in the second embodiment, and the evaluation method is completely different from the third to eighth embodiments. That is, in the present embodiment, the differential direction value of the feature line extracted by the image processing is obtained, and the change pattern of the differential direction value when the inspection object is a non-defective product is stored in advance as a non-defective product pattern. The change pattern of the differential direction value of the inspection object is compared with a non-defective pattern, and the presence / absence of a difference in the continuity of the differential direction value is detected to detect a concave / convex defect.

As described above, the image processing apparatus shown in FIG. 2 has a function of storing the density change direction as a differential direction value image. The differential direction value is 0 as the density change direction.
It has values in 16 directions from F to F. FIG. 20 shows the relationship between the density change direction and the differential direction value. In the laser line light characteristic image (in the case of a non-defective product) shown in FIG. 21, the differential direction values are as shown in the figure. In the figure, the background is black,
A portion having the brightness of the laser line light is white. The differential direction value is defined to indicate a direction in which black changes to white. In the case of a non-defective product, as shown in FIG. 21, the differential direction values around the characteristic line indicate continuous values from A to E on the upper side and 2 to 6 on the lower side. However, when there is a defect, as shown in FIG. 22, there is a characteristic that the continuity of the differential direction value is lost. By utilizing the feature, only the concave / convex defect appearing on the laser line light feature image can be extracted.

In other words, in the case of a non-defective product, the upper differential direction value pattern is A-A from the left of the image as shown in FIG.
It changes to BCDE. This change pattern is stored in advance as a non-defective pattern, and is compared with a change pattern obtained by searching a feature image during inspection. If there is a defect, the pattern of the differential direction value is ABCCABDCEDE from the left of the image as shown in FIG.
Since it is changed to C-D-E, it is possible to determine the acceptability by obtaining the difference from the non-defective pattern. For example, it is assumed that D always comes after C, and when this rule is violated, it is determined that the device is defective, so that the inspection can be easily performed.

(Embodiment 10) FIG. 24 shows Embodiment 1 of the present invention.
11 is a flowchart showing a processing procedure of a visual inspection method of No. 0; In this embodiment, the pass / fail judgment is made by using the data of the differential direction value described in the second embodiment. In the ninth embodiment, the change pattern of the differential direction value is used. The example is different in that the length (full length) of the change pattern of the differential direction value is used. That is, in the case where the non-defective pattern has the length shown in FIG. 25A and the unevenness defect exists, the pattern of the differential direction value is changed as shown in FIG.
It becomes longer as shown. By utilizing this feature, the quality can be determined by comparing the total lengths of the patterns of the differential direction values. This is based on the fact that, in the case of a non-defective image, the same waveform can be obtained when the inspection object is rotated and imaged. Further, by using the comparison of the change pattern of the differential direction value described in the ninth embodiment and the comparison of the total length of the change pattern according to the present embodiment together, a stable inspection can be performed.

(Embodiment 11) FIG. 26 shows Embodiment 1 of the present invention.
4 is a flowchart showing a processing procedure of the first visual inspection method. In this embodiment as well, the pass / fail judgment is made using the data of the differential direction value described in the second embodiment, and the differential direction value at the reference position set on the change pattern of the differential direction value and this reference position In this case, the irregularity defect is detected by comparing the differential direction value at a position away from the pixel by a preset value. That is, the differential direction value at a position apart from the reference point by a predetermined pixel is obtained while tracking the characteristic image described in the second embodiment. In the case of a non-defective product, using the fact that the differential direction value continuously increases, decreases, or shows the same value, the obtained two differential direction values are compared, and a change different from the non-defective pattern (for example,
Change in the direction of becoming smaller, although it should be larger). If the pattern is different from the non-defective pattern, it is determined to be defective.

FIG. 27 illustrates a specific inspection method. The pixel is tracked one by one from the left side of the feature image, and is compared with the differential direction value of the preceding pixel by a predetermined number of pixels from there. Normally, it should have the same value or a large value (it may be smaller depending on the image). Therefore, the differential direction value of the pixel at the reference position is compared with the differential direction value of the pixel separated from the reference position by a predetermined pixel, and is larger than the differential direction value of the reference position, or
If they are the same, it is determined to be good. When the value is larger than the differential direction value of the reference position, it is preferable that the range can be set. For example, if the differential direction value is A to E, it is determined to be good.

(Twelfth Embodiment) A twelfth embodiment of the present invention relates to the removal of noise, and corresponds to the twelfth embodiment. When foreign matter such as dust adheres to the surface of the inspection object, the pattern of the differential direction values may be disturbed. For example, when the non-defective pattern of the differential direction value is as shown in FIG. 28A from the left of the feature image, the pattern of the differential direction value for the defective image may be as shown in FIG. 28B. is there. As in this example, when the differential direction value is disturbed over one or two pixels due to the influence of dust or the like,
Judging from the continuity of the differential direction value, it is determined to be defective,
When the feature of the irregularity defect is designed to have a size of 10 to 20 pixels (depending on the inspection optical system), for example, the continuity of the differential direction value and the width of the defect candidate portion are reduced. By storing as inspection parameters in advance, noise of about one pixel (resolution of 400 microns) can be removed. That is, the number of pixels in the defect candidate portion is compared with the threshold value, and if the number of pixels in the defect candidate portion is larger than the threshold value, it is determined that the pixel is defective.

Embodiment 13 FIG. 29 shows Embodiment 1 of the present invention.
3 is a system configuration diagram of FIG. The image captured by the CCD camera 4 is subjected to preprocessing by the image processing unit 60, and is transferred to the image memory unit 7A. A laser line light characteristic image is extracted based on the transferred image. The extracted characteristic images are further transferred to the image memory unit 7B and accumulated. After completing the accumulation of all the images obtained by rotating the object to be inspected by 180 degrees, the processing of the second to tenth embodiments is performed on the accumulated multiple feature images by using the arithmetic CPU 80 to obtain the defective portion. Is extracted.

FIG. 30 shows an example of a characteristic image obtained by rotating the object to be inspected from 0 to 180 degrees when there is a defective portion. In this case, an example of the stored multiple feature image is shown in FIG. in this way,
By accumulating and superimposing a plurality of characteristic images, it is possible to emphasize a defective portion, and it is possible to improve detection accuracy.

Regarding the method of accumulating each characteristic image, the position of the image projected on the screen does not change even if the dome-shaped inspection object is rotated by the laser line light. What is necessary is just to integrate the concentrations of. If the position of the image displayed on the screen fluctuates slightly with the rotation of the inspection object, a process of matching both ends of the characteristic image between the images may be performed.

[0048]

According to the first to thirteenth aspects of the present invention, the surface of the dome-shaped inspection object is irradiated with a laser beam in a linear manner.
Since the reflected light is projected on the screen and the characteristics of the irregularities appearing on the characteristic line are recognized and the quality of the inspected object is determined, the irregularities generated on the surface of the dome-shaped inspected object are determined. There is an effect that defects can be determined by visual inspection.

According to the second aspect of the present invention, an inspection line having substantially the same shape is created by shifting the characteristic line obtained by the image processing, and the intersection of the characteristic line and the inspection line is obtained, thereby forming the defect candidate portion. Because it was detected,
An area where a rapid change on the feature line should be evaluated can be limited only to the vicinity of the intersection, and the inspection speed can be improved.

According to the third, fifth or seventh aspect of the present invention, an evaluation window is set in the vicinity of the intersection detected as a defect candidate portion, and the density difference and the magnitude of the differential absolute value are compared with the reference value. As a result, the irregularity defect was determined by
It is possible to reliably determine whether or not it is a true irregularity defect. According to the invention of claim 4, 6 or 8, when there are a plurality of intersections, a mask is set at each intersection, and the sum of the differential absolute values and the sum of the density differences for each mask are obtained. By adding the data obtained every time, the difference from the data of the non-defective product can be increased, and the inspection accuracy can be improved.

According to the ninth aspect of the present invention, a differential direction value of the characteristic line extracted by the image processing is obtained, and a change pattern of the differential direction value when the inspection object is a non-defective product is stored as a non-defective pattern in advance. Since the pattern of change in the differential direction of the test object to be measured is compared with a non-defective pattern, and the difference is detected to detect unevenness defects, advanced image processing such as enlargement or reduction of feature lines is performed. There is an advantage that the presence / absence of a defect can be detected only by comparing the patterns of the numerical data without performing, and the load on the processing device can be reduced.

According to the tenth aspect of the present invention, the entire length of the characteristic line is used as a criterion for determining unevenness defects. Can be improved. According to the eleventh aspect, the differential direction value of the characteristic line extracted by the image processing is obtained, and the differential direction value at the reference position set on the change pattern of the differential direction value, and the differential direction value set in advance from this reference position are determined. Since the unevenness defect is detected by comparing the value with the differential direction value at a position apart from the pixel, there is an advantage that it is not necessary to previously store a non-defective pattern.

According to the twelfth aspect of the present invention, even if the pattern of the differential direction value of the characteristic line on the surface of the object to be inspected is disturbed due to dust or the like, the irregularity defect can be detected while maintaining the inspection accuracy. . According to the thirteenth aspect, the obtained characteristic line is obtained for each inspection screen, and the defect portion is emphasized by superimposing, so that the inspection accuracy can be further improved.

[Brief description of the drawings]

FIG. 1 is a perspective view showing the appearance of an inspection optical system for performing an appearance inspection method of the present invention.

FIG. 2 is a block diagram of an image processing system for implementing the appearance inspection method of the present invention.

FIG. 3 is an explanatory diagram showing an example of an original image obtained by imaging a screen by the appearance inspection method of the present invention.

FIG. 4 is an explanatory diagram of a local parallel window used in the appearance inspection method of the present invention.

FIG. 5 is an explanatory diagram showing an example of a differential absolute value image used in the appearance inspection method of the present invention.

FIG. 6 is an explanatory diagram of edge extension processing used in the appearance inspection method of the present invention.

FIG. 7 is an explanatory diagram showing a projected image in a case where the inspection object according to the present invention is a non-defective product.

FIG. 8 is an explanatory view showing a projected image when an object to be inspected is a defective product in the present invention.

FIG. 9 is an explanatory diagram showing a reflected light path on the surface of the inspection object in the present invention.

FIG. 10 is a flowchart illustrating a processing procedure of an appearance inspection method according to the first embodiment of the present invention.

FIG. 11 is a flowchart illustrating a processing procedure of an appearance inspection method according to a second embodiment of the present invention.

FIG. 12 is an explanatory diagram illustrating a method for creating an inspection line according to the second embodiment of the present invention.

FIG. 13 is an explanatory diagram showing a relationship between an inspection line and a feature line before shifting the phase according to the second embodiment of the present invention.

FIG. 14 is an explanatory diagram illustrating a relationship between an inspection line and a feature line after a phase shift according to the second embodiment of the present invention.

FIG. 15 is an explanatory diagram showing an intersection of an inspection line and a feature line in Embodiment 2 of the present invention.

FIG. 16 is a flowchart illustrating a processing procedure of an appearance inspection method according to a third embodiment of the present invention.

FIG. 17 is an explanatory diagram showing a stick mask used in Embodiment 3 of the present invention.

FIG. 18 is an explanatory diagram showing an arrangement of a stick mask used in Embodiment 3 of the present invention on an inspection line.

FIG. 19 is a flowchart showing a processing procedure of an appearance inspection method according to a ninth embodiment of the present invention.

FIG. 20 is an explanatory diagram showing a relationship between a differential direction value and a density change direction used in Embodiment 9 of the present invention.

FIG. 21 is an explanatory diagram in which differential direction values used in a ninth embodiment of the present invention are written and shown on a feature image for a non-defective product.

FIG. 22 is an explanatory diagram showing a differential direction value used in the ninth embodiment of the present invention written on a feature image for a defective product.

FIG. 23 is an explanatory diagram showing the continuity of the differential direction value used in the ninth embodiment of the present invention for a good product and a defective product.

FIG. 24 is a flowchart showing a processing procedure of an appearance inspection method according to a tenth embodiment of the present invention.

FIG. 25 is an explanatory diagram showing the length of the pattern of the differential direction value used in the tenth embodiment of the present invention for a non-defective product and a defective product, respectively.

FIG. 26 is a flowchart showing a processing procedure of an appearance inspection method according to an eleventh embodiment of the present invention.

FIG. 27 is an explanatory diagram illustrating a specific inspection example of the appearance inspection method according to the eleventh embodiment of the present invention.

FIG. 28 is an explanatory diagram of an appearance inspection method according to Example 12 of the present invention.

FIG. 29 is a system configuration diagram of a thirteenth embodiment of the present invention.

FIG. 30 is an explanatory diagram showing an example of a characteristic image obtained by rotating an inspection object having a defect portion on its surface from 0 ° to 180 ° in Example 13 of the present invention.

FIG. 31 is an explanatory diagram illustrating an example of a multiple feature image used in Embodiment 13 of the present invention.

[Description of sign]

 DESCRIPTION OF SYMBOLS 1 Inspection object 2 Line light source 3 Screen 4 CCD camera L Projection image

Claims (13)

[Claims] 1. A method of irradiating a laser beam in a line on the surface of a dome-shaped inspection object, and recognizing a characteristic of a concave / convex defect portion appearing on a characteristic line obtained by projecting a reflected light on a screen. A visual inspection method characterized by determining the quality of an object. 2. The method according to claim 1, wherein an image including the characteristic line projected on the screen is captured, and the characteristic line is extracted as a substantially arched curve by image processing. Create an inspection line by slightly translating it in a substantially vertical direction or moving it in a scaled and similar manner, and slightly translating it along the longitudinal direction of the arch to create an inspection line and the original features. An appearance inspection method, wherein a defect candidate portion is detected by finding an intersection with a line. 3. The method according to claim 2, wherein a plurality of pairs of windows arranged substantially symmetrically along the characteristic line with respect to the intersection determined as the defect candidate portion are set, and the density between the paired windows is set. An appearance inspection method, wherein a sum of differences is obtained and compared with a preset reference value to detect unevenness defects on a feature line. 4. A method according to claim 2, wherein when there are a plurality of intersections obtained as defect candidate portions, a plurality of pairs of windows are set substantially symmetrically along the characteristic line with respect to each intersection. Then, the sum of the density differences between each paired windows is obtained for each intersection, all of them are added, and the sum is compared with a preset reference value to detect unevenness defects on the feature line. Appearance inspection method. 5. A method according to claim 2, wherein a plurality of pairs of windows arranged substantially symmetrically along the characteristic line with respect to the intersection determined as the defect candidate portion are set, and a differential absolute value of a pixel included in each window is set. An appearance inspection method characterized in that unevenness defects on a feature line are detected by calculating the sum of the values and comparing the sum with a preset reference value. 6. A method according to claim 2, wherein when there are a plurality of intersections determined as defect candidate portions, a plurality of pairs of windows are set substantially symmetrically along the characteristic line with each intersection as a center. Then, the sum of the differential absolute values of the pixels included in each window is obtained for each intersection, and the sum of all the sums is compared with a preset reference value to detect unevenness defects on the feature line. Features a visual inspection method. 7. A method according to claim 2, wherein a plurality of pairs of windows arranged substantially symmetrically along the characteristic line with respect to the intersection determined as the defect candidate portion are set, and the density between each pair of windows is set. Finding the sum of the differences and the sum of the differential absolute values of the pixels included in each window, and comparing the sum with a preset reference value, detects irregularities on the feature line. Appearance inspection method. 8. A method according to claim 2, wherein when there are a plurality of intersections determined as defect candidate portions, a plurality of pairs of windows are arranged substantially symmetrically along the characteristic line with each intersection as a center. Then, the sum of the density differences between each pair of windows is obtained for each intersection, and all are added.The sum of the differential absolute values of the pixels included in each window is obtained for each intersection, and all are added. An appearance inspection method characterized by detecting unevenness defects on a feature line by comparing the sum of both addition values with a preset reference value. 9. The method according to claim 1, wherein an image including the characteristic line projected on the screen is captured, the characteristic line is extracted as a substantially arched curve by image processing, and a differential direction value of the extracted characteristic line is obtained. The change pattern of the differential direction value when the test object is a good product is stored in advance as a good product pattern, and the change pattern of the differential direction value of the test object to be measured is compared with the good product pattern to detect a difference. An appearance inspection method characterized by detecting unevenness defects by using the method. 10. The method according to claim 1, wherein an image including the characteristic line projected on the screen is captured, the characteristic line is extracted as a substantially arched curve by image processing, and the total length of the extracted characteristic line is obtained. An appearance inspection method characterized by storing in advance the entire length of a characteristic line when the inspection object is a non-defective product, and detecting an irregularity defect by comparing with the entire length of the characteristic line of the inspection object to be measured. 11. The method according to claim 1, wherein an image including the characteristic line projected on the screen is captured, the characteristic line is extracted as a substantially arched curve by image processing, and a differential direction value of the extracted characteristic line is obtained. , The differential direction value at the reference position set on the differential direction value change pattern,
An appearance inspection method characterized by detecting a concave / convex defect by comparing with a differential direction value at a position apart from the reference position by a preset pixel. 12. The visual inspection method according to claim 9, wherein when there is a disturbance of a predetermined number of pixels or less on the change pattern of the differential direction value, the disturbance is removed as noise. 13. The imaging method according to claim 1, wherein an image including the characteristic line projected on the screen is captured, and the characteristic line is extracted as a substantially arch-shaped curve by image processing. A feature inspection method for emphasizing a defective portion by superimposing the extracted feature lines and detecting an uneven defect based on the multiple feature lines obtained by the superimposition.