JP2000105831A - Device and method for surface defect inspection - Google Patents

Abstract

PROBLEM TO BE SOLVED: To decide a rugged defect such as a dent or sticking foreign matter on a metal surface and a defect such as rust or dirt by forming its image by one kind of lighting system forming the image of the defect part by irradiation with a slit pattern, extracting an area which is different from the direction of the slit pattern from the directional vector image, and performing inspection. SOLUTION: Reflected light from an object 2 is picked up by a television camera 5 and inputted to an image processing part, and a processing area setting means 10 sets a range to be processed as a processing area in the A/D converted image. Then a binary image generating means 11 threshold processes the image data in the process area and starts scanning at a slit pattern scan start position through a directional vector filter to generate a vector image. When a feature quantity is not decided, a directional vector area which is different in the direction of the slit pattern is extracted as a candidate area for a defect and the total number of the pixels in the defect candidate area is found to make a quality decision according to the total value.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to, for example, surface irregularities such as scratches, dents, foreign matter adhesion, rust, etc. on metal surfaces such as defect inspection of metal films of diaphragms of condenser microphones of portable telephones. The present invention relates to a surface defect inspection apparatus and method for inspecting defects such as dirt by image processing.

[0002]

2. Description of the Related Art In recent years, there has been an increasing demand for automation of exterior inspection of basic mechanical parts and products, and research and development of inspection automation by non-contact image processing using a TV camera or the like have been actively conducted. Hereinafter, an example of the above-described conventional defect inspection apparatus will be described with reference to the drawings. FIG. 5 is a configuration diagram of a conventional defect inspection apparatus using image processing. A plurality of illumination devices 23 (1) to 23 (n) for inputting an object 22 to be inspected placed on a positioning table 21
(N is the number of lighting devices), and a TV camera 24 is mounted on the movable TV camera support 25. here,
The television camera 24 is controlled by the television camera control means 26.

A video signal input from a television camera 24 is converted into an analog-to-digital (hereinafter referred to as an A / D converter) 27 by an analog-to-digital converter (0-255 (25)) depending on the density of an image.
The data is digitized into image data such as (6 gradations) and input to an image processing unit including a CPU (control unit), ROM, RAM and the like. The image processing unit includes a processing area setting unit 29 for setting an area to be processed, a binarizing unit 31 for binarizing an image and extracting a defect candidate portion, an area, a width, and a length of the extracted region. Amount detecting means 32 for detecting the characteristic amount of
And a pass / fail judgment circuit 33 for judging pass / fail of the candidate area based on the feature, a lighting control means 30 for controlling the lighting devices 23 (1) to 23 (n), and a CPU 28.

The operation of the above-structured apparatus will be described below. First, as shown in the flowchart of FIG. 6, an object 22 to be inspected placed on a positioning table 21 is imaged by a television camera 24 under an illumination device 23 (1) and input to an image processing unit. A /
D-converted (step 21), and set a processing area for the input image as a processing area (step 2).
2), binarize the image in the processing area to extract a defect candidate portion (step 23), and determine the area of the extracted region,
A feature amount such as a width and a length is detected (step 24), and a pass / fail judgment based on the feature amount is performed to inspect a flaw (step 2).
5). Next, an image is formed under a necessary lighting device among the other lighting devices 23 (2) to 23 (n), and the same processing is repeated. However, it is difficult to image all the defects using only one illumination device, and a method of performing an inspection process by switching a plurality of illuminations (changing the illumination irradiation angle) to form an image is usually adopted.

[0005]

However, in the above-described apparatus configuration, there is a certain correspondence between illumination irradiation and imaging of a defect. Illumination in different directions is prepared, switched, imaged, and inspection processing is performed. Therefore, even if various defect contents are handled, guaranteeing the detection performance requires a considerably large number of illuminating devices, which increases the introduction cost and the processing time. SUMMARY OF THE INVENTION The present invention solves the above-mentioned conventional problems, and is a defect in which a defect such as a dent on a metal surface or the like, a concavo-convex defect such as adhesion of a foreign substance, or rust or dirt is imaged and inspected by one type of illumination system. It is an object to provide an inspection device and method.

[0006]

In order to achieve this object, the present invention is configured as follows. First of the present invention
According to the aspect, in an image processing apparatus for inspecting a defect on a metal surface or the like, a dent, or a concavo-convex defect of foreign matter attachment, rust, or a dirt defect, a slit pattern light projecting a slit pattern light on an inspection object Means, image input means for capturing reflected light of the slit pattern light from the object, and direction vector image creation for creating a direction vector image from a change in the density of an image from the input image input to the image input means Means, a defect candidate area extracting means for extracting an area different from the direction of the slit pattern from the direction vector image as a defect candidate, and a good or bad judgment means for judging the quality of the defect candidate area. Provide a surface defect inspection device.

According to a second aspect of the present invention, there is provided the surface defect inspection apparatus according to the first aspect, wherein the pass / fail judgment means judges pass / fail from the total sum of the defect candidate areas.

According to the third aspect of the present invention, the area combining means for recognizing the area for each lump from the extracted defect candidate area and performing the area combining process based on the surrounding situation, and the area after each combining Means for detecting a plurality of feature values selected from among the feature values of the area, density, width, and length, wherein the pass / fail determination means determines pass / fail from the detected plurality of feature values. A surface defect inspection device according to the first or second aspect, wherein the determination is performed.

According to a fourth aspect of the present invention, there is provided an image processing method for inspecting a metal surface or the like for scratches, dents, or irregularities due to foreign matter adhesion, rust, or dirt defects. Irradiate the object, image the reflected light from the target object of the slit pattern light, to create a direction vector image from the density change of the image from the captured image,
A surface defect inspection method characterized in that an area different from the direction of the slit pattern is extracted as a defect candidate from the direction vector image, and the quality of the defect candidate area is determined.

[0010] According to a fifth aspect of the present invention, there is provided the surface defect inspection method according to the fourth aspect, wherein the pass / fail judgment is made based on the total sum of the defect candidate areas when the pass / fail judgment is performed.

According to the sixth aspect of the present invention, a region is recognized for each lump from the extracted defect candidate region, a region combining process is performed based on a surrounding situation, and an area, A fourth or fifth aspect in which a plurality of feature amounts selected from among the feature amounts of density, width, and length are detected, and at the time of the pass / fail determination, pass / fail is determined from the detected plurality of feature amounts. And a surface defect inspection method according to (1).

[0012]

Embodiments of the present invention will be described below in detail with reference to the accompanying drawings. An embodiment of the present invention will be described below with reference to FIGS.

(First Embodiment) FIG. 1 is a configuration diagram of a surface defect inspection apparatus and method according to a first embodiment of the present invention. A slit pattern coaxial epi-illumination device 3 is installed to irradiate an object 2 to be inspected placed on a positioning table 1, and a television camera 5 and a lens 6 are installed as a set on a movable television camera support 4. I have. here,
The television camera 5 is controlled by the television camera control means 7. A video signal input from the television camera 5 is converted into an analog-to-digital (hereinafter referred to as an A / D converter) section 8 by an analog-to-digital conversion section 8 according to the density of the image.
Image data such as gradation, CPU, ROM,
The data is input to an image processing unit including a RAM and input / output. The slit pattern coaxial epi-illumination device 3 mainly includes a light source 3a, a lens 3b for converting illumination light into parallel light, a slit pattern 3c, a lens 3d for enlarging or reducing the slit pattern light, and a half mirror 3e.

The image processing section is a judgment control means (CPU) 9 to which a command is given from the main controller or the operation panel.
Processing area setting means 10 for designating an area to be processed
And 2 of the image data A / D converted by the A / D converter 8
A binary image creating means 11 for performing a binarization, a contour point detecting means 12 for detecting a contour point of a slit pattern based on an input from the binary image creating means 11, and a direction vector is created by tracking the contour points. Direction vector image creating means 13;
Defect candidate area extraction means 14 for extracting a direction vector area different from the direction of the slit pattern as a defect candidate area based on an input from the direction image vector image creation means 13
And a pass / fail judgment means 15 for judging pass / fail from the total sum of the defect candidate areas. The operation of the surface defect inspection apparatus configured as described above, that is, the surface defect inspection method will be described with reference to FIG.

An object 2 to be inspected placed on a positioning table 1 is irradiated with a slit pattern 3c by a slit pattern coaxial epi-illumination 3, as shown in FIG. The reflected light is picked up by the TV camera 5 and is input to the image processing unit via the TV camera control means 7 and A / D converted by the A / D conversion unit 8 (step 1). Here, the principle of inspection by slit pattern irradiation will be described. When irradiating the slit pattern 3c in various irregularities of the object 2 and when the irregularities are gentle as shown in FIG. 3C, the reflected light is imaged by the camera 5 as shown in FIG. As shown in FIG. 30C, the slit pattern 3c is slightly distorted and the image is taken. As shown in FIG. 3D, as the unevenness becomes sharper, the reflected light of the slit pattern 3c escapes farther and is captured by the television camera 5. As a result, the slit pattern is interrupted as shown by 30d in FIG. The width of the slit of the slit pattern is determined by the size of the defect and the resolution in image processing. For example, when detecting a 10 μm flaw, it is assumed that the resolution in image processing is 1/10. At this time, the product of the size of the flaw and the resolution, that is,
Even if a slit having a width smaller than 1 μm, which is one tenth of 10 μm, cannot be used for defect detection, the slit width is set to 1 μm or more. Preferably, the slit width is about one half of the size of the flaw, that is, 5 μm. The narrower the slit width, the higher the accuracy of defect detection and the more detailed the state of the defect can be detected.

For the image input to the image processing unit and A / D converted, a processing area is set as a processing area by the processing area setting means 10 (step 2).
The processing area is set according to the shape of the object 2. Next, the image data in the processing area is binarized by the binary image creating means 11 (step 3).

Next, as shown in FIG. 4A, scanning of the binarized image of the slit pattern is started from the slit pattern scanning start position by a direction vector filter as shown in FIG. 4B. (Step 4), a vector image is created (step 5). FIG. 4C shows an example of a direction vector image based on the slit pattern as shown in FIG. Here, the direction vector creation filter substitutes 8 if the pixel of interest is not a white region, and if the pixel of interest is a white region, focuses on a direction value that changes clockwise from 0 to 1 around the pixel of interest. This is to be assigned to the pixel. That is, in the slit pattern of FIG. 4A, while moving along the horizontal direction from the scanning start position at the upper left end to the right end from the left end, and from the right end to the left end in the pattern one step below, as described above, Direction vectors for eight surrounding directions are determined. More specifically, for example, when the pixel of interest is 255 (when it is bright), a location that changes from 0 to 255 is searched clockwise from the position of the direction vector 0 in FIG. , Part 2
The direction vector of the 55th point is set as the direction vector of the target pixel. If not found, the direction vector is set to 8.

Next, in the case where the characteristic amount determination is not performed in step 6 (the case of performing the characteristic amount determination will be described later as a second embodiment), a direction vector region (0, 4) different from the direction of the slit pattern 3c is used. , 8) are extracted as defect candidate areas (step 7). FIG.
In the direction vector image of (C), the part with a circle is the part to be extracted as a defect candidate area. Next, the sum of the number of pixels in the defect candidate area is obtained, and the pass / fail judgment is made from the sum (step 8). That is, in order to roughly grasp the state of a defect, the sum of the number of defective pixels is calculated, and if the value of the sum exceeds a threshold value, it is determined that there is a defect. This threshold is determined empirically or by trial and error so as not to detect noise.

According to the first embodiment, the slit pattern projecting means for irradiating the object, the image input means for capturing the reflected light of the slit pattern, and the direction vector based on the change in the density of the image from the input image A direction vector image creating means for creating an image, a defect candidate area extracting means for extracting an area different from the direction of the slit pattern from the direction vector image as a defect candidate, and a pass / fail judgment based on the sum of the defect candidate areas And has a function of imaging and inspecting defects such as dents on a metal surface, irregularities such as foreign matter adhesion, rust, or dirt using one type of illumination system. be able to.

(Second Embodiment) As shown in FIG. 1, a surface defect inspection apparatus and method according to a second embodiment of the present invention include a defect candidate area in the configuration of the inspection apparatus and method of the first embodiment. A region combining unit 16 for recognizing a region for each lump from the group and performing a region combining process based on a surrounding situation;
Means 17 for detecting a feature quantity such as area, density, width or length from the region after each combination is added;
5 is configured to judge pass / fail from the plurality of detected feature amounts.

More specifically, in the case where the feature amount is determined in step 6 in the apparatus of the first embodiment, a fixed range is set after extracting a defect candidate area as in step 7 (step 9). The pitch and directionality of the slit pattern, such as combining the defect candidate areas in the certain range as the same or combining defect candidate areas in the vicinity of 8 or 16 with the direction vector difference in a certain range. (Step 1)
0). Next, a plurality of feature values of the combined defect candidate area are detected (step 11), and pass / fail judgment is performed based on the feature values (step 8).

According to the second embodiment, a region is recognized for each lump from the defect candidate region, and a region combining process is performed by the region combining means based on a surrounding situation. Width, or a means for detecting a feature amount such as a length, and a pass / fail judgment means for judging pass / fail from the plurality of detected feature amounts. It can have the effect of identifying the content of the defect.

[0023]

According to one aspect of the present invention, a slit pattern is irradiated even when inspecting a plurality of defects such as dents on a metal surface or the like, irregularities such as foreign matter adhesion, rust, or dirt. To provide a surface defect inspection apparatus and method for inspecting all defects with one type of illumination system in order to image a defect portion, extract a region different from the direction of the slit pattern from the direction vector image, and perform inspection. Is possible. Further, according to another aspect of the present invention, since the areas are combined from the defect candidate areas and the quality is determined based on a plurality of feature amounts, the defect content can be identified.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a surface defect inspection apparatus according to a first embodiment of the present invention.

FIG. 2 is a flowchart of a surface defect inspection method which is an operation of the inspection apparatus of the first embodiment.

FIG. 3 is a diagram illustrating the principle of imaging a defect by irradiating a slit pattern in the first embodiment.

FIG. 4 is an explanatory diagram of a method for extracting a defect candidate area from a slit pattern image in the first embodiment.

FIG. 5 is a configuration diagram of a conventional surface defect inspection apparatus.

FIG. 6 is a flowchart of the operation of the conventional example.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Positioning table, 2 ... Object, 3 ... Slit pattern coaxial epi-illumination device, 3a ... Light source, 3b ... Lens, 3
c: slit pattern, 3d: lens, 3e: half mirror, 4: television camera support, 5: television camera, 6
... Lens, 7 ... TV camera control means, 8 ... A / D converter, 9 ... CPU (judgment control means), 10 ... Processing area setting processing area, 11 ... Binary image creating means, 12 ... Contour point detecting means, 13 ... direction vector image creation means, 14 ... area extraction means, 15 ... pass / fail judgment means, 16 ... area combination means, 17 ... feature amount detection means.

Continued on front page F-term (reference) 2G051 AA37 AA61 AA90 AB01 AB07 BA00 CA04 CB01 CD04 EA11 EA12 EA14 EB01 EB02 ED07 5B047 AA11 AB02 BA01 BC11 CB23 DC09 5B057 AA01 BA02 BA15 CA08 CA12 CA16 CC01 DC03 DB02 DB02

Claims (6)

[Claims] An image processing apparatus for inspecting scratches, dents on metal surfaces or the like, asperity defects due to foreign matter adhesion, rust or dirt defects, wherein a slit pattern light is emitted to an inspection object with slit pattern light. Means (3); image input means (5) for imaging reflected light of the slit pattern light from the object; and a direction vector image based on a change in density of an image from the input image input to the image input means. A direction vector image creating means (13) to be created; a defect candidate area extracting means (14) for extracting an area different from the direction of the slit pattern from the direction vector image as a defect candidate; and determining whether the defect candidate area is good or bad. Pass / fail judgment means (1
5) A surface defect inspection device characterized by comprising: 2. The surface defect inspection apparatus according to claim 1, wherein the pass / fail judgment means (15) judges pass / fail from the sum of the defect candidate areas. 3. A region combining means (16) for recognizing a region for each lump from the extracted defect candidate region and performing a region combining process based on a surrounding situation; (17) for detecting a plurality of feature amounts selected from among the feature amounts of width, length, and length. The surface defect inspection apparatus according to claim 1 or 2, wherein the determination is made. 4. An image processing method for inspecting a surface defect such as a scratch, a dent, or a foreign matter attached to a metal surface or the like, a rust, or a dirt defect. An image of light reflected from the object is taken, a direction vector image is created from a change in density of the image from the taken image, and an area different from the direction of the slit pattern is extracted from the direction vector image as a defect candidate. A defect inspection method for determining whether the defect candidate area is good or bad; 5. The surface defect inspection method according to claim 4, wherein the pass / fail judgment is made based on the sum of the defect candidate areas. 6. A region is recognized for each lump from the extracted defect candidate region, a region combining process is performed based on a surrounding situation, and an area, a density, a width, and a length are determined from each combined region. 6. The surface defect inspection method according to claim 4, wherein a plurality of feature amounts selected from the feature amounts are detected, and at the time of the pass / fail determination, pass / fail is determined from the detected plurality of feature amounts.