JP2000266687A - Device and method for visual inspection - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make appropriately discriminable a mark from a defective part being executed to a target surface by detecting a mark candidate line and the number of continuous lines by scanning the main scanning line of the binary image of the surface to be inspected. SOLUTION: The image of an end face RA of an outer track ring R is picked up by a CCD camera 3, and an image signal is converted to binary image data. Then, in the binary image, scanning is made along the radial direction of the outer track ring R with a specific interval over the entire periphery of the end face RA. In this process, when a mark candidate line that is a main scanning line including a black pixel is detected, the number of continuous mark candidate lines is examined. Then, if the value is within the range of the maximum and minimum values of the number of lines where the mark spans, it is judged that the mark is provided at the region of the mark candidate line. Otherwise, it is judged that a defective part exists at the region, thus drastically reducing inspection time since it is not necessary to recognize the shape of each region represented by a black pixel.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an appearance inspection apparatus and an appearance inspection method for inspecting the appearance of an inspection object.

[0002]

2. Description of the Related Art For example, in a product inspection of a bearing, scratches, rust, black scale residue (after heat treatment of an outer race or an inner race) is formed on an end surface or a peripheral surface of an outer race or an inner race of a bearing. An appearance inspection device that automatically inspects whether or not a defective portion such as a discolored surface remaining on a polished surface) is used.

In an inspection using this appearance inspection apparatus, an inspection target surface is imaged by a CCD camera, and an image signal corresponding to the imaged inspection target surface is output from the CCD camera. Then, the image signal output from the CCD camera is converted into binary image data by binarizing the image signal with a certain threshold value. Based on the binary image data, the presence or absence of a defective portion on the inspection target surface is determined. Is determined.

For example, in a binary image obtained based on the converted binary image data, a defective portion is represented by a black pixel and an area other than the defective portion is represented by a white pixel. The presence or absence of the part can be determined.

[0005]

By the way, the surface to be inspected, such as the end surface of the outer race of the bearing, is often engraved with, for example, the name of the country of manufacture, the name of the manufacturer, and the model number. If the mark is applied to the surface to be inspected, it is difficult for the conventional visual inspection apparatus described above to distinguish between the mark and the defective portion, so that the presence or absence of the defective portion cannot be accurately determined. Was.

Therefore, when a mark is applied to the surface to be inspected, the shape of each region represented by black pixels is recognized from the binary image of the surface to be inspected, and the shape is actually applied to the surface to be inspected. It is conceivable to perform a so-called pattern search in which an area having a shape other than the shape of the stamp is determined as a defective portion. However, in this pattern search, it is necessary to recognize the shape of each area represented by black pixels.
Inspection time is long. If the quality of the stamp is low, the stamp may be erroneously determined as a defective portion.

It is an object of the present invention to solve the above-mentioned technical problems and to provide an appearance inspection apparatus and an appearance inspection method capable of satisfactorily discriminating a mark and a defective portion formed on a surface to be inspected. . Another object of the present invention is to provide a visual inspection apparatus and a visual inspection method capable of satisfactorily determining the quality of a surface to be inspected.

[0008]

According to the first aspect of the present invention, there is provided an appearance inspection apparatus for inspecting the appearance of an object having an engraved inspection object surface. An imaging unit that images the inspection target surface and outputs an image signal corresponding to the imaged inspection target surface, and binarizes the image signal output from the imaging unit and converts the image signal into binary image data And a plurality of main scanning lines set at substantially equal intervals in a predetermined sub-scanning direction in the binary image of the inspection target surface represented by the binary image data generated by the converting unit. Scanning detecting means for scanning and detecting a marking candidate line which is a main scanning line including pixels of binary image data having a predetermined value; and detecting a continuous number of marking candidate lines detected by the scanning detecting means. Identification for identifying whether a mark is applied or a defective portion is formed in the area of the continuous marking candidate line based on the number of consecutive marking candidate lines detected by the line number detecting means and the number of marking candidate lines detected by the line number detecting means. Means for visual inspection.

According to the present invention, a plurality of main scanning lines set at substantially equal intervals in a predetermined sub-scanning direction are sequentially scanned in a binary image of a surface to be inspected, and binary image data of a predetermined value is obtained. By detecting an engraving candidate line which is a main scanning line including pixels, and further detecting the continuous number of the engraving candidate line, the area of the continuous engraving candidate line is engraved based on the detected number of consecutive engraving candidate lines. Can be determined as to whether or not a defect has occurred. Therefore, unlike the pattern search in which the shape of each region represented by a black pixel is recognized and the region having a shape other than the shape of the mark actually applied to the inspection target surface is determined as a defective portion, the pattern search is performed using the black pixel. Since it is not necessary to recognize the shape of each region represented, the inspection time can be greatly reduced. Further, even when the quality of the marking is low, it is possible to satisfactorily distinguish between the marking and the defective portion.

If the number of lines detected by the line number detecting means is within a predetermined numerical range, the identifying means determines that the area detected by the scanning detecting means is marked. If the number of lines detected by the line number detecting means is out of the predetermined numerical range, it is preferable that it is determined that a defective portion has occurred in the area detected by the scanning detecting means. .

Preferably, the main scanning line is a line substantially perpendicular to the sub-scanning direction along an erect direction (main scanning direction) of the mark on the inspection target surface. For example, when a plurality of inscriptions are formed side by side on an arc-shaped inspection target surface, the individual inscriptions are provided along the radial direction of the arc-shaped inspection target surface, and thus the arc-shaped inspection target surface is formed. Is the main scanning direction, and the circumferential direction substantially orthogonal to the main scanning direction is the sub-scanning direction.

According to a second aspect of the present invention, the inspection is performed by storing the number of inscriptions to be actually applied to the inspection target surface, and counting the number of inscriptions identified by the identification means. The number of markings detecting means for detecting the number of markings applied to the entire area of the target surface, and comparing the number of markings stored in the storage means with the number of markings detected by the number of markings detecting means 2. The appearance inspection apparatus according to claim 1, further comprising a quality judgment unit for judging whether the inspection target surface is a non-defective product surface or a defective product surface.

According to the present invention, the number of stamps stored in the storage unit is compared with the number of stamps detected by the stamp number detecting unit. If the numbers are not the same, the inspection target surface is determined to be defective. Thus, even if the identification unit erroneously determines that the defective portion is a stamp, there is no possibility that the quality of the inspection target surface is erroneously determined.

According to a third aspect of the present invention, the conversion means includes:
The grayscale image of the inspection target surface obtained based on the image signal output from the imaging unit was divided into a plurality of regions, and image signals corresponding to pixels included in the plurality of regions were set for each region. 3. The visual inspection apparatus according to claim 1, wherein the image signal output from the image pickup means is converted into binary image data by binarizing the image data with a threshold value.

According to the present invention, even when unevenness of illumination or the like occurs on the inspection target surface, the grayscale image of the inspection target surface is divided into a plurality of regions, and each region is appropriately determined. By binarizing the image signal output from the imaging unit based on the threshold value, a good binary image of the inspection target surface can be obtained. Thereby, the quality of the inspection target surface can be more accurately determined.

According to a fourth aspect of the present invention, there is provided an appearance inspection apparatus for inspecting the appearance of an object having a surface to be inspected,
An imaging unit that images the inspection target surface and outputs an image signal corresponding to the imaged inspection target surface; and a plurality of grayscale images of the inspection target surface obtained based on the image signal output from the imaging unit. Converting means for dividing the image signal corresponding to the pixels included in the plurality of areas into binarized data by using a threshold value set for each area and converting the image signal into binary image data; An appearance inspection apparatus characterized by including a pass / fail determination unit that determines whether the inspection target surface is a non-defective surface or a defective surface based on the converted binary image data.

According to the present invention, even when unevenness of illumination or the like occurs on the surface to be inspected, the grayscale image of the surface to be inspected is divided into a plurality of regions, and each region is appropriately determined. By binarizing the image signal output from the imaging unit based on the threshold value, good binary image data corresponding to the inspection target surface can be obtained. This allows
The quality of the inspection target surface can be more accurately determined.

According to a fifth aspect of the present invention, there is provided an appearance inspection method for inspecting the appearance of an object having an engraved inspection object surface, wherein the inspection object surface is imaged and the imaged inspection object surface is taken. , An image signal acquired in the imaging step, a conversion step of binarizing the image signal and converting the image signal into binary image data, and an image signal represented by the converted binary image data. Inspection surface 2
In the value image, a plurality of main scanning lines set at substantially equal intervals in a predetermined sub-scanning direction are sequentially scanned to detect a marking candidate line which is a main scanning line including pixels of binary image data having a predetermined value. A scanning detection step, a line number detection step for detecting the continuous number of the marking candidate lines detected in the scanning detection step, and the continuous number of the marking candidate lines detected in the line number detecting step. An identification step of identifying whether an inscription or a defective portion has occurred in the area of the inscription candidate line.

According to this method, the same effect as that of the first aspect can be obtained. According to a sixth aspect of the present invention, in the inspection, the number of inscriptions to be actually applied to the surface to be inspected is stored in a storage means, and the number of inscriptions identified in the identification step is counted. An engraving number detecting step for detecting the number of engravings applied to the entire area of the target surface, and comparing the number of engravings detected in the engraving number detecting step with the number of engravings stored in the storage means. 6. The appearance inspection method according to claim 5, further comprising: a quality determination step of determining whether the inspection target surface is a non-defective product surface or a defective product surface.

According to this method, the same effect as that of the second aspect can be obtained. According to a seventh aspect of the present invention, in the conversion step, the grayscale image of the inspection target surface obtained based on the image signal obtained in the imaging step is divided into a plurality of regions, and pixels included in the plurality of regions are divided. The image signal obtained in the imaging step is converted into binary image data by binarizing the image signal corresponding to the image data with a threshold value set for each region. An appearance inspection method according to 5 or 6.

According to this method, the same effect as that of the third aspect can be obtained. The invention according to claim 8 is an appearance inspection method for inspecting the appearance of an object having an inspection target surface, wherein the inspection target surface is imaged, and an image signal corresponding to the imaged inspection target surface is acquired. An imaging step and dividing the grayscale image of the inspection target surface obtained based on the image signal acquired in the imaging step into a plurality of regions, and converting an image signal corresponding to a pixel included in the plurality of regions into each region. A conversion step of binarizing the image data with a threshold value set for each of them, and converting the image data into binary image data; and, based on the converted binary image data, determining whether the inspection target surface is a non-defective surface or a defective surface. A quality determination step of determining whether or not there is a visual inspection.

According to this method, the same effect as that of the fourth aspect can be obtained.

[0023]

Embodiments of the present invention will be described below in detail with reference to the accompanying drawings. FIG. 1 is a cross-sectional view showing a simplified configuration of a visual inspection apparatus according to an embodiment of the present invention. This visual inspection device is, for example,
This is a device for inspecting the end face RA of the outer race R of the bearing, and has a holder 1 for holding the outer race R.

The holder 1 has a recess 11 into which the outer race R can be fitted.
Is fitted into the recess 11 with the end face upward, and the end face RA is held in a state of being substantially along the horizontal plane. An illumination optical system 2 for illuminating the end face RA of the outer race R held by the holder 1 is disposed above the holder 1. The illumination optical system 2 includes, for example, a plurality of LEDs.
Are arranged in an annular shape, and a light source 21
And a diffusing plate 22 for diffusing the illumination light from the outside and guiding it to the end surface RA of the outer race R held by the holder 1.

Further above the illumination optical system 2, a holder 1
On the central axis of the outer race R mounted on the camera, a CCD camera 3 that photoelectrically converts reflected light from the end surface RA of the outer race R and outputs an image signal is disposed. This CCD camera 3
Is supplied to an image processing unit 4 composed of a microcomputer including a CPU, a ROM and a RAM, for example.

The image processing section 4 includes a conversion processing section 41 and a pass / fail judgment section 42. First, the image signal given from the CCD camera 3 to the image processing unit 4 is first converted by the conversion processing unit 4.
1 is input. The conversion processing unit 41 converts the input image signal into binary image data by binarizing the input image signal with a predetermined threshold, and converts the converted binary image data into the image processing unit 4. Stored in the RAM. Pass / fail judgment unit 4
2 is based on the binary image data stored in the RAM,
It is determined whether the end surface RA of the outer race R is a non-defective product surface or a defective product surface.

Further, the image processing section 4 displays a binary image of the end face RA of the outer race R on the monitor 5 based on the binary image data stored in the RAM. Thus, an annular black and white image representing the end surface RA of the outer race R is displayed on the monitor 5. FIG. 2 is a flowchart illustrating the flow of the pass / fail determination process executed by the pass / fail determination unit 42. FIG. 3 is a diagram for explaining the pass / fail determination processing shown in FIG. In this pass / fail determination process, the end surface RA of the outer race R which is engraved with, for example, the name of the manufacturing country, the name of the manufacturer, and the manufacturing model number has a defective portion such as a scratch, rust or black scale residue. This is a process for determining whether the surface is a non-defective surface or a defective surface having a defective portion.

In the following description, the scanning line S
L1 to SLn are lines along the radial direction (main scanning direction) of the outer race R in the binary image of the end face RA of the outer race R represented by the binary image data, as shown in FIG. Say. The scanning lines SL1 to SLn are
Are set over the entire circumference of the end face RA at predetermined intervals (for example, 0.5 ° intervals) in the circumferential direction (sub-scanning direction).

When the pass / fail judgment process is started, first, a scan (radial scan) of a scan line SL1 at a predetermined rotational direction position is performed (step S1).
It is determined whether all the pixels included in L1 are white pixels (step S2). Specifically, the scanning line S
The binary image data of the pixels included in L1 is read from the RAM, and it is determined whether or not all the read binary image data is data representing white pixels. In other words, it is determined whether or not the binary image data read from the RAM includes data representing a black pixel.

The first scanning line SL1 is scanned.
Is always set to a scanning line SL that does not include black pixels. For example, when the scanning line SL arbitrarily extracted is scanned and a black pixel is included in the scanning line SL, the scanning line SL is not set to the scanning line SL1, but is set to the scanning line SL. Scans different scan lines SL. Then, it is checked whether or not the scan line SL includes a black pixel. If the scan line SL also includes a black pixel, another scan line SL is scanned. A scan line SL that does not include black pixels is found, and the found scan line SL is set as the scan line SL1.
Therefore, at this time, the determination as to whether all the pixels included in the scanning line SL1 are white pixels is affirmed.

Next, for example, the count value L of a line number counter provided in the RAM is examined, and it is determined whether or not this count value L satisfies a condition Lmin ≦ L ≦ Lmax (Lmin and Lmax are natural numbers). Is determined (step S
3). The above numerical values Lmin and Lmax are respectively the outer raceway R
Is set based on the circumferential width of the marking actually applied to the end face RA, that is, the minimum value and the maximum value of the number of scanning lines SL straddling the marking, for example, as shown in FIG. Is applied, the numerical value Lmin is set to "3 (the number of scanning lines SL straddling the engraving" N ")", and the numerical value Lmax is set to "5 (scanning straddling the engraving" A "). Line SL).
Further, the line number counter is reset to “0” at the start of the pass / fail determination processing. Therefore, at this time,
The determination whether the count value L of the line number counter satisfies the above condition is denied.

If the determination as to whether or not the count value L satisfies the above condition is denied, it is determined whether or not the count value L is "0" (step S4). Count value L
Is "0", then all the scanning lines SL1 to SL1
It is determined whether or not SLn has been scanned (step S).
5). If all the scanning lines SL1 to SLn have not been scanned, the scanning of the scanning line SL2 adjacent to the scanning line SL1 is performed (step S6).
The process returns to step S2, and it is determined whether or not all the pixels included in the scanning line SL2 are white pixels. After that, until the radial line SL including the black pixel is detected,
The processing of steps S2 to S6 described above is repeated, and the scanning lines SL after the scanning line SL2 are sequentially scanned one by one.

For example, in the pass / fail judgment of the end face RA corresponding to the binary image shown in FIG. 3A, the scanning line SL7 including the black pixels is engraved while the scanning line SL is sequentially scanned one by one. It is detected as a candidate line (NO in step S2). When the scanning line SL7 is detected, the count value L of the line number counter is incremented (+1) (step S7). Then, the process proceeds to step S5, and after determining that all the scanning lines SL1 to SLn have been scanned is denied, scanning of the next scanning line SL8 is performed (step S6). Then,
It is determined whether all the pixels included in the scan line SL8 are white pixels (step S2). In the binary image shown in FIG. 3A, since the scanning lines SL7 to SL10 include black pixels, the above-described steps S1 to S4 until a scanning line SL11 containing no black pixels is detected in step S2.
Steps S2, S7, S5, and S6 are repeated.

When the scanning line SL11 containing no black pixel is detected (YES in step S2), the count value L of the line number counter is checked, and the count value L satisfies the above condition Lmin ≦ L ≦ Lmax. It is determined whether or not there is (step S3). At this time, steps S2 and S
The processing of steps S7, S5, and S6 is repeated, and scanning lines SL7 to SL10 including black pixels are detected as marking candidate lines, so that the count value L of the line number counter is "4". The value L is the above condition Lmi
It is determined that n ≦ L ≦ Lmax is satisfied (YES in step S3). As a result, it is determined that the marking is performed on the scanning lines SL7 to SL10 detected as the marking candidate lines, and the count value N of the marking number counter provided in the RAM is incremented (+1) (for example). Step S8). Thereafter, the count value L of the line number counter is cleared to "0" (step S9),
Proceeding to step S5, it is determined whether all the scanning lines SL have been scanned. At this time, since scanning of all the scanning lines SL has not been completed, the process proceeds to step S6, and scanning of the scanning lines SL after the scanning line SL11 is performed.

As described above, the scanning lines SL are sequentially scanned one by one, and the engraved marks on the end face RA are detected one after another. Then, all the scanning lines SL1 to SL1
When scanning of SLn is performed (YES in step S5),
The count value N of the stamp number counter is checked, and it is determined whether or not this count value N is "0" (step S).
10). If the count value N is not “0” (step S
Next, it is determined whether or not the count value N matches the number of stamps actually applied to the end face RA of the outer race R (hereinafter, referred to as “actual stamp number”). (Step S11). The actual engraved number is input before the start of the pass / fail judgment processing, and is stored in, for example, a RAM.

If the count value N matches the actual engraved number (YES in step S11), the end face R of the outer race R
A is determined to be a non-defective product (step S12). If the count value N does not match the actual engraved number (NO in step S11), it is determined that the engraved part has been erroneously identified as a defective portion occurring on the end face RA of the outer race R, and this end face RA is It is determined that the surface is defective (Step S1)
3).

For example, in the binary image shown in FIG.
A defective portion D1 exists on the four scanning lines SL2 to SL5. Therefore, in the pass / fail determination of the end face RA corresponding to the binary image shown in FIG. 3B, it is determined in step S3 that the count value L of the line number counter satisfies the above condition Lmin ≦ L ≦ Lmax. Thus, the defective portion D1 on the scanning lines SL2 to SL5 is erroneously determined to be a mark. Then, the count value N of the engraved number counter is incremented (+1), and as a result,
When the scanning of all the scanning lines SL1 to SLn is completed, the count value N of the marking number counter becomes larger than the actual marking number. Therefore, when the count value N does not match the actual engraved number, the end surface RA is determined to be a defective surface, and a relatively large defective portion such as the defective portion D1 is generated on the end surface RA. In addition, it is possible to eliminate the possibility that the end surface RA is erroneously determined to be a good surface.

Further, all the scanning lines SL1 to SLn
If the count value N is "0" at the time when the scanning is completed, it is determined that the end face RA of the outer race R is a non-marked face on which no marking is performed (step S14). On the other hand, for example, in the pass / fail judgment of the end face RA corresponding to the binary image shown in FIG. 3C, the scanning line SL5 including the black pixel is used as the marking candidate in the process of sequentially scanning the scanning line SL line by line. It is detected as a line (NO in step S2). When the scanning line SL5 is detected, the count value L of the line number counter is incremented (+1).
Is performed (step S7). Then, the process proceeds to step S5, where it is determined whether or not all the scanning lines SL1 to SLn have been scanned. All scan lines SL1 to SLn
Is not performed, the next scanning line SL
6 are performed (step S6), and the process returns to step S2 to determine whether all the pixels included in the scan line SL6 are white pixels.

In the binary image shown in FIG. 3C, since the scanning line SL6 does not include a black pixel, step S2
It is determined that all the pixels included in the scan line SL6 are white pixels, and then it is determined whether or not the count value L of the line number counter satisfies the above condition Lmin ≦ L ≦ Lmax (step S3). At this time, since the count value L of the line number counter is “1”, it is determined that the count value L does not satisfy the above condition Lmin ≦ L ≦ Lmax (NO in step S3), and the count value L is “0”.
Is determined (step S4). And
The determination as to whether or not the count value L is "0" is also denied, whereby it is determined that a defective portion D2 has occurred on the scanning line SL5 including the black pixel detected as the marking candidate line, and The end surface RA of the bearing ring R is determined to be a defective surface (step S13).

As described above, in this embodiment, the end surface RA of the outer race R is imaged by the CCD camera 3, and
The image signal output from the D camera 3 is converted into binary image data by binarization. And this 2
Scanning along the radial direction of the outer race R in the binary image of the end face RA represented by the value image data (radial scanning)
Is a predetermined interval in the circumferential direction (for example, 0.5 ° interval)
Is performed over the entire circumference of the end face RA. In this process, when a marking candidate line that is a main scanning line including a black pixel is detected, the number L of consecutive marking marking lines is checked.
Then, this line continuation number L is equal to the above condition Lmin ≦ L ≦ L
If max is satisfied, it is determined that the continuous marking candidate line area has been marked. On the other hand, if the line continuation number L does not satisfy the above condition Lmin ≦ L ≦ Lmax, it is determined that a defective portion has occurred in that area, and the end surface RA is determined to be a defective surface.

Therefore, unlike the pattern search in which the shape of each area represented by black pixels is recognized and an area having a shape other than the shape of the mark actually applied to the inspection target surface is determined as a defective portion, Since there is no need to recognize the shape of each region represented by black pixels, the inspection time can be significantly reduced. Also, even if the quality of the engraving is low,
It is possible to satisfactorily distinguish between the stamp and the defective portion.

In this embodiment, if the number of areas determined to be engraved matches the actual engraved number at the time when scanning of all the scanning lines SL1 to SLn is completed, the outer raceway R
Is determined to be a non-defective surface, and if they do not match, the end surface RA of the outer race R is determined to be a defective surface.
Accordingly, even if the defective portion is relatively large, and the mark is erroneously determined to be a stamp, there is no possibility that the quality of the end face RA is erroneously determined.

The quality of the inspection target surface is determined by determining whether or not the number of regions determined to be engraved matches the actual engraved number, and determining whether the inspection target surface is good or not. It need not be limited. For example, in the pass / fail judgment of the inspection target surface shown in FIG. 4, the number of the marking lines actually applied to the inspection target surface (in this case, the number of the marking lines = 3) and the number of the marking lines included in each of the marking lines are determined. Number (in this case, the number of stamps on the stamp row S1 = 5, the number of stamps on the stamp row S2 = 5, the number of stamps on the stamp row S3 =
3) is input in advance. Then, the scanning line SL1
In the scanning process of ~ SLn, if the interval between the area determined to be engraved and the area next determined to be engraved is smaller than the predetermined interval, it is determined that those areas are a series of engraved rows,
The number of stamp lines and the number of stamps included in each stamp line are detected. When the scanning of all the scanning lines SL1 to SLn is completed, the detected number of engraved lines and the number of engraved lines of each engraved line coincide with the previously input number of engraved lines and the number of engraved lines of each engraved line, respectively. Then, the inspection target surface may be determined to be a non-defective product, and if they do not match, the inspection target surface may be determined to be a defective product surface.

In this embodiment, the end surface RA of the outer race R of the bearing is set as the inspection target surface.
The outer peripheral surface or the inner peripheral surface of the outer race R may be the inspection target surface. When the outer peripheral surface or the inner peripheral surface of the outer race R is the inspection target surface, the appearance inspection device shown in FIG. 5 or FIG. 6 may be used. FIG. 5 shows the outer race R
FIG. 2 is a simplified cross-sectional view showing a configuration of a visual inspection device for inspecting an outer peripheral surface of the device. In FIG. 5, portions corresponding to the respective portions in FIG. 1 described above are denoted by the same reference numerals as in FIG.

In this appearance inspection apparatus, a reflection mirror 6 having an inverted conical reflection surface 61 is arranged so as to surround the side of the holding position of the outer race R by the holder 1.
As a result, the illumination light from the illumination optical system 2 is reflected by the reflection surface 61 of the reflection mirror 6 and guided to the outer peripheral surface RO of the outer race R held by the holder 1. The light reflected by the outer peripheral surface RO of the outer race R is further reflected by the reflection surface 61 of the reflection mirror 6 to become scattered light, and a part of the scattered light is captured by the CCD camera 3. An annular black and white image representing the outer circumferential surface RO of the outer race R is displayed on the monitor 5.

FIG. 6 is a cross-sectional view showing a simplified configuration of an appearance inspection device for inspecting the inner peripheral surface of the outer race R. In FIG. 6, portions corresponding to the respective portions in FIG. 1 described above are denoted by the same reference numerals as in FIG. In this visual inspection device, the illumination optical system 2 is
The inner peripheral surface RI of the outer race R held by the holder 1 is illuminated by the illumination light from the illumination optical system 2. Then, the light reflected on the inner peripheral surface RI of the outer race R is captured by the CCD camera 3, and the monitor 5 displays the inner peripheral surface R of the outer race R on the monitor 5.
An annular black-and-white image representing I is projected.

FIG. 7 is a diagram for explaining a further preferred embodiment of the present invention. For example, when the area of the inspection target surface is relatively large, there is a possibility that the illumination optical system 2 may cause uneven illumination on the inspection target surface. Further, in the appearance inspection device for inspecting the outer peripheral surface RO of the outer race R, the vicinity of the lower end of the outer peripheral surface RO is shaded as shown in FIG. In the grayscale image of the outer peripheral surface RO obtained based on the above, the density near the inner circumference becomes higher than the density near the outer circumference. Therefore, when the image signal output from the CCD camera 3 is binarized with a certain threshold value, the outer peripheral surface R
Pixels near the inner periphery of the O grayscale image are determined as black pixels, and there is a possibility that the outer peripheral surface RO is determined to be a defective product surface in the above-described quality determination processing.

Therefore, as shown in FIG.
The O grayscale image is divided into an area A1 near the inner circumference and an area A2 near the outer circumference, and an appropriate binarization threshold is set for each of the areas A1 and A2. Then, the image signals corresponding to the pixels included in each of the areas A1 and A2 are converted into binary image data by binarizing the image signals with binarization thresholds individually set. Thereby, the binarization of the image signal output from the CCD camera 3 can be favorably performed, and the quality determination of the outer peripheral surface RO in the quality determination processing can be performed more accurately.

In this embodiment, the grayscale image of the outer peripheral surface RO is divided into two regions A1 and A2 and binarized. However, the grayscale image of the outer peripheral surface RO is divided into three or more regions. The image signal output from the CCD camera 3 may be binarized based on a threshold value set for each of the divided areas. Although some embodiments of the present invention have been described above, the present invention is not limited to the above embodiments. For example, in the above-described embodiment, an example is described in which the end surface, the outer peripheral surface, or the inner peripheral surface of the outer race of the bearing is set as the inspection target surface. However, the present invention is not limited to the outer race of the bearing. The end surface, the outer peripheral surface, or the inner peripheral surface of the ring may be the inspection target surface.
Further, the end surface, the outer peripheral surface, or the inner peripheral surface of the completed bearing may be the inspection target surface.

Furthermore, the surface to be inspected is not limited to an annular surface such as an end surface of a bearing, but any surface of any shape can be used as the surface to be inspected as long as it can be imaged as a two-dimensional image by the imaging means. In addition, various design changes can be made within the scope of the matters described in the claims.

[Brief description of the drawings]

FIG. 1 is a simplified cross-sectional view showing a configuration of a visual inspection apparatus according to an embodiment of the present invention.

FIG. 2 is a flowchart illustrating a flow of a pass / fail determination process.

FIG. 3 is a diagram for explaining pass / fail judgment processing;

FIG. 4 is a diagram for explaining another pass / fail judgment method.

FIG. 5 is a simplified cross-sectional view showing a configuration of an appearance inspection device for inspecting an outer peripheral surface of an outer race.

FIG. 6 is a simplified cross-sectional view showing a configuration of an appearance inspection device for inspecting an inner peripheral surface of an outer race.

FIG. 7 is a diagram for explaining a further preferred embodiment of the present invention.

[Explanation of symbols]

3 CCD camera 4 Image processing section 41 Conversion processing section (conversion means) 42 Pass / fail judgment section (scanning detection means, line number detection means, identification means, storage means, stamp number detection means, pass / fail judgment means) A1, A2 area D1, D2 Defective part R Outer ring (object) RA End surface (surface to be inspected) RI Inner surface (surface to be inspected) RO Outer surface (surface to be inspected) SL1 to SLn Scan line (main scanning line)

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Kenshi Sakamoto 3-5-8 Minamisenba, Chuo-ku, Osaka-shi Inside Koyo Seiko Co., Ltd. (72) Inventor Kazuaki Yamada 3-58-8 Minamisenba, Chuo-ku, Osaka-shi F-term (reference) in Seiko Co., Ltd. BA02 CA08 CA12 CA16 CB06 CB12 CB16 CC01 CC03 DA03 DA08 DA13 DB02 DB08 5L096 AA06 AA07 BA03 CA02 DA03 EA43 FA52 FA53 FA73 GA28 GA51 JA28

Claims (8)

[Claims] An appearance inspection device for inspecting the appearance of an object having an engraved inspection target surface, wherein the inspection target surface is imaged, and an image signal corresponding to the imaged inspection target surface is generated. Imaging means for outputting; converting means for binarizing an image signal output from the imaging means to convert the image signal into binary image data; and the inspection object surface represented by the binary image data generated by the conversion means In the binary image, a plurality of main scanning lines set at substantially equal intervals in a predetermined sub-scanning direction are sequentially scanned, and a marking candidate line which is a main scanning line including a pixel of binary image data having a predetermined value. Scanning detecting means for detecting the number of lines of engraving candidate lines detected by the scanning detecting means, and the number of lines of engraving candidate lines detected by the line number detecting means. Based on the number, the area of the continuous stamping candidate lines, the appearance inspection apparatus characterized by comprising identifying means for identifying whether stamped Do defect has been subjected has occurred. 2. A storage means for storing the number of inscriptions to be actually applied to the inspection object surface, and counting the number of inscriptions identified by the identification means, and applying the number to the entire area of the inspection object surface. Comparing the number of stamps stored in the storage unit with the number of stamps detected by the stamp number detection unit, and comparing the number of stamps detected by the stamp number detection unit with the number of stamps detected by the stamp detection unit. 2. The visual inspection apparatus according to claim 1, further comprising a quality judgment unit for judging whether the surface is a non-defective product or a defective product. 3. The conversion means divides a grayscale image of the inspection target surface obtained based on an image signal output from the imaging means into a plurality of regions, and corresponds to pixels included in the plurality of regions. 2. An image signal output from the image pickup means is converted into binary image data by binarizing the image signal with a threshold value set for each region. Or the visual inspection device according to 2. 4. An appearance inspection apparatus for inspecting the appearance of an object having an inspection target surface, comprising: an imaging unit which images the inspection target surface and outputs an image signal corresponding to the imaged inspection target surface. Dividing the grayscale image of the inspection target surface obtained based on the image signal output from the imaging means into a plurality of regions, and setting an image signal corresponding to a pixel included in the plurality of regions for each region; Conversion means for converting the image data into binary image data by converting the image data into binary image data based on the threshold values, and converting the inspection target surface into a non-defective product surface or a defective product surface based on the binary image data converted by the conversion device. A visual quality inspection device, comprising: a pass / fail judgment means for judging whether or not there is an object. 5. An appearance inspection method for inspecting an appearance of an object having an engraved inspection target surface, wherein the inspection target surface is imaged, and an image signal corresponding to the imaged inspection target surface is obtained. An image capturing step to be obtained, and an image signal obtained in this image capturing step is binarized to 2
A plurality of main scanning lines set at substantially equal intervals in a predetermined sub-scanning direction in a binary image of the inspection target surface represented by the converted binary image data. In order to detect a marking candidate line which is a main scanning line including pixels of binary image data having a predetermined value, and a continuous number of marking candidate lines detected in the scanning detection step. Based on the line number detecting step to be performed and the number of consecutive marking candidate lines detected in the line number detecting step, it is determined whether the area of the continuous marking candidate line has been marked or has a defective portion. And an identification step. 6. A storage step of storing in a storage means the number of inscriptions to be actually performed on the inspection target surface, and counting the number of inscriptions identified in the identification step, and An engraving number detecting step of detecting the number of engravings applied to the entire area; and comparing the number of engravings stored in the storage means with the number of engravings detected in the engraving number detecting step. 6. The appearance inspection method according to claim 5, further comprising: a quality determination step of determining whether the target surface is a non-defective surface or a defective surface. 7. The converting step divides a grayscale image of the inspection target surface obtained based on the image signal obtained in the imaging step into a plurality of areas, and corresponds to pixels included in the plurality of areas. 7. The image signal obtained in the imaging step is converted into binary image data by binarizing the image signal with a threshold value set for each region. Visual inspection method described. 8. An appearance inspection method for inspecting the appearance of an object having an inspection target surface, comprising: an imaging step of imaging the inspection target surface and acquiring an image signal corresponding to the imaged inspection target surface; Dividing the grayscale image of the inspection target surface obtained based on the image signal obtained in this imaging step into a plurality of regions,
Image signals corresponding to the pixels included in the plurality of regions are
Binarize with the threshold set for each area
A converting step of converting the image data into value image data; and a quality determining step of determining whether the inspection target surface is a non-defective product surface or a defective product surface based on the converted binary image data. Appearance inspection method.