JPH08334478A - Seal inspection system - Google Patents

Abstract

PURPOSE: To inspect the applied state of a sealant from the condition of its image. CONSTITUTION: This seal inspection system includes a means 12 for picking up image of a work joining face 40 to which a sealant 41 is applied, an image memory 17 in which data from the image-picking up means 12 is stored as image data, and a processing device 25 for performing variable density image process on the image data. The processing device has the function of specifying a portion for inspection, by which that image position on an inspection width which corresponds to the sealant application position is inspected, and has the function of judging, from picture elements from each end of the inspection width to an inspection point within the image data, the positions of those picture elements at which density variation is at or above a set value to be both ends (edges) of the sealant along its cross direction. Further, the processing device has a function of determining cut-line errors, by which, if both ends of the sealant are not detected, the inspection point is judged to be the cut line of the sealant.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a seal inspection system, and more particularly to an improvement of a seal inspection system for determining a coating state of a sealing material coated in a bead shape by using an automatic seal coating device or the like.

[0002]

2. Description of the Related Art Recently, a sealing method using a silicone resin or the like has been adopted for sealing between various mechanical parts.

Particularly, in the process of assembling an engine or the like, it can be easily applied to an assembling line and the working environment can be improved. Therefore, from a nozzle provided so as to be movable along a joint surface of a work. A device that discharges a liquid sealing material and automatically applies the sealing material to the bonding surface has come into widespread use.

In the coating apparatus described above, when air or the like is mixed in the seal material for some reason, air is ejected when the seal material is discharged from the nozzle,
There is a disadvantage that a so-called blank ejection state is partially obstructed to continuously discharge the sealing material, or a partial abnormality of the coating width occurs.

Therefore, after the automatic application, it is necessary to inspect the state of the sealing material applied to the joint surface. This inspection requires visual observation by an operator, and C
A coating inspection by image processing using an image pickup means such as a CD camera has also been adopted.

[0006]

However, since coating abnormalities or defects rarely occur, it is not possible to detect this when an actual defect occurs due to the belief that there is no defect in the visual inspection. It is easy to miss the case.

Further, in the visual inspection, since the abnormality of the coating is judged by the inspector's subjectivity, it is difficult to make uniform the criterion for judging whether the coating state is normal or abnormal. That is, in the visual inspection, a coating state that is judged to be abnormal by one inspector may be determined to be normal by another inspector, and the evaluation of the coating state of the sealing material cannot be quantified.

On the other hand, in the inspection by the image processing, the density value in the photographed image tends to vary, and in the inspection in which the image processing is performed on the binarized image data, it is difficult to secure the accuracy of the inspection result. There was an inconvenience. This will be described in more detail. An image obtained when there is no irregular reflection of illumination at the time of shooting is as shown in FIG.
The seal material 41 can be clearly identified with respect to zero. This is because the bonding surface 40 is normally imaged brightly, whereas the sealing material is imaged darker than this brightness.

Therefore, when the image shown in FIG. 4A is binarized by a predetermined threshold value, the joining surface 40 and the sealing material 41 are well separated as shown in FIG. 4B. Therefore, it is possible to capture the sealing material 41 on the joint surface 40. Further, by counting the number of pixels in the dotted frame in the figure and comparing the number of pixels with a set value, the area where the sealing material is applied can be captured and the application state can be inspected.

However, in the field where the sealing material is applied, it is a part of the manufacturing process, and the installation position of the illumination is also limited. Therefore, the inspection environment such as the ambient brightness is not necessarily good. Therefore, in actual shooting,
As shown in FIG. 4C, the illumination may be reflected by a part of the seal material 41 to be bright, and the brightness of a part of the seal material 41 and the joint surface 40 may be similar. Therefore, when binarizing the image shown in FIG.
As shown in (D), the portion separated as the sealing material 41 is smaller than the actual sealing material 41 portion, and even if the number of pixels surrounded by the dotted line in the figure is counted, the number of pixels is less than the preset number. It is determined to be defective.

That is, depending on the inspection environment, the reflection of the illumination occurs on the sealing material 41, and the reflection of the sealing material portion causes an erroneous determination of the coating state. Therefore, when the bonding surface 40 is imaged in an ideal state, the coating state can be determined by image processing, but in the actual manufacturing process, the sealing material 41 is reflected, so that automatic inspection without visual inspection is performed. However, the reliability was low.

Further, in the method of judging the presence or absence of the inspection object based on the number of pixels in the window of the binarized image as shown in FIG. 4, it is difficult to set the threshold value for binarization. This causes the inconvenience that the threshold value that works well in a certain inspection environment does not separate the sealing material portion and the joint surface portion in another environment. Further, it is too complicated to set the threshold value every time the illumination position or the intensity is changed, which hinders the automation of the inspection of the coating state of the sealing material by the image processing. Further, the automatic setting of the threshold value by the discriminant analysis method requires a long time for the processing.

As described above, in the inspection system for the coating state of the sealing material, even if the density in the image data is not uniform and slightly changes due to the illumination position and the illuminance changing depending on the inspection environment, It is desirable to have a method capable of separating the sealing material portion in the image data even if the external light is reflected in a part of the area and has the same brightness as the joint surface.

It has been found that a portion where the sealing material cannot be identified due to reflection actually tends to occur in the central portion along the coating direction of the sealing material. That is, due to the thickness of the sealing material and the like, both end portions in the width direction of the sealing material 41 are less likely to be reflected by the illumination and are imaged darkly. That is, the sealing material 41
While the density of the central part of the is easily affected by the illumination and is unstable, the density difference between the both ends is stable with respect to the bonding surface 40. Therefore, the concentration at both ends of the sealing material in the width direction can be clearly identified with respect to the concentration at the joint surface.

[0015]

It is an object of the present invention to detect the presence / absence of a sealing material on the basis of changes in the concentration distribution along the widthwise ends of the sealing material in the image data of the workpiece joint surface, and this detection is predetermined. Another object of the present invention is to provide a seal inspection system capable of performing the coating state of the sealing material at high speed and with high accuracy by performing the inspection at each inspection point.

[0016]

In order to achieve the above-mentioned object, the present invention provides a means for picking up an image of a work joint surface coated with a sealing material, and an image for storing the data taken in by the image pick-up means as image data. A memory and a processing device that is connected to the image memory and that performs grayscale image processing on the image data are provided, and the processing device further includes an inspection width portion centered on a predetermined inspection point on the image data. The inspection portion specifying function of searching for the pixel position of, and the position of the pixel from the both ends of the inspection width in the image data to the inspection point where the density change value is equal to or more than the set value is the both ends in the width direction of the sealing material. Both-sides detection function that determines the position, and a break that determines that the inspection point is a seal break when both ends in the width direction are not detected by the both-sides detection function And a color determination function, and employs a configuration that.

Further, the processing device in the seal inspection system has a coating state judging function for judging that the coating state is defective when a break error continuously occurs at a plurality of inspection points. Further, it is preferable to adopt a configuration in which the inspection point is set at the center in the width direction of the normally applied sealing material and both ends of the inspection width are set on the work joining surface.

Further, the processing apparatus calculates the width of the sealing material at the inspection point based on the widthwise both end positions detected by the both side detecting function, and the coating width calculating function. In addition, when the coating width is smaller than a predetermined minimum set width, a coating width error determination function for determining a coating width error at the inspection point is provided.

Further, the processing apparatus is provided with a coating state judging function for judging that the coating state is defective when a coating width error continuously occurs at a plurality of inspection points.

Further, the processing apparatus uses a deviation amount calculation function for judging the deviation amount of the coating position based on the both end positions in the width direction detected by the both side detection function and the position of the inspection point, and this deviation amount calculation function. It is also possible to employ a configuration in which a deviation error determination function for determining a deviation error at the inspection point is provided when the calculated deviation amount is larger than a predetermined set deviation amount, whereby various variations of the coating state can be individually adopted. Can be inspected. At this time, the processing apparatus may be configured to have a coating state determination function that determines that the coating state is defective when the coating width error continuously occurs at a plurality of inspection points.

[0021]

Before performing the grayscale image processing, the inspection points are taught by using the reference work on which the sealing material is normally applied, and the inspection points are stored in the processing device. After that, the image data is captured by the image pickup means with respect to the inspection object having the seal material applied to the work joining surface, and the image data is stored in the image memory. The processing device performs grayscale image processing on the image data in the image memory in the seal width direction for each inspection point. Here, whether or not both ends of the seal material have been detected is determined by comparing the density of the work joint surface and the density of both ends of the seal material. The density difference can be set based on the color of the work joint surface and the seal material, or the reflectance of light, and the presence or absence of both ends of the seal can be determined based on the set value of the density difference.

When both ends of the sealing material at a specific inspection point cannot be detected, the seal breakage at the inspection point is counted. Then, when the both ends of the seal material cannot be detected at the next inspection point, the count of seal breakage is added up. In this way, when the count number of consecutive seal breaks exceeds the allowable set number of times, it is considered as a seal break and it is determined that the application is abnormal.

In addition, in the structure in which the processing device is provided with the function of detecting the minimum set width of the seal material, it is possible to detect the seal application width and the width deviation at each inspection point at the same time in addition to the seal breakage.

[0024]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the seal inspection system according to the present invention will be described below with reference to the drawings.

FIG. 1 shows a block diagram of the seal inspection system. In this figure, inspection work 1
0 is intended for an automobile engine head cover or the like, and a seal material 41 such as a silicone resin is applied to the joint surface 40 of the work 10. The application of the sealing material 41 is performed by a robot (not shown) for applying the sealing material 41. This sealing material application robot has a nozzle that discharges the sealing material 41 onto the joint surface 40 and the nozzle that joins the joint surface 4
Nozzle drive means for moving along 0,
The nozzle driving means is driven and controlled by the robot controller 11. The robot controller 11 teaches nozzles on a coating locus along the joint surface 40 of the work 10 to continuously produce the work 1.
It has a function of sequentially applying the sealing material 41 to the bonding surface 40 of 0.

Above the inspection position of the work 10,
A plurality of CCD cameras 12 and an illuminating device 13 forming an image pickup unit are arranged. In this embodiment, the CCD camera 12 has an effective pixel number of 768 (H) × 494 (V), and images the joint surface 40 in the inspection work 10 in monochrome. Further, the photographing distance between the CCD camera 12 and the work 10 is set so that the seal can be photographed to a size of, for example, 4 to 6 pixels within this effective number of pixels. The work to be inspected and the position of the CCD camera 12 are fixed,
The joint surface 40 of the work 10 is always imaged at the same position.

The CCD camera 12 has an image switching device 14
The image switching device 14 sequentially switches the CCD cameras 12 according to the position of the imaging target, and inputs the analog image data captured by the connected CCD camera 12 to the digitizer 15.

The digitizer 15 includes an A / D converter 16 for converting analog image data output by the CCD camera 12 into digital image data, and an image memory 17 for storing the image data output by the A / D converter 16. , And a D / A converter 18 for analog-converting the image data in the image memory 17 for monitor display. Here, the A / D converter 16 converts the analog image data into 2
It is converted into digital image data of 56 gradations (density value "0" to "255"). Hereinafter, the 256-tone digital image data is simply referred to as image data.

On the other hand, the D / A converter 18 is connected with internal and external monitors 19 and 20 and a printer 21. Since the image data once stored in the image memory 17 is D / A converted and displayed on the monitors 19 and 20, the coordinates of the image displayed on the monitor correspond to the coordinates based on the address of the image memory 17. Therefore, the position designated by the operator with respect to the images displayed on the monitors 19 and 20 corresponds to the address of the image memory 17, and further, the CCD camera 1
2 corresponds to the position of the joint surface 40 imaged.

Moreover, here, since the position of the CCD camera 12 is fixed, it is specified how many millimeters of the joint surface 40 the vertical and horizontal length of one pixel of the image data in the image memory 17 corresponds to. It Further, since the actual size of this one pixel is specified, the actual length of the image data on the monitor can be displayed together.

The D / A converter 18 has a printer 21
Is installed side by side and is configured so that image data, inspection results, etc. can be printed out as required.

A microcomputer 25 as a processing device is connected to the image memory 17. The microcomputer 25 performs grayscale image processing on the image data stored in the image memory 17, and also gives instructions such as the timing of drive control by the robot controller 11. The microcomputer 25 is actually provided with a storage means such as a hard disk or a ROM (not shown) in which a program in which a control procedure is described is stored in advance.

The microcomputer 25 outputs a predetermined coating start signal and inspection end signal to the robot controller 11 and inputs a photographing start signal output from the robot controller 11 via the I / O board 26. It has and. In addition, a predetermined parameter is input to the microcomputer 25 by the keyboard 24 connected to the outside. Specifically, the inspection start position, the inspection start direction, the inspection end position, the allowable cut width, the allowable width, and the like according to the inspection target when the teaching program is executed can be arbitrarily set as variable information.

The microcomputer 25, in accordance with various programs, defines an inspection position specifying section 25A for defining an inspection point of the sealing material application state on the image data, and a seal of the inspection point p defined by the inspection position specifying section 25A. The coating state inspection unit 25B for determining the coating state of the material by image processing is provided.

The inspection position specifying unit 25A specifies the position where the sealing material should be applied. That is,
In the 512 × 512 image data captured by the CCD camera 12, the coordinates at which the presence of the sealing material should be detected are specified. When the sealant is extracted well, a continuous line having a thickness of 4 to 6 pixels is detected in this image data. The position of this line is determined by the shape of the work and the position where the sealing material 41 is applied to the joint surface 40.

The inspection position specifying unit 25A first monitors the image data of the joint surface 40 to which the seal material is normally applied by the monitor 1.
It is displayed on 9, 20. Alternatively, the position where the sealing material should be normally applied is painted black, for example, and the bonding surface 40
The image data of is displayed on the monitor. Then, the operator receives the inspection start position and the inspection end position as coordinate information on the monitor. Further, when an instruction of a search direction is received from the operator, the application position of the seal material conforming to the shape of the work is automatically detected as the coordinates of the image memory by the search processing.

Here, it is performed by tracing the pixels judged as the sealant in the image data one by one. In this search processing, first, the number of consecutive pixels in a straight line in the search direction is obtained. If the number of continuous pixels is larger than the reference value, one pixel is advanced in that direction and the coordinates are recorded.

Further, when the number of continuations is smaller than the reference value, it is determined that the seal is a bent portion and a new search direction is determined. If the vehicle cannot go straight, the direction of 45 degrees to the left and right is checked, and the new direction is determined by comparing the density values.

Next, the average of the densities at the current position and the densities of the pixels adjacent in the traveling direction is set as the reference density value, and the average of the densities of the second and third pixels from the current position is set as the comparative density value. When the reference density value is equal to or higher than the comparison density value, one pixel is advanced in that direction and the coordinates are recorded with the advanced position as the current position.

Each time the coordinate is searched and the coordinates are recorded, it is confirmed whether or not the position is within the range of the end position. If the position is within the end range, it is determined to be normal and the process ends. On the other hand, search is for sealant 4
Considering the case where the search cannot be performed normally beyond 1 and if the end position is not reached even after repeating the search a certain number of times, it is determined to be abnormal and the process ends.

As a result, the line coated with the sealing material is extracted. Furthermore, coordinates are specified by inspecting points at regular intervals on this line. At this time, the inspection point is set at the center of a certain width.

The coating state inspecting section 25B determines the position where the density change value in the width direction of the sealing material 41 at the inspection point defined in advance on the image data becomes equal to or larger than the set value, the both ends positions in the width direction of the sealing material. Both side detection function to judge
It is provided with a break error judging function for judging a seal break at the inspection point when both ends in the width direction are not detected by the both side detecting function.

Further, the coating state inspecting section 25B has a coating width calculating function for calculating the width of the sealing material at the inspection point based on the widthwise both end positions detected by the both side detecting function, and this coating width calculating function. And a coating width error determining function for determining a coating width error at the inspection point when the coating width calculated by is smaller than a predetermined minimum set width.

Further, the coating state inspecting section 25B has a shift amount calculating function for determining the shift amount of the coating position based on the widthwise both end positions detected by the both side detecting function and the position of the inspection point, and this shift. A deviation error determination function for determining a deviation error at the inspection point when the deviation amount calculated by the quantity calculation function is larger than a preset set deviation amount is provided.

Further, in the present embodiment, when each error occurs continuously at a plurality of inspection points, it is determined that the coating state is defective. Therefore, the coating state inspection section 25B counts the number of each error. It has an error counter.

Next, the coating state inspection unit 25B will be described in more detail.

The coating state inspecting section 25B is provided in the image memory 17
The presence of the sealing material is confirmed by detecting the portion of the image data accumulated in the area where the density change is large. Here, as shown in FIG. 2, as the inspection of the coating state, three elements, that is, the cut of the sealing material, the coating width, and the shift of the sealing material coating position are confirmed. First, a method of extracting a sealing material portion by image processing will be described with reference to FIG. 3, and then a processing step of the coating state inspection unit 25B that inspects the coating state of the sealing material will be described.

The inspection point p is defined at the center position in the width direction of the seal material 41 applied to the work joining surface 40. The inspection points p are arranged at regular intervals along the coating direction of the sealing material 41. Further, an inspection width w is defined for the inspection point p. this is,
The width of the sealing material is set to be a certain distance wider from the inspection point, and both ends of the inspection width w are located in the work joining surface 40 here.

The coating state inspection unit 25B detects a change in density from the pixels at both ends of the inspection width w to the pixel at the inspection point p. As described above, since the edge portion in the width direction of the sealing material 41 is not easily affected by the reflection of illumination, the coating state inspection unit 25B determines the pixels adjacent to each other in the inspection point p direction from the density of the pixels at both ends of the inspection width w. The density change is calculated, and if the density change is equal to or more than the set value, it is determined that the edge of the sealing material.

Among the pixels from the both ends of the inspection width w to the inspection point p, the positions of the pixels whose density change is equal to or more than the set value are determined to be the both ends of the seal material, so the coating state inspection unit 25B It is possible to extract the width at one point (inspection point) of the sealing material applied on the bonding surface 40 captured by the CCD camera 12. Since the detected width of the sealing material is calculated as the number of pixels, by comparing with the number of pixels according to the width when the coating is normally applied, the C
It is possible to quantitatively determine whether the width of the sealing material at the inspection point p captured by the CD camera 12 is thinner or thicker than normal.

Further, since the inspection point p is set at the center position in the width direction of the sealing material which is normally applied, it is determined that the inspection point p is on both sides (edges) of the sealing material. By comparing the position of the applied pixel with the position of the applied pixel, it is possible to quantitatively determine how much the seal material deviates from the position where it should be applied.
Since the CCD camera 12 images the work joining surface 40 so that the width of the seal material that is normally applied is 4 to 6 pixels, the width and the amount of deviation of the seal material can be detected in millimeters.

The coating state inspecting unit 25B detects the width of the sealing material at each inspection point arranged at a constant interval in the sealing material coating direction. At this time, the coating state inspecting unit 25B determines that a break has occurred in the coating direction of the sealing material when the detection of the width of the sealing material fails continuously at a plurality of inspection points p.

Now, the step of inspecting the coating state of the sealing material having the structure shown in FIG. 1 will be described with reference to the flowchart of FIG.

First, an illuminance inspection is performed (step S
1). This is because when the lighting gets dark, the sealing material 41
Since the density of the edge portion of the sealing material and the density of the bonding surface 40 are similar to each other and both ends (both sides) in the width direction of the sealing material based on the density change cannot be detected well, the illuminance before the image pickup by the CCD camera 12 It is an inspection. If the illuminance test is NG, it is determined that the illumination is defective and the process is terminated without performing the test (step S2).

Next, with respect to the inspection point p defined in advance for the image data picked up by the CCD camera 12, the detection processing of both sides of the sealing material is performed. further,
The coating state is inspected based on this detection result.

Steps S3 to S1 shown in FIG.
Up to 8, the process flow is for one inspection point p. Since a plurality of inspection points p are defined for the image data along the work joint surface 40, both side detection processing of all the inspection points p defined for this image data in step S18 ends. It is determined whether or not.

In addition, in the inspection of the coating state shown in FIG. 2, the break of the sealing material coating (steps S3 to S6)
And the width of the applied sealing material (steps S8-S
9), deviation of the sealing material application position (steps S13 to S)
16) Inspect. In the present embodiment, in each inspection, an error (a defective coating state) is determined when a break or the like occurs not at one inspection point but at continuous inspection points.

In step S3, the presence / absence of a pixel having a density change greater than a set value among the pixels from the inspection width w to the inspection point p is detected. If neither side of the seal width is detected, the value of the break error counter is incremented by "1" (step S4). When the value of the break error counter is equal to or greater than the set value (step S5), that is, when it is determined that there is a break at the inspection points p that are equal to or larger than the predetermined number of inspection points, the NG flag is set to "ON".
Then, a break / shift error is determined (step S6).
There is a possibility of a misalignment error because it is assumed that the sealing material 41 is applied outside the inspection width w.

When both sides of the seal material are detected in step S3, first, the break error counter is initialized (step S7). Further, the number of pixels between both sides of the sealing material is calculated, and it is confirmed based on the number of pixels whether the width of the sealing material is equal to or larger than the minimum set width (step S).
8). When the width is less than the minimum set width, that is, when the number of pixels is 4 to 6 pixels or less in the above example, the value of the width error counter is incremented by "1" (step S9). Then, it is confirmed whether or not the value of the width error counter is equal to or larger than the set value (step S10). That is, it is confirmed whether the inspection points p that are less than the minimum set width are continuous. When it is determined that the width of the sealing material is continuously small at the inspection points p equal to or larger than the preset number of inspection points, the NG flag is set to "ON" and a width error is generated (step S11).

If it is determined in step S8 that the width is greater than or equal to the minimum set width, first, the value of the width error counter is initialized (step S12). Further, based on the pixel positions on both sides of the sealing material and the pixel positions of the inspection points, it is confirmed whether or not the both sides are located within the setting deviation (step S13). If it exceeds the setting deviation range, the deviation error counter value is set to "1" as in other cases.
It is increased (step S15). Further, when it is determined that the deviation of the coating position occurs at consecutive inspection points p (step S15), the NG flag is set to "ON" and a deviation error occurs (step S11).

When it is determined in step S13 that the positions of both sides of the seal material are within the set deviation, first, the deviation error counter is initialized (step S17). If each error is not equal to or more than the set number of times in the processing from steps S3 to S17, it is determined that the coating state is good, the next inspection point is selected, and the processing is returned to step S3.

On the other hand, if it is determined to be NG, it may be possible to externally display the fact that the sealing material coating state is NG as it is and terminate the processing. The next inspection point is selected and the process is returned to the step S3 in order to obtain the information as to whether the occurrence has occurred. In this embodiment, the malfunction of the coating robot is specified based on the error information for each inspection point. Further, the operation of the coating robot may be controlled based on this error information, and the sealing material may be coated again. That is, the microcomputer 25 may cause the robot controller 11 to control the application of the sealing material again based on the position corresponding to the inspection point and the error type of the inspection point.

When the processing is completed for all the inspection points (step S18), the state of the NG flag is confirmed (step S19), and if the NG flag is ON, NG is output (step S20), while the NG flag is output. If is not ON, OK is output (step S21). The inspection result of the coating state of the sealing material automatically performed by image processing can be displayed externally. This external display is displayed on the monitors 19 and 20.
Alternatively, the printing may be performed by the printer 21.

CCD camera 12 according to the size of the work
In the case where there are a plurality of images, after the processing for one image data is completed, the input of the CCD camera 12 is switched by the image switching device 14 and the processing for the next image data is performed.

As described above, according to the inspection process shown in FIG. 2, the presence or absence of the sealing material is detected by detecting the edge portion of the sealing material. Both sides can be detected, and therefore, the application state of the sealing material can be inspected by image processing without creating an ideal illumination environment.

Further, since the pixels whose density change from the both ends of the inspection width toward the inspection point is equal to or more than the set value are determined as the both ends of the sealing material, the width of the sealing material can be quantitatively captured by the number of pixels. Since it is possible to specify how many millimeters of one pixel corresponds to the actual bonding surface 40 by specifying the CCD camera position, the width of the sealing material defined by this number of pixels is actually several millimeters. Can be accurately calculated.

Moreover, since both ends of the inspection width are set as start positions and the amount of change in density is detected based on the density value of the pixel at the position, even if the brightness on the surface to be inspected is not uniform, that is, the illumination environment is Even if the reflection is not good and some irregular reflection occurs on the joint surface 40, accurate determination can be stably performed.

Further, since the inspection point corresponds to the actual position on the bonding surface 40, the sealing material application operation of the robot controller is controlled based on the error situation at each inspection point, so that the application is not properly performed. It is possible to apply the sealing material again automatically and accurately at each position.

[0069]

The present invention is constructed as described above, and
According to this, according to this, the both-sides detection function determines the position of the pixel from the both end positions of the inspection width in the image data to the inspection point where the density change value is the set value or more in the width direction of the seal material. Since it is determined to be the both end positions, it is possible to detect the presence or absence of the seal material 41 based on the portions of the both ends of the seal material 41 which are always imaged darkly. Furthermore, the break error determination function is performed by the both side detection functions. Since it is determined that the inspection point is a seal break when both ends in the width direction are not detected, it is possible to automatically detect a seal break, which is a defective application state of the sealing material 41, by image processing. It is possible to provide a seal inspection system that produces an effect.

Further, since both ends of the inspection width are set on the work joining surface 40, when the processing device retrieves the pixel position of the inspection width portion centering on the predetermined inspection point on the image data, this inspection is performed. The positions at both ends of the width are pixels on the joint surface 40. Therefore, the gradation value that is the reference for obtaining the density change value can be the gradation value on the joint surface 40. Even if the density value changes due to the illumination, the sealing material 41 can be detected based on the changed density value. Therefore, the sealing material 41 can be detected.
The inspection of the coating state can be stably performed without being affected by the change of the inspection environment.

Further, since the coating width calculating function calculates the width of the sealing material at the inspection point based on the both ends in the width direction detected by the both side detecting function, the sealing material on the imaged joint surface 40 is calculated. The width of 41 can be calculated as the number of pixels, and this pixel is
Since it corresponds to the above size, the coating width of the sealing material 41 on the bonding surface 40 can be calculated based on the number of pixels, and the coating width error calculation function is calculated by this coating width calculation function. When the coating width is smaller than the predetermined minimum setting width, it is determined that the inspection width is a coating width error. Therefore, the state of the coating width can be determined by a high-speed image processing such as comparison of the number of pixels.

Further, when the error of the coating width is continuously generated at a plurality of inspection points, it is judged that the coating state is defective. It is possible to satisfactorily detect it according to the shape etc. of the coating, and if it is judged that the coating state is defective when it occurs continuously at multiple inspection points, it is mistaken as a coating width error due to noise etc. at one inspection point. Even if it is determined, this is not determined to be a defective coating state, and thus a defective coating state can be determined without being affected by noise.

Further, since the deviation amount calculation function determines the deviation amount of the coating position based on the both end positions in the width direction detected by the both side detection function and the position of the inspection point, the position of the predetermined inspection point is determined. Thus, it is possible to provide a seal inspection system that has an excellent effect that has not been obtained in the past, in which the amount of deviation of the actually applied seal material 41 can be quantitatively determined as the number of pixels.

Further, the processing apparatus has a coating state determination function,
Conventionally, it is possible to determine the coating state without being affected by noise or the like at one inspection point, since it is determined that the coating state is defective when a coating deviation error continuously occurs at a plurality of inspection points. It is possible to provide a seal inspection system with excellent effects not found in the market.

[Brief description of drawings]

FIG. 1 is a block configuration diagram showing an embodiment of a seal inspection system according to the present invention.

FIG. 2 is an inspection flowchart of the seal inspection system.

FIG. 3 is a schematic plan view of a seal inspection point.

4A to 4D are explanatory views of a conventional example in which image data is binarized and a seal inspection is performed.

[Explanation of symbols]

 12 CCD camera 15 Digitizer 17 Image memory 25 Microcomputer as a processing device 40 Bonding surface 41 Sealing material

Front page continued (72) Inventor Hiroyuki Okamoto 14-4 Nishiiuma-cho, Matsue-shi, Shimane Prefecture Matsue National College of Technology

Claims (7)

[Claims] 1. A means for picking up an image of a work joining surface coated with a sealing material, an image memory for storing the data taken in by the image pick-up means as image data, and the image data connected to the image memory. In a sticker inspection system including a processing device for performing grayscale image processing, the processing device specifies an inspection portion for searching a pixel position of an inspection width portion centered on a predetermined inspection point on the image data. Both sides that determine the function and the position of the pixel from the both end positions of the inspection width in the image data to the inspection point where the density change value is equal to or greater than the set value as the both end positions in the width direction of the sealing material. A detection function and a cut that determines that the inspection point is a seal break when both ends in the width direction are not detected by the both side detection function. Seal inspection system is characterized in that a color determination function. 2. The seal according to claim 1, wherein the processing device has a coating state determination function that determines that the coating state is defective when the break error occurs continuously at a plurality of inspection points. Inspection system. 3. The inspection point is set at the center of the normally applied sealing material in the width direction, and both ends of the inspection width are set on the work joining surface. The seal inspection system according to item 1. 4. The coating device has a coating width calculating function for calculating the width of the sealing material at the inspection point based on the widthwise both end positions detected by the both side detecting function, and the coating width calculating function. 4. A coating width error determination function for determining a coating width error at the inspection point when the calculated coating width is smaller than a predetermined minimum set width.
Seal inspection system described. 5. The processing apparatus according to claim 4, further comprising a coating state determination function for determining a coating state defect when the coating width error continuously occurs at a plurality of inspection points. Seal inspection system. 6. The deviation amount calculation function of the processing device, which determines the deviation amount of the coating position based on both end positions in the width direction detected by the both side detection function and the position of the inspection point, and the deviation amount. 4. A shift error determination function for determining a shift error for the inspection point when the shift amount calculated by the calculation function is larger than a preset set shift amount. Seal inspection system described. 7. The processing apparatus according to claim 6, further comprising a coating state determining function for determining that the coating state is defective when the coating deviation error continuously occurs at a plurality of inspection points. Seal inspection system.