JPH0650904A - Method for correcting size of surface defect - Google Patents

Abstract

PURPOSE:To detect the distance and the angle between an optical-system sensor and a material to be inspected and to correct the size of a defect, whose image is picked up, based on the position relationship with respect to the method for correcting the size of the optically detected defect from the surface of the material to be inspected. CONSTITUTION:An optical-system sensor 12 having a light source 4, a mirror 5 and a camera 6 is provided at the upper side of a vehicle body 2, which is a material to be inspected. Non-contact type distance sensors 71-74 are arranged around the light source 4. Distance sensors 71-74 independently measure the distance between the optical-system sensor 12 and the body 2, respectively. The angle of the body 2 with respect to the optical-system sensor 12 and the distance between both parts are computed based on the measured distance data. The picture, which is picked up with the camera, is corrected based on the computed position relationship. Therefore, the size of the recognized defective part is not fluctuated by the fluctuation of the position relationship between the optical-system sensor 12 and the body 2.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a surface defect size correction method, and more particularly to a method for correcting the size of a defect optically detected from the surface of an object to be inspected.

[0002]

2. Description of the Related Art Conventionally, various proposals have been made for optically detecting defects on the surface of the surface to be inspected. For example, the present applicant proposes a device for detecting a defect by irradiating a predetermined light on the coating film surface, automatically capturing the reflected light, and performing image processing in order to automatically inspect the coating film surface. (Japanese Patent Application Laid-Open No. 1-213509).

Generally, defects on the surface of a coating film include a protrusion called "butsu" which is generated by coating with foreign matter such as dust adhered thereto, and a "drip" which is caused by over-spraying of the paint. Limited. Therefore, the apparatus described in the above publication detects the protrusion by optically imaging the surface of the coating film, and judges that the coating quality is poor when the detected protrusion is larger than a predetermined size.

Further, in this apparatus, an optical system sensor including a light source for irradiating the coating film surface with light and a camera for imaging the coating film surface is configured to scan the coating film surface. Therefore, according to the apparatus described in the above publication, it is possible to automatically detect the spots and sags on the entire coating film surface.

[0005]

However, in the above-mentioned conventional apparatus, the positional relationship between the coating film surface and the optical system sensor is not taken into consideration. That is, regardless of the distance or angle between the optical system sensor and the coating film surface, the quality of the coating film is judged by the size of the defective portion captured by the camera.

Therefore, if the coating film surface is a curved surface or the distance between the coating film surface and the optical system sensor varies, the size of a defect of the same size when captured by a camera is different even if the defect has the same size. There are cases. Therefore, when the quality of the coating film surface is judged from the size of the detected defect, the judgment may be erroneous. That is, the above-mentioned conventional device has a problem that the reliability of the inspection is low and it cannot be completely automated in practical use.

The present invention has been made in view of the above points, and detects the distance and angle between the optical system sensor and the surface of the object to be inspected, and based on the distance and the angle, the surface of the object to be inspected is detected. It is an object to provide a method for correcting a defect size.

[0008]

As shown in the principle diagram of FIG. 1, the above-mentioned problems are caused by projecting light on the surface of an object to be inspected using a predetermined light source and reflecting the light on the surface of the object to be inspected. A method for correcting the size of a surface defect of the inspected object captured by capturing an image of light with a predetermined image capturing means, the method comprising:
In step M1, the distance between the optical system sensor including the light source and the image pickup device and the object to be inspected is measured, and in the second step M2, the optical system sensor is detected based on the measured distance data. By a positional defect size correction method for detecting a positional relationship such as a distance and an angle with the inspection object and correcting the defect size on the surface of the inspection object based on the detected positional relationship in a third step. Will be resolved.

[0009]

According to the above construction, in the first step, the distance between the object to be inspected and the optical system sensor is measured.

In the second step, the positional relationship such as the distance and the angle between the object to be inspected and the optical system sensor is detected based on the distance data measured in the first step.

In the third step, the size of the defect of the inspection object picked up by the image pickup means is corrected based on the positional relationship detected in the second step. That is,
The difference in defect size due to the difference in imaging distance or imaging angle is eliminated, and the defect size is not overestimated or underestimated.

[0012]

Next, in order to further clarify the structure of the method for correcting the surface size according to the present invention, the surface defect size of the inspection object is corrected by one embodiment of the method according to the present invention, and A surface defect detection device that determines the quality of the surface of an object will be described.

FIG. 2 is a system configuration diagram of a coating film surface inspection apparatus to which the method of this embodiment is applied. As shown in the figure,
The vehicle body 2 after the painting work is mounted on the line conveyor 1. Behind the line conveyor 1, a flowing car body 2
A recognition sensor 3 for reading the vehicle body number is arranged.

Optical system sensor 12 in the apparatus of this embodiment
Is composed of a light source 4, a mirror 5 and a camera 6. The light source 4 irradiates the vehicle body 2 flowing on the line conveyor 1 with predetermined light. At this time, the light reflected by the vehicle body 2 is reflected by the mirror 5 and is incident on the camera 6 corresponding to the image pickup means.

Further, the light source 4 is provided with four non-contact type distance sensors 7 _{1 to} 7 ₄ for measuring the distance to an object using laser or ultrasonic waves as a medium. These distance sensors 7 _{1-7 4} is connected to the image processing apparatus 8 with the camera 6.

That is, the image processing device 8 processes the video signal transmitted from the camera 6 to detect a defect on the surface of the vehicle body 2, and based on the distance signal transmitted from the distance sensors 7 _{1 to} 7 _4. Correct the detected defect size,
Judges whether the coating surface is good or bad. Note that this point will be described in detail later.

In the figure, reference numeral 9 indicates a gate type robot. The robot 9 has a pillar portion 9a that moves back and forth, left and right, and up and down, and a holding portion 9b fixed to the pillar portion 9a.
An optical system sensor 12 is held in. Also, reference numeral 1
Reference numeral 0 denotes a robot control panel, which sends a command to the robot 9 to move the optical system sensor 12 along the unevenness of the vehicle body 2 flowing on the line conveyor 1.

Further, the computer 11 uses the recognition sensor 3
Vehicle type information relating to the shape and color of the vehicle body 2 is transmitted to the image processing device 8 from the vehicle body number read by. In response to this, the image processing device 8 transmits a start start signal to the robot control panel 10. Therefore, in the apparatus of this embodiment, it is possible to flow a plurality of vehicle bodies having different shapes on the same line.

FIG. 3 is a perspective view showing the structure around the optical system sensor. The configuration around the optical system sensor, which is a main part of the apparatus of this embodiment, will be described below with reference to FIG. still,
In the figure, the same parts as those in FIG. 2 are designated by the same reference numerals.

In the figure, reference numeral 12 indicates an optical system sensor in which the light source 4, the mirror 5 and the camera 6 are held in a predetermined positional relationship. The optical system sensor 12 is fixed to the holding portion 9b of the robot 9 by a stay 12b at the center of the main body 12a.

As shown in the figure, distance sensors 7 _{1 to} 7 ₄ are provided at four corners of the light source 4. These distance sensors 7 _{1-7 4,} a non-contact type sensor as described above, each independently, can measure the distance between the vehicle body 2. Therefore, it is possible to simultaneously measure the distances of the _four distance sensors 7 _{1 to} 7 _{4 from} the vehicle body 2 at the respective positions.

Further, in order to calculate the distance between the optical system sensor 12 and the vehicle body 2 and the angle of the vehicle body 2 with respect to the optical system sensor 12, the distance between the optical system sensor 12 and the vehicle body 2 is determined by three points that are not on a straight line. Just measure. As described above, in the device of this embodiment, the distance sensor exists at each corner of the quadrangle. Therefore, there are four methods for selecting any three distance sensors to measure the position of the vehicle body 2, and the vehicle body 2 can be detected as a composite of four planes.
Therefore, according to the device of the present embodiment, the surface condition of the vehicle body 2 can be accurately grasped.

As described above, the camera 6 and the distance sensors 7 _{1 to} 7 ₄ are connected to the image processing device 8. Less than,
The image processing device 8 and the computer 11 will be described with reference to the configuration diagram shown in FIG.

As shown in the figure, the image processing device 8 has a C
The PU 8a, the ROM 8b, and the RAM 8c are mainly included in the bus 8 and the A / D converter 8d and the external input / output circuit 8e.
It is configured as a logical operation circuit mutually connected by f.

The above-mentioned robot control device 10 is connected to the external input / output circuit 8e of the image processing device 8, and
The external input / output circuits 11d of the computer 11 are connected to each other. The A / D converter 8d, above the camera 6 and the distance sensor 7 _{1-7 4} is connected.

The computer 11 has a CPU 11
A, the ROM 11b, and the RAM 11c are mainly connected to the external input / output circuit 11d by a bus 11e. The external input / output street 11d is mutually connected to the external input / output circuit 8e of the image processing device 8 as described above, and is also connected to the recognition sensor 3.

5 and 6 are flowcharts of the processes executed by the image processing apparatus 8 and the computer 11. The operation of the apparatus of this embodiment will be described below with reference to the drawings. The "computer main routine" shown in FIG. 5 (A) shows the processing executed by the computer 11.
The “image processing main routine” shown in FIG. 6B and the “defect determination processing routine” shown in FIG. 6 show processing executed by the image processing apparatus 8.

The coated vehicle body 2 is carried by the line conveyor 1 at a constant speed. When the vehicle reaches a predetermined position, the recognition sensor 3 reads the manufactured vehicle body number. Then, the computer 11 takes in the vehicle body number read by the recognition sensor 3 via the external input / output circuit (step 101). Then, a process for searching the vehicle type of the vehicle body 2, the color of the applied coating, and the like from this manufacturing number is performed (step 102).

When the vehicle type and color information of the vehicle body 2 are obtained, this information is output to the image processing device 8 via the external input / output circuit (step 103). After that, the processing in the computer 11 is in a state of waiting for an input from the image processing device 8 (step 104).

On the other hand, as shown in FIG. 1B, the image processing apparatus 8 is connected to the computer 1 via the external input / output circuit 8e.
When the information of the vehicle body 2 is supplied from 1 (step 10
5) Based on this information, the robot 9 is moved to a predetermined position for each vehicle type (step 106).

When the camera 6 is moved to a predetermined position which is predetermined for each vehicle type along with the movement of the robot 9, subsequently, a defect determination process described later is executed (step 10).
7). The defect information obtained as a result is the external input / output circuit 8
It is output to the computer 11 via e (step 1
08).

After that, the process returns to the above-mentioned step 105, and the image processing device 8 is in a waiting state for the vehicle body information from the computer 11.

Since the computer 11 is waiting for the input from the image processing device 8, when the defect information of the coating film surface is supplied to the computer 11, the process proceeds to step 109. In step 109, the defect information is output to the painting process together with the information such as the inspection date, the vehicle type, and the painting color.

Therefore, according to the apparatus of this embodiment, the coating result can be fed back to the coating condition, and the stable coating quality can be always secured. Further, when the quality of the coating film suddenly changes, a history of defective coating films is always left, so that the cause can be easily investigated.

Next, the defect determination processing routine (step 107 above), which is the main part of the processing performed by the apparatus of this embodiment, will be described.

As shown in FIG. 6, when this process is started, the image processing device 8 firstly detects the vehicle body 2 imaged by the camera 6.
The image signal of the surface of is acquired (step 201). Next, this signal is subjected to a known process such as a differential process or a binarization process to classify the captured image into a smooth region and a non-smooth region (step 202).

Most of the coating film defects of the vehicle body 2 are "bugs" caused by adhesion of dust or the like in the coating process, or "sag" caused by defective coating conditions. To be detected. That is, the quality of the coating film surface can be evaluated by the smoothness of the coating film surface, and the non-smooth area in the above step 202 is determined to be a defective area due to spots or sagging.

Next, the size of each non-smooth region, that is, the size of each spot or sag is calculated from the image signal captured from the camera 6 (step 203).

By the way, FIG. 7A is a diagram showing a path through which the light emitted from the light source reaches the camera 6. Further, FIG. 1B shows the vehicle body 2 imaged by the camera 6.
The top defects 20-22 are shown.

As shown in FIG. 1A, the camera 6 images the light reflected from the vehicle body 2 from the light source 4, that is, the light emitted at the position of the virtual image light source 14. At this time, if there are defects 20 to 22 on the vehicle body, diffuse reflection of light occurs at these points, and the reflected light does not reach the camera 6. Therefore, the defect can be detected by imaging the reflected light with the camera 6.

However, as shown in FIG.
In order for the light emitted from the vehicle to be reflected by the vehicle body 2 and reach the mirror 5, the light from the light source 4 must be incident from the oblique direction of the vehicle body 2. That is, with respect to the optical path from the virtual image light source 14 to the virtual image camera 16, the vehicle body 2 which is the surface to be inspected
Cannot be vertical.

Therefore, the defects 20 to 22 on the vehicle body 2 are
When scattered in the Y direction as shown in FIG.
A difference occurs in the distance from 22 to the camera 6 (virtual image camera 16). Therefore, if the sizes of the defects 20 to 22 are the same, the defect 20 located closest to the camera 6 is detected.
Is imaged the largest, and the defect 22 located farthest from the camera 6 is imaged the smallest.

As described above, when the optical path to be imaged is not perpendicular to the object to be inspected, the image captured by the camera 6 is distorted. Therefore, in the apparatus of the present embodiment, the image distortion caused by such an optical path difference is corrected at a rate set in advance based on the image coordinates of the camera 6 (step 204).

Further, as described above, the distance sensors 7 _{1 to} 7 ₄ are connected to the A / D converter 8 d of the image processing apparatus 8, and the distance measurement values from the respective sensors 7 _{1 to} 7 ₄ are measured. α _{1 to} α ₄ are supplied.

In the apparatus of this embodiment, as the first step, in step 205, these distance measurement values α _{1 to} α ₄ are fetched. Then in the second step, step 206,
The distance from the optical system sensor 12 to the vehicle body 2 and the angle of the vehicle body 2 with respect to the optical system sensor 12 are calculated based on α _{1 to} α ₄ .

In steps 205 and 206, the optical system sensor 1
When the positional relationship between the vehicle body 2 and the vehicle body 2 is clarified, as the third step, the image signal corrected in the above step 204 is further corrected.

That is, the positional relationship between the optical system sensor 12 and the vehicle body 2 varies to some extent due to the limitations of the molding accuracy of the vehicle body 2 and the operating accuracy of the robot 9. Due to this variation, if the positional relationship between the two deviates from the relationship during teaching of the robot 9, the camera 6 will take a distorted screen as in the case described above.

Therefore, in the apparatus of the present embodiment, the image of the defective portion actually captured by the camera 6 is rotated or enlarged or reduced based on the positional relationship between the optical system sensor 12 and the vehicle body 2, thereby Is corrected to an image that would be picked up in the case of the reference positional relationship (step 207).

As described above, in the apparatus of this embodiment, the change in the positional relationship between the optical system sensor 12 and the vehicle body 2 and the vehicle body 2
The image distortion caused by the difference in the distance from the camera to the camera 6 is corrected.

Therefore, due to the problem of the molding accuracy of the vehicle body 2 or the operation accuracy of the robot 9, the optical system sensor 12 and the vehicle body 2
Even if the positional relationship between and changes, the size of the detected defective portion does not change.

Next, the size of the defect portion thus calculated is compared with a predetermined threshold value (step 208), and it is determined whether or not the imaged defect corresponds to a defective coating film (step 209). . That is, it is determined that the coating film is defective when the size of the imaged defective portion is equal to or larger than a predetermined threshold value, and that the coating film is normal when the size is smaller than the threshold value.

Thereafter, the process returns to step 108 of the image processing main routine, and waits for this routine to be started again. In this embodiment, after the defect portion data without image distortion is calculated in step 207, this data is used only for comparison with the threshold value, but the present invention is not limited to this. It is also possible to take out as a list and add a configuration used for analysis of coating conditions and the like.

Further, in the apparatus of this embodiment, four distance sensors are used to detect the positional relationship between the optical system sensor 12 and the vehicle body 2. However, for example, one distance sensor is moved to obtain a plurality of points. The distance data may be measured, or the distance data at four points may be measured by moving two distance sensors.

[0054]

As described above, according to the present invention, the positional relationship between the optical system sensor and the inspection object is clarified by the distance between the optical system sensor and the inspection object measured at a plurality of points. Based on this, the image captured by the image capturing means is corrected.

Therefore, in the conventional method, when the distance or the angle between the optical system sensor and the inspection object changes, the size of the defective portion imaged on the inspection object changes accordingly. Therefore, according to the method of the present invention, the imaged image is always standardized, and the size of the recognized defective portion does not change due to the change in the distance or angle between the two.

Therefore, by using the method of correcting the surface defect size according to the present invention, the accuracy of judging the quality of the surface to be inspected is significantly improved. For this reason, the inspection process can be completely automated, and the labor saving can be achieved.

[Brief description of drawings]

FIG. 1 is a principle diagram of a method of correcting a surface defect size according to the present invention.

FIG. 2 is a configuration diagram of an example of a defect detection apparatus using a surface defect size correction method according to the present invention.

FIG. 3 is a configuration diagram of an example of the periphery of an optical system sensor used in the apparatus of this embodiment.

FIG. 4 is a configuration diagram of an example of a control device used in the device of the present embodiment.

FIG. 5 is a flowchart of an example of processing executed by the control device of the apparatus according to the present embodiment.

FIG. 6 is a flowchart of an example of a main part of a process executed by the control device of the device of this embodiment.

FIG. 7 is a diagram showing a path of light when a defect is detected by the apparatus of this embodiment.

[Explanation of symbols]

M1 first stage M2 second stage M3 third stage 2 vehicle body 4 light source 5 mirror 6 camera 7 _{1 to} 7 ₄ distance sensor 8 image processing device 11 computer 12 optical system sensor

Claims (1)

[Claims] 1. A surface of the object to be inspected, which is captured by projecting light on the surface of the object to be inspected using a predetermined light source and imaging the light reflected by the surface of the object to be inspected by a predetermined imaging means. A method for correcting the size of a defect, which measures a distance between an optical system sensor including the light source and the imaging unit, and the inspection object, and based on the measured distance data, the optical system sensor and the A surface defect size correction method for detecting a positional relationship with an object to be inspected and correcting the defect size on the surface of the object to be inspected based on the detected positional relationship.