JP2956230B2 - Surface defect inspection equipment - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a surface defect inspection apparatus for inspecting defects on the surface of a work after processing, and more particularly to detecting a flaw or the like on a painted surface of a work after coating with high accuracy. The present invention relates to a surface defect inspection device that can perform the inspection.

[0002]

2. Description of the Related Art In general, a manufacturing process in a production plant or the like is provided with an inspection process for inspecting the quality of processed products. Therefore, the quality of the product is inspected using an optical device such as a camera.

Such a quality inspection includes a surface defect inspection for inspecting whether a defect such as a scratch is present on the surface of a processed product, and includes, for example, coating of a painted panel of an automobile. Inspection of the surface.

[0004] As a conventional painted surface inspection apparatus, for example, there is one as shown in FIG. This painted surface inspection device
A light source 10 for irradiating light (for example, slit laser light) to a surface to be inspected of a coating panel W as a workpiece, and a one-dimensional sensor 11 for receiving light reflected from the surface to be inspected and converting the light into an electric signal
And processing the signal from the one-dimensional sensor 11 to
And an image processing unit 12 for determining the presence / absence of the image.

[0005] The image processing section 12 detects a flaw in the coating panel W in accordance with, for example, the procedure shown in FIG. For example, when a flaw F exists in the paint panel W, the image processing unit 1
2 first inputs a video signal having a waveform as shown in FIG. 5A from the one-dimensional sensor 11, and then, in a binarization circuit, a DC voltage of a level as shown in FIG. After the threshold level is generated, the threshold level is compared with the input video signal to generate a binarized signal having a pulse waveform as shown in FIG. 5B. Then, it is determined that there is a flaw F based on the binarized signal.

[0006]

However, in such a conventional surface defect inspection apparatus, the so-called undulation generally occurs on the surface of the painted surface after the work is painted. When binarizing the video signal for detection, it is necessary to set the threshold level to a level that does not affect the swell,
It was difficult to detect a relatively small flaw.

For example, a painted panel having a small undulation (that is, a small surface roughness) (see FIG. 6A) and a painted panel having a large undulation (that is, a large surface roughness) (FIG. 6)
Looking at the case where the same small flaw F is detected while comparing with (b), if the undulation is small, a video signal having a waveform as shown in FIG. Since there is almost no surface unevenness, the threshold level for binarization can be set high, so that even a small flaw F can be detected. On the other hand, when the undulation is large, a video waveform as shown in FIG. 6D is obtained, and the depression at the flaw F portion does not fall much less than the uneven portion on the surface. In addition, it becomes impossible to selectively cut out only the flaw F portion by the binarization process, and it becomes impossible to detect a small flaw F.

SUMMARY OF THE INVENTION The present invention has been made in view of such problems of the prior art, and provides a surface defect inspection apparatus capable of detecting small scratches regardless of the roughness of a work surface. With the goal.

[0009]

SUMMARY OF THE INVENTION In order to achieve the above object, the present invention provides a surface roughness detecting means for detecting the roughness of a surface to be inspected on a surface of a work, and the surface roughness detected by the surface roughness detecting means. Storage means for storing roughness information, image input means for inputting an image of the surface to be inspected, and filter means for removing a roughness component stored in the storage means from an image signal output from the image input means And determination means for inputting an output signal from the filter means and determining presence / absence of a surface defect on the surface to be inspected of the work.

[0010]

According to the present invention thus constructed, the surface roughness detection means detects the roughness of the surface to be inspected, the roughness information is stored in the storage means, and the image input means is used for the image input means. Enter an image. Then, the filter unit removes the roughness component stored in the storage unit from the image signal from the image input unit,
The judging means judges the presence or absence of a surface defect on the surface to be inspected based on the smoothed signal from the filter means.

Accordingly, the determination means can set a high threshold level for binarizing the image signal by an amount corresponding to the removal of the roughness component, so that a smaller surface defect can be detected.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a schematic configuration diagram of a surface defect inspection apparatus according to one embodiment of the present invention. Here, a painted panel will be described as an example of a workpiece to be inspected. The same members as those in FIG. 3 are denoted by the same reference numerals.

This surface defect inspection apparatus includes a surface roughness meter 13 for detecting the roughness (that is, undulation) of a painted surface of a painted panel W, which is a work, and a light (for capturing an image on the inspected surface of the painted surface). Light source 10 for irradiating, for example, a slit laser beam)
And a one-dimensional sensor 11 that receives reflected light and captures an image of the surface to be inspected, and a painted panel W (or light source 10 and one-dimensional sensor 11) so as to capture image data of all portions of the surface to be inspected. Motor 14 to be moved and motor 14
Encoder 15 for detecting the rotation angle and rotation speed of
Display device 1 for displaying the operating status of various devices, inspection results, etc.
6 and an inspection apparatus main body 17 to which these devices are externally connected.

The inspection apparatus main body 17 is provided with an input / output device 18 for transmitting and receiving input / output signals to and from the external devices 10, 11, 13 to 16. The input / output device 18 has a surface roughness. A roughness information storage unit 19 for storing roughness information of the painted surface detected by the total 13 (for example, the swell of a normal painted surface is composed of a constant frequency component, and the frequency component of the swell); Processing unit 20 for processing the image signal output from the camera, surface roughness meter 13, one-dimensional sensor 11, light source 10, motor 14, and display device 1.
6 is connected to a device control unit 21 for controlling the operation of the image processing unit 6. Further, the roughness information storage unit 19 and the image processing unit 20 store the video signal from the image processing unit 20 in the roughness information storage unit 19. A filter processing unit 22 for performing a filtering process for removing the frequency component of the roughness is connected to the filter processing unit 22. A flaw detection unit 23 for detecting the presence / absence of flaws is connected. The flaw detector 23 has a binarization circuit. The device control unit 21 controls the operations of the surface roughness meter 13, the motor 14, the display device 16, and the like based on signals from the roughness information storage unit 19, the image processing unit 20, and the flaw detection unit 23. Has become.

The surface roughness detecting means is a ideographic roughness meter 13.
The storage means is formed by the roughness information storage unit 19, the image input means is formed by the one-dimensional sensor 11, the filter means is formed by the filter processing unit 22, and the judgment means is formed by the flaw detection unit 23.

Next, the operation of the thus configured surface defect inspection apparatus will be described with reference to the operation flowchart shown in FIG.

First, when the start of work is instructed and the program is started, the operation states of the inspection apparatus main body 17 and all the devices connected thereto are initialized, and preparation for surface defect inspection processing of the coating panel W is performed. And the painting panel W
Is disposed at a predetermined inspection position, the device control unit 21
The surface roughness meter 13 is operated via the input / output device 18 (step 1 above).

The surface roughness meter 13 detects the roughness of all portions of the surface to be inspected on the painted surface of the painted panel W in the form of numerical values (clear values). Then, the clear image value from the surface roughness meter 13 is input as the roughness information of the painted surface through (step 2).

The roughness information storage unit 19 stores the roughness information (including the undulation frequency component as described above) of the painted surface input in step 2 as sharpness data of the surface to be inspected. . Then, when all the storage processes are completed, a signal is sent to the device control unit 21, and the device control unit 21 ends the operation of the surface roughness meter 13 via the input / output device 18 (Step 3 above).

Next, the device control section 21 controls the input / output device 1
The light source 10 is turned on via the motor 8 and the motor 14 is driven to irradiate the inspection surface of the coating panel W with light. Then, the one-dimensional sensor 11 receives the reflected light from a predetermined range on the surface to be inspected, photoelectrically converts the reflected light into an image of the surface to be inspected, and outputs the image signal to the input / output device 18. (Step 4).

The image processing section 10 stores the signal output from the one-dimensional sensor 11 in step 4 as image data. When the storage process is completed, the image processing unit 20 outputs a signal to the device control unit 21, and the device control unit 21 further operates the motor 14 via the input / output device 18. Thereby, the one-dimensional sensor 11 receives reflected light from different portions of the surface to be inspected (above,
Step 5).

The device control section 21 determines whether or not image data of all portions of the surface to be inspected of the painted panel W has been taken by the one-dimensional sensor 11 based on a signal from the encoder 15. If the result of this determination is that image data has not been completely captured for the surface to be inspected, the processing from step 4 to step 6 is repeated. To the motor 14 via the input / output device 18.
Is stopped, and the process proceeds to the next step (step 6 above).

Next, the image processing section 20 processes all the image data stored in the processing of step 4 and step 5 and converts the data to the filter processing section 2 in the form of a video signal.
2 (step 7). The waveform of the video signal output from the image processing unit 20 in step 7 is, for example, as shown in FIG. 6 (c) (when the undulation is small) and FIG. ) (When the swell is large).

After inputting the video signal from the image processing unit 20 in step 7, the filtering unit 22 reads from the roughness information storage unit 19 the sharpness data (particularly the undulation Frequency component) is read (step 9).

Then, the filter processing section 22 associates the video signal from the image processing section 20 with the sharpness data from the roughness information storage section 19 and converts the video signal into a roughness component corresponding to the video signal. (Frequency component) is removed, and the signal after this filter processing is output to the binarization circuit of the flaw detection unit 23. By this filtering, the undulation component is removed from the video signal, and the unevenness of the signal is smoothed except for the flaw F (step 9). Looking at the waveform of the signal after the filter processing in step 9,
For example, the signal of FIG. 6C is corrected to a signal having a waveform as shown in FIG. 6E, and the signal of FIG. 6D is corrected to a signal having a waveform as shown in FIG.

When the signal subjected to the filtering process in step 9 is input to the binarizing circuit, the flaw detecting section 23 sets a threshold level for binarizing the signal. At this time, as shown in FIGS. 6 (e) and 6 (f), the signal after the filter processing was substantially smoothed except for the flaw F portion regardless of the magnitude of the undulation of the painted surface of the painted panel W. Because of the waveform, the threshold level can always be set high (step 10).
As a result, even when the undulation of the painted surface is large, small scratches can be detected.

Then, the flaw detector 23 binarizes the video signal after the filtering process in the binarization circuit with a threshold level (step 11).

The flaw detection section 23 determines the presence or absence of flaws on the painted surface based on the binarized signal, and outputs the result to the equipment control section 21. Then, the device control unit 21
The display device 16 is operated via the input / output device 18 to display the inspection result on the display device 16 (step 1).
2).

As described above, according to the present embodiment, the roughness component (undulation frequency component) of the inspected surface of the painted surface is measured by the surface roughness meter 13.
And the filtering unit 22 removes the roughness component from the video signal before binarizing the video signal on the surface to be inspected, so that the video signal is removed regardless of the roughness of the painted surface. Is smoothed except for the flaw F, and the threshold level for binarization can always be set high. Therefore, even when the painted surface has a large roughness, a small flaw can be detected, and the accuracy of detecting the flaw on the painted surface is improved.

In this embodiment, a painted panel is exemplified as an example of a work. However, the present invention is not limited to this, and a surface defect inspection apparatus according to the present invention can be used as long as the surface roughness can be detected by a signal. Is possible.

Also, an example has been shown in which image processing is performed after storing all image data on the surface to be inspected of the workpiece, but the image data obtained by the one-dimensional sensor 11 is configured to be processed in real time. Of course, it may be possible.

[0032]

As is apparent from the above description, according to the present invention, it is possible to reliably detect a small surface defect regardless of the roughness of the surface to be inspected on the work surface. Inspection accuracy of surface defects is improved.

[Brief description of the drawings]

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

FIG. 2 is an operation flowchart of the apparatus of FIG. 1;

FIG. 3 is a schematic view showing a conventional surface defect inspection apparatus.

FIG. 4 is a block diagram showing an operation procedure of the device shown in FIG. 3;

FIG. 5 is a diagram illustrating an example of each of a video signal and a binary signal.

FIG. 6 shows examples of video waveforms in various cases.

[Explanation of symbols]

 11: one-dimensional sensor (image input device) 13: surface roughness meter (surface roughness detection means) 19: roughness information storage section (storage means) 22: filter processing section (filter means) 23: scratch detection section (judgment) Means) W ... Painted panel (work) F ... Scratch (surface defect)

Claims (1)

(57) [Claims] 1. A surface roughness detecting means for detecting the roughness of a surface to be inspected on a surface of a work; a memory means for storing roughness information detected by the surface roughness detecting means; Image input means for inputting an image, filter means for removing a roughness component stored in the storage means from an image signal output from the image input means, and an output signal from the filter means, Determining means for determining the presence or absence of a surface defect on the surface to be inspected.