JP3412366B2 - Coating smoothness inspection device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for inspecting the smoothness of a painted surface, for example, an apparatus for inspecting the smoothness of a painted surface of an automobile body by image processing.

[0002]

2. Description of the Related Art As a conventional coating film smoothness inspection apparatus, there is one disclosed in Japanese Patent Laid-Open No. 1-213509. In this method, a predetermined light-dark stripe pattern is projected on the surface to be inspected, and the smoothness is detected by utilizing the irregularity of the light-dark pattern due to the unevenness of the orange peel on the surface to be inspected.
Specifically, a predetermined light / dark pattern image projected on the surface to be inspected is captured, and the light / dark pattern of the obtained light-receiving image is binarized,
The smoothness of the painted surface to be inspected is determined by the shape of the bright or dark part of the bright and dark pattern formed by superimposing a predetermined pattern on it. Further, when the surface to be inspected is a curved surface, the above-mentioned shape variation is calculated separately for a flat surface portion, a curved surface, and a portion, and determination is made using the average value of both.

The above-mentioned yuzu skin refers to the following states. That is, since the concentration of the paint is not strictly constant due to eddy convection that occurs in the process of evaporation of the paint solvent, extremely thin (low) unevenness is periodically generated in the thickness of the coating film. This unevenness has, for example, a gap between mountains of 0.
It is about 1 to 1 cm, and the height of the irregularities is about several μm. Such extremely thin irregularities are usually not noticeable and do not become defects. However, since it may look like "Yuzu" or the surface of orange under the control of light, it is so-called "Yuzu skin" or "Orange skin".

[0004]

However, the conventional coating film smoothness inspection device as described above has the following problems. That is, in the conventional coating film smoothness inspecting apparatus, it is necessary to set the pitch of the light / dark pattern of illumination to about 0.15 cm in order to correspond to the size of the unevenness to be detected as described above. Considering the number of stripes reflected and the number of pixels (resolution) of the imaging means (video camera), the shooting range (field of view) of the video camera is 5 cm.
It is a narrow range of about × 5 cm. Therefore, it is difficult to inspect a wide range in a short time, and there is a problem that a large-area object such as a car body cannot be inspected efficiently. In addition, in the curved surface inspection method as described above, in the case of a complicated curved surface such as an automobile body,
There is also a problem that the inspection efficiency is deteriorated because the effect of the curved surface cannot be completely removed and the processing time becomes long.

The present invention has been made to solve the problems of the prior art as described above, and a first object of the present invention is to provide a coating film smoothness inspection apparatus capable of efficiently inspecting an object to be inspected in a large area. Is to provide. A second object is to provide a coating film smoothness inspecting device capable of inspecting accurately even on a curved surface.

[0006]

In order to achieve the above object, the present invention is constructed as described in the claims.

First, in the invention described in claim 1,
A predetermined light-dark pattern is projected on the surface to be inspected, and the received light image is converted into image data of electric signals. Then, a boundary area between a light portion and a dark portion in the image of the light-dark pattern is extracted from the image data, and the width of the light-dark boundary area in the binarized image obtained by binarizing the extracted boundary area is adjusted to the degree of variation (corresponding to the magnitude of the disturbance). The smoothness of the surface to be inspected is determined accordingly. The degree of smoothness is a degree determined by performing sensory evaluation in advance according to the quality required in the actual painting work, and is displayed in several stages such as "good, good, bad". Yes, or 5 grades from 1 to 5 or 1
It can be displayed in a 10-step evaluation of 10 to 10. The above configuration corresponds to, for example, the embodiment shown in FIG. 6 described later.

Further, in the invention described in claim 2,
Thinning means for thinning the binarized image in the light-dark boundary region to extract the center line of the image is provided, and the smoothness is determined using the width in the direction perpendicular to the center line in the light-dark boundary region as the width of the light-dark boundary region. It is configured to do.

Further, in the invention described in claim 3,
By providing expansion / contraction means for performing expansion / contraction processing in which the light / dark boundary area binarized by the binarizing means is once expanded and then contracted, the disturbed portion (the boundary portion of the light-receiving image of the stripe due to orange peel or the like is provided. Image is removed, and a logical operation is performed with the image binarized by the binarizing means (for example, exclusive OR is obtained).
Is performed, and the smoothness of the surface to be inspected is determined based on the number or the total area of the regions corresponding to the disturbed portions extracted as a result. The above configuration corresponds to, for example, the embodiment shown in FIG. 7 described later.

Further, in the invention described in claim 4,
Smoothing means for smoothing the light-dark boundary area obtained by the light-dark boundary extracting means is provided, and a logical operation is performed on the binarized image of the light-dark boundary area and the smoothed binarized image, and the result is extracted. The smoothness of the surface to be inspected is determined based on the number or the total area of the regions corresponding to the disordered portion. The above configuration corresponds to, for example, the embodiment shown in FIG. 8 described later.

According to the invention of claim 5,
The difference between the light-dark boundary area itself and the image with the light-dark boundary area smoothed (actually, the subsequent processing is easy if the absolute value of the difference is obtained)
And a smoothing of the surface to be inspected based on the number or total area of the regions corresponding to the disturbed portion obtained as a result of binarizing the detection result of the difference detecting unit with a predetermined threshold value. It is configured to determine sex. The above configuration corresponds to, for example, the embodiment shown in FIG. 9 described later.

According to the invention of claim 6,
An expansion / contraction means for binarizing the received light image and expanding / contracting the binarized received light image to expand and contract once is provided, and the binarized image and the expansion / contraction process described above are provided. The image is subjected to a logical operation, and the smoothness of the surface to be inspected is determined based on the number of areas or the total area corresponding to the disturbed portions extracted as a result. The above configuration corresponds to, for example, the embodiment shown in FIG. 10 described later.

According to the invention of claim 7,
A smoothing means for smoothing the received light image is provided, and a logical operation is performed between the binarized image of the received light image itself and the binarized image of the smoothed image, and as a result, the disturbed portion extracted is obtained. The smoothness of the surface to be inspected is determined based on the number or the total area of the corresponding regions. The above configuration corresponds to, for example, the embodiment shown in FIG. 11 described later.

Further, in the invention described in claim 8,
A smoothing means for smoothing the received light image and a difference detecting means for obtaining a difference between the received light image itself and the smoothing processing result are provided, and the disturbed portion obtained as a result of binarizing the detection result of the difference detecting means. The smoothness of the surface to be inspected is determined based on the number of areas or the total area corresponding to. The above configuration corresponds to, for example, the embodiment shown in FIG. 12 described later.

In the invention described in claim 9,
In the configuration according to any one of claims 1 to 8, the threshold value in the binarizing means is configured to use a predetermined value according to the coating color of the surface to be inspected.

Further, in the invention described in claim 10, in the structure described in any one of claims 1 to 8, the threshold value in the binarizing means is automatically set according to the physical quantity relating to the brightness of the received light image. It is provided with a means for setting.

As described above, in the present invention, the disturbance of the light-dark boundary region caused by minute irregularities on the surface of the coating film such as orange peel is detected from the received light image and the magnitude of the disturbance is detected.
The smoothness of the coating film surface is determined based on the number of disordered regions, the total area of the disordered regions, and the like. Therefore, unlike the conventional device, it is not necessary to detect the yuzu skin itself, so that the stripe pitch of the illumination device can be made larger than that of the conventional device. For example, the wavelength (pitch) of yuzu skin is generally about 0.1 to 1 cm, and in the conventional smoothness inspection device, the pitch of the light-dark pattern of illumination is set to about 0.15 cm in order to detect such yuzu skin. Therefore, considering the number of stripes in the received light image and the resolution of the video camera, the shooting range of the video camera is 5c.
It was a narrow range of about m × 5 cm (area about 25 cm ² ). However, in the present invention, since the smoothness is determined by the disturbance of the stripes as described above, the pitch of the stripes can be set to about 5 cm, and therefore the imaging range of the imaging means (video camera) is about 20 × 12c.
It has become possible to increase the size to about m (about 240 cm ^{2 in} area). This value is about 9 to 10 times the conventional value, and it becomes possible to quickly inspect a large area.

Further, as described above, since the smoothness of the coating film surface is judged by the size of the disturbance in the light-dark boundary area, the number of the disturbance areas, the total area of the disturbance areas, etc.,
Even if the surface to be inspected is a curved surface and the stripe width or pitch in the received light image is not constant, the smoothness can be accurately determined. In particular, according to the second aspect, in the configuration in which the determination is made by the width in the direction perpendicular to the center line, the smoothness can be accurately measured even when the surface to be inspected is a three-dimensional curved surface and the stripe image in the received light image is oblique. You can judge.

When the threshold value of the binarizing means is set according to the coating color as described in claims 9 and 10, when the brightness or contrast changes depending on the coating color. However, an accurate judgment is possible.

[0020]

DETAILED DESCRIPTION OF THE INVENTION The present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing an overall configuration according to an embodiment of the present invention. In FIG. 1, reference numeral 1 denotes an illuminating device, which is arranged so as to irradiate the surface 3 to be inspected with light of a predetermined bright and dark pattern. 2 is a video camera (eg CCD
A camera or the like), and is arranged so as to capture an image of the surface 3 to be inspected on which the bright and dark patterns are projected. Also, 4 is a camera control unit, here a video camera 2
An image signal of the received light image picked up in is generated and output to the image processing device 5. Reference numeral 6 denotes a host computer, which has a function of externally displaying or outputting control of the image processing apparatus 5 and processing results. A monitor 7 displays a screen imaged by the video camera 2.

Next, FIG. 2 is a block diagram showing the configuration of the image processing apparatus 5. In FIG. 2, the image signal from the camera control unit 4 is converted into a digital value by the A / D converter 9 via the buffer amplifier 8. Further, the MPU (microprocessor) 10 performs predetermined calculation, processing, etc. on the image data. Reference numeral 11 denotes a memory that stores image data and processing results.
It can be output to and displayed on the monitor 7 via the / A converter 12.

Next, FIG. 3 is an exploded perspective view showing the details of the illuminating device 1. In the present embodiment, description will be made assuming that bright and dark stripe pattern illumination is used. In FIG. 3, reference numeral 1a is a light source, and the light is diffused by the diffusion plate 1b and is irradiated on the surface to be inspected through the stripe plate 1c. The diffusion plate 1b is, for example, like frosted glass, and uniformly irradiates the surface 3 to be inspected with light. The stripe plate 1c is a transparent or diffuser plate having black stripes at predetermined intervals. A similar effect can be obtained by using a structure in which a stripe pattern is directly provided on the diffusion plate 1b without using the stripe plate 1c (for example, a matte black tape is attached in a stripe shape). By using the illumination device 1 as described above, a bright and dark (black and white) stripe pattern can be projected on the surface 3 to be inspected. The structure of the illuminating device is not limited to the above embodiment as long as it can project a stripe pattern on the surface to be inspected.

Next, the operation will be described. A light-receiving image obtained by imaging the surface 3 to be inspected on which the bright and dark stripes are projected is as shown in FIG. When the surface 3 to be inspected is a curved surface as shown in FIG. 1, even if the pitch of the stripe pattern of the illuminating device 1 is constant, the stripe pitch in the received light image is shown in FIG. 4 according to the curvature of the surface 3 to be inspected. It changes as shown. In addition, when the surface 3 to be inspected has minute irregularities such as orange peel, the boundary line of the stripe is disturbed as shown in FIG. In the light-receiving image of FIG. 4, the upper left corner of the screen is the origin (0, 0), and the coordinate axes x,
When y is determined, the density (luminance) cross section in the x direction at an arbitrary y value (for example, y = ya) is as shown in FIG.

As described above, the received light image shown in FIG. 4 is captured as the original image S0 in the image processing apparatus 5 by a signal as shown in FIG. For example, when the image signal from the video camera 2 is A / D converted and the brightness level is converted to a digital value of 8 bits, 255 is the maximum value on the vertical axis, which is white,
The minimum value of 0 is black. Also, one screen is 512 x 480
When captured at the pixel resolution, the horizontal axis represents the number of pixels of 0 to 512.

In the experiment conducted by the present inventor, the shooting range (field of view) of the video camera 2 is about 20 cm in the x direction and y.
The direction was about 12 cm (area about 240 cm ² ), and the stripe pitch was about 5 cm. Generally, the wavelength (pitch) of yuzu skin is about 0.1 to 1 cm, and in order to detect the yuzu skin as described above with the conventional smoothness inspection device, the pitch of the stripe should be equal to or less than that of the yuzu skin. Had to do. However, in the present invention, since the smoothness is detected by a principle different from the conventional one, as described above, it is possible to use stripes having a pitch significantly larger than the pitch of orange peel. Therefore, the field of view of the video camera 2 is also 5 cm × 5 cm (about 25 c in area).
Since the area ratio can be increased from about m ² ) to about 9 to 10 times as described above, it is possible to quickly inspect a large area.

The contents of image processing in the image processing apparatus will be described below. 6 to 12 are block diagrams showing an embodiment of image processing according to the present invention. First, the image processing shown in FIG. 6 will be described. This embodiment corresponds to the structure of claim 1. Note that FIG.
FIG. 14 is a diagram showing images at various processing stages in image processing, which will be described below with reference to FIG. Referring to FIG. 6, first, in step s1, the video camera 2 is used as shown in FIG.
The original image shown in 3 (a) is input.

Next, in step s2, the light-dark boundary area of the stripe is extracted (a method of extracting the light-dark boundary area will be described later), and binarized in step s3 to obtain the result shown in FIG.
Obtain the image of 3 (b). Next, in step s4, the white area (bright / dark boundary area) of the image (b) is labeled (labeled: numbered in order), and the width d of each white area is calculated in step s5. At this time, the surface to be inspected 3
The poorer the smoothness of, the larger the amount of disturbance in the white area (bright / dark boundary area), and thus the greater the change in the width d. That is,
As shown in FIG. 14, the poorer the smoothness of the coating film on the surface 3 to be inspected (the greater the degree of the orange peel skin), the larger the difference between the narrow and wide portions of the light-dark boundary region becomes, and the dispersion of the above-mentioned widths is increased. As shown in FIG. 15, the relationship between and is smoothing upward.

Actually, for example, in the image of FIG. 13B, the width d in the x direction is calculated for 480 lines, and the width d
The smoothness can be obtained from the result and the characteristic previously obtained in FIG. 15 by performing a statistical process for obtaining the variance. The above process is performed in step s6. Note that FIG.
The relationship between the dispersion of the width d and the smoothness shown in (3) may be obtained in advance by a sensory evaluation using a yuzu skin test piece by a plurality of inspectors. Further, the degree of smoothness is a degree determined by performing a sensory evaluation in advance according to the quality required in the actual painting work, for example, “good,
Display in several steps such as “Yes, No” or 1 to 5
It can be displayed in a five-level evaluation of 1 or a 10-level evaluation of 1 to 10. By performing the image processing as described above, even if the surface 3 to be inspected is a curved surface as shown in FIG. 2 and the stripe width or pitch of the received light image is not constant, the smoothness can be quickly measured. Can be done.

Next, the case where the surface 3 to be inspected is a three-dimensional curved surface will be described. This corresponds to the structure of claim 2.
FIG. 16 is a diagram showing processed images at various stages in image processing when the surface 3 to be inspected is a three-dimensional curved surface. In FIG. 16, (a) is a light-receiving image, (b) is an extraction result of a light-dark boundary region, (c) is an enlarged view of the light-dark boundary region, (d) is a thinning processing result of (c). Is.

When the surface 3 to be inspected is a three-dimensional curved surface,
The received light image appears distorted as shown in FIG. In this case, as shown in FIG. 16C, the true width of the light-dark boundary area is d ′, but since the width of the screen in the x direction is detected as described above, the x direction can be detected. Will be erroneously detected as the width d of the light-dark boundary area. In this case, FIG.
As can be seen from (c), since d '<d, it is judged that the smoothness is worse than the actual smoothness. In order to avoid such a state, as shown in FIG.
The extracted light / dark boundary area may be thinned to extract the center line, and the width d ′ in the direction orthogonal to the direction of the center line in the light / dark boundary area in FIG. 16C may be detected. This processing is performed in the width calculation processing in step s5 of FIG.

The above-described thinning processing is a processing method generally used in image processing, and is processing for obtaining the center line of a figure as a line in which a line width of 1 pixel is connected. Note that the smoothness of the coating surface was very poor, and the disturbance in the light-dark boundary area was large and complicated, so the center line obtained by the thinning process was disturbed and the accurate direction could not be recognized. An accurate direction can be obtained by performing processing for smoothing the center line, for example, moving average processing. Alternatively, expansion / contraction processing is performed on the binarized image of the light / dark boundary area (described later in FIG. 7).
It is also possible to reduce the turbulence by applying the above, and perform thinning processing on the result to obtain the center line. Further, by repeating the expansion / contraction process described above, the disorder in the light-dark boundary region is almost eliminated, and if it becomes an oblique straight line,
It is possible to find the direction without thinning.

Next, an example of the bright / dark boundary extraction processing in step s2 will be described. FIG. 17 is a diagram showing a luminance signal waveform and an image during the light-dark boundary extraction processing. Figure 1
In FIG. 7, when the original image S0 is subjected to a predetermined smoothing process (for example, passed through a simple averaging filter) to remove minute noise, an image S1 with slightly blurred edges is obtained. Next, a predetermined differentiation (or difference) process (for example, sobel
When the absolute value of the result (specifically, the sum of the absolute values in the x and y directions or the square root of the sum of the squares) is calculated, the area where the brightness changes (edge component)
Only the image S2 can be extracted. This is binarized with a predetermined threshold value, and for example, a binary image S3 in which the light-dark boundary area is white (level 255) and the other background is black (level 0) is created. In this way, the bright and dark boundary areas can be extracted. The method of extracting the light and dark boundaries is not limited to the above method.

Next, the image processing shown in FIG. 7 will be described. This embodiment corresponds to the structure of claim 3. Note that this case will be described with reference to FIG. In FIG. 7, the processing of steps s1 to s3 is the same as that of FIG. Next, in step s7, expansion / contraction processing is performed on the extracted and binarized light-dark boundary area.
This expansion processing means that at least one graphic pixel (white with a brightness of 1: level 25) in the vicinity Q1 to Q8 of a certain background pixel Q0 (black with a brightness of 0: level 0) of an image as shown in FIG.
5), the background pixel Q0 is converted to white with a brightness of 1 when there is. Therefore, the connected portion of the graphic image in the image expands outward by one pixel, that is, expands. In this expansion processing, the concave portion of the light-dark boundary area waveform changes to white with a brightness of 1 first, so the turbulent waveform is gradually shaped by repeating the expansion processing. Next, the contraction process is the reverse of the above process, and is a process of changing a graphic pixel into a background pixel, and the connected portion of the graphic pixel is contracted inside by one pixel.

By performing the expansion processing as described above, the disorder of the bright-dark boundary region (the disorder of the white line in FIG. 13b) due to the orange peel skin is reduced, and the linear bright-dark boundary region is obtained.
It should be noted that the number of times of expansion processing should be experimentally obtained in advance by using a sample in which the largest amount of yuzu skin occurs in the expected range and performing the processing until there is no disturbance in the light-dark boundary area. Good. Next, by repeating the contraction processing the same number of times as the expansion processing described above, as shown in FIG. 13C, a linear shape having a width substantially the same as that of the light-dark boundary area before the expansion / contraction processing is performed. A light-dark boundary area is obtained.

Next, in step s8, as shown in FIG.
By performing a logical operation for obtaining the exclusive OR (EX-OR) of the images of (b) and (c), it is possible to extract only the portion of the disorder due to the orange peel as shown in FIG. 13 (d). . The larger the wavelength and height of the unevenness of the orange peel skin (the poorer the smoothness of the coating film surface), the larger the disturbance of the light-dark boundary region of the stripe, and the larger the number and total area of the extraction regions in FIG. There is a relationship. Therefore, the relationship between the degree of yuzu skin (smoothness of the coating film surface) and the number or total area of the above-mentioned extracted regions is obtained in advance by experiments and quantified, and the number or total area of the above-mentioned extracted regions is obtained and applied. It is possible to judge the smoothness of the coating film surface. In FIG. 7, in step s9, labeling is performed on each extraction region of FIG. 13 (d), in step s10, the number of extraction regions or the total area is calculated, and in step s11, the process of applying the above experimental results. The smoothness is determined by performing.

Next, a case where the number of stripes displayed on the screen changes depending on the curvature of the surface 3 to be inspected in the above-mentioned smoothness judging method will be described. As described above, in the processing of steps s10 and s11, in the case of the method of determining the smoothness based on the number of extraction regions in FIG. 13D, it is desirable that the number of stripes per screen is the same. The reason for this is that the number of light-dark boundary areas changes depending on the number of stripes per screen, so the number of extraction areas also increases as the number of stripes per screen increases, even if the smoothness is the same, and the smoothness determination result This is because there is a possibility that an error may occur in the.

As an example of the measures for preventing the above-described error from occurring, a method of making a judgment on a fixed number of light-dark boundary areas in one screen will be described. FIG. 13
In (3), when (d) is obtained by performing the logical operation of (b) and (c), it is performed only for a predetermined area, for example, the two central areas (b ₁ , b ₂ and c ₁ , c ₂ ) every time. By doing so, even if the number of stripes displayed on the entire screen changes due to the curved surface as shown in FIG. 19, the processing is always performed only on a certain area of the two central areas, so that an accurate determination is made every time. You can Further, the number of processing regions per screen may be determined in advance from the maximum curvature of the surface to be inspected and the minimum number of stripes on the screen. The area may be specified by specifying the label number of the light / dark boundary area or by a position in the screen such as "center of screen". As another method, the curvature of the surface to be inspected may be calculated from the number and width of stripes shown in the received light image, and the number of processing areas may be determined or the processing areas may be designated based on the curvature. Alternatively, the number of extraction regions may be directly corrected based on the calculated curvature.

Next, the image processing shown in FIG. 8 will be described. This embodiment corresponds to the structure of claim 4. Note that this case will be described with reference to FIG. In FIG. 8, the processing of steps s1 to s3 is the same as that of FIG. Next, in step s12, the light-dark boundary area is smoothed, and in step s13, the binarization is performed, and the result is used to perform the calculation after step s8 in FIG. The smoothing process in the step s12 is a process for flattening the disturbance of the bright and dark boundaries due to the orange peel skin. If this result is binarized, it is as shown in FIG. 13 (c) like the expansion / contraction process. An image is obtained.

The specific contents of the smoothing process are, for example, a maximum value filter and a minimum value filter. The maximum value filter is, for example, in the case of a 3 × 3 size as shown in FIG. 18, that sets the maximum value of the brightness values of the nine pixels Q0 to Q8 as the new brightness of Q0, and the minimum value. On the contrary, the filter uses the minimum value as the new luminance. Such maximum value filtering is performed in S2 of FIG.
When applied to such a grayscale image, a region having a larger brightness value than the surroundings spreads, so that the light-dark boundary region spreads around. At this time, since the concave / convex portion of the light / dark boundary region waveform is changed to a region having a higher luminance first, the disturbed waveform is gradually shaped by repeating the maximum value processing. This process is repeated until the disorder due to Yuzu skin is completely filled. Next, the minimum value filter processing is performed the same number of times as the maximum value filter processing described above to return the width of the region to almost the original value, and then binarized to obtain a straight line as shown in FIG. A light and dark boundary area is obtained. That is, even if the maximum value / minimum value filter processing is performed, the same result as the expansion / contraction processing can be obtained.

Next, the image processing shown in FIG. 9 will be described. This embodiment corresponds to the structure of claim 5. Note that this case will be described with reference to FIG. In FIG. 9, the processes in steps s1 and s2 are as follows.
This is the same as in FIG. Next, in step s14, smoothing processing (similar to step 12 in FIG. 8) is performed on the light-dark boundary area obtained in step s2, and step s15
Then, the absolute value of the difference between the smoothing processing result and the light-dark boundary area obtained in step s2 is obtained. This result shows that the brightness value at the part where the light-dark boundary is disturbed remains higher than that at other parts, so if the value is binarized at a predetermined threshold value in step s16, the image as shown in FIG. can get. After that, the smoothness can be determined by performing the processing of steps s9 to s11 similarly to the above. It should be noted that in step s15, the difference between the two may be simply obtained, but in that case the post-processing becomes complicated because it is necessary to pay attention to the sign. Therefore, the post-processing is facilitated by obtaining the absolute value of the difference as described above.

Next, the image processing shown in FIG. 10 will be described. This embodiment corresponds to the structure of claim 6. This case will be described with reference to FIG. In FIG. 10, if the bright / dark boundary extraction processing in step s2 is not performed, but the binarization processing in step s3 is immediately performed on the image input in step s1, the image in FIG. 20 (a) is obtained. Next, in step s17, expansion / contraction processing is performed on either the white area or the black area of the binarized image. For example, if the target is a black area, the black area is first expanded to eliminate the distortion due to orange peel skin (disturbance of the boundary area between white and black), and then the contraction processing is performed to reduce the width of the black area. Is returned to almost the original size, an image as shown in FIG. 20 (b) is obtained. Next, in step s8, the logical operation (for example, EX-OR) between the image after expansion / contraction and the binarized image in step s3 is performed, and the result of FIG.
An image of only the disordered area as shown in FIG.
The same image) is obtained. After that, the smoothness can be determined by performing the processing of steps s9 to s11 in the same manner as described above.

Next, the image processing shown in FIG. 11 will be described. This embodiment corresponds to the structure of claim 7. Note that this case will also be described with reference to FIG. Figure 11
In step 1, the input image input in step s1 is binarized in step s3, the input image input in step s1 is smoothed in step s18, and then binarized in step s19. The logical operation of s8 is performed. The above-mentioned smoothing processing is processing for flattening the disturbance of the light-dark boundary area due to orange peel skin, and the specific content is the same as the smoothing processing of step s12 in FIG. Since the image as shown in FIG. 20C is obtained also by such processing, it is possible to determine the smoothness by performing the processing of step s9 and subsequent steps as described above.

Next, the image processing shown in FIG. 12 will be described. This embodiment corresponds to the structure of claim 8. Note that this case will also be described with reference to FIG. 12
In step 1, the absolute value of the difference between the result of smoothing the input image input in step s1 in step s20 and the input image itself is obtained (also in this case, the difference between the two may be simply obtained as in FIG. 9). . This result shows that the luminance value at the portion where the light-dark boundary is disturbed remains higher than that at the other portions, so if the threshold value is binarized at a predetermined threshold value in step s22,
An image such as 0 (c) is obtained. After that, the smoothness can be determined by performing the processing of steps s9 to s11 similarly to the above.

Next, an embodiment for preventing the influence of the coating color on the surface to be inspected will be described. This embodiment corresponds to the structure of claim 9. If the brightness difference between the bright portion and the dark portion in FIG. 17 (S1), that is, the contrast, changes depending on the coating color of the surface 3 to be inspected, FIG.
Since the differential level in the extraction result of the light-dark boundary region shown in 2) also changes, if this is binarized with a fixed threshold value, the width d of the light-dark boundary region changes depending on the coating color. As a first method for preventing this, a threshold value suitable for the coating color is previously obtained by an experiment and set to the threshold value corresponding to the coating color.

The second method is to recognize the coating color from the received light image and automatically set the binarization threshold value to the optimum value accordingly. This embodiment corresponds to the structure of claim 10. This method will be described below. As shown in FIG. 21, the brightness and the contrast are different between the light paint color (white or silver metallic) and the dark color (black, navy blue, gun metallic, etc.),
The higher the brightness of the paint itself, the higher the brightness value as a whole and the smaller the contrast. Therefore, when these are differentiated (actually, a difference filter such as a sobel filter is used) and the absolute value thereof is obtained, it becomes as shown in FIG. 22, and it is necessary to change the binarization threshold value according to the difference in contrast. I understand. Note that, in FIG. 22, the broken line indicates the characteristic of high contrast, and the solid line indicates the characteristic of low contrast. Further, ThL is a threshold value suitable for high contrast, and ThD is a threshold value suitable for low contrast.

As a method of changing the binarization threshold value according to the difference in contrast, a physical quantity relating to the luminance of the original image as shown in FIG. 21, for example, the maximum value or average value of the luminance in the image is obtained, A method is conceivable in which the coating color is determined according to the size and the threshold value is determined. Alternatively, it is also possible to obtain the contrast (brightness / darkness difference) of the stripes from the maximum value and the minimum value in the luminance characteristic of FIG. 21 and determine the threshold value accordingly. With the above-mentioned configuration, even if the coating color is different, the optimum threshold value can be automatically set based on the luminance information of the received light image, and the accurate smoothness can always be determined.

[0047]

As described above, according to the present invention, the disturbance of the light-dark boundary region caused by the minute unevenness of the coating film surface such as orange peel is detected from the received light image, and the magnitude of the disturbance is detected. Since the smoothness of the coating film surface is determined by the number of disordered regions, the total area of the disordered regions, etc., the pitch of the stripes of the illuminating device can be made larger than that of the conventional device. It is now possible to significantly increase the shooting range of. Therefore, it becomes possible to inspect a large area quickly.
The effect is obtained.

Further, as described above, the smoothness of the coating film surface is determined by the size of the disturbance in the bright / dark boundary area, the number of the disturbance areas, the total area of the disturbance areas, etc.
Even if the surface to be inspected is a curved surface and the stripe width or pitch in the received light image is not constant, the smoothness can be accurately determined.

Further, in the case where the threshold value of the binarizing means is set according to the color of the coating, accurate determination can be performed even when the brightness or contrast changes depending on the color of the coating. The effect is obtained.

[Brief description of drawings]

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

FIG. 2 is a block diagram of an example of the image processing apparatus 5 in FIG.

3 is an exploded perspective view of an example of the lighting device 1 in FIG.

FIG. 4 is a diagram showing a striped light-receiving image.

5 is a signal waveform diagram showing the image signal of FIG.

FIG. 6 is a flowchart showing the processing contents in the first embodiment.

FIG. 7 is a flowchart showing the contents of processing in the second embodiment.

FIG. 8 is a flowchart showing the processing contents in the third embodiment.

FIG. 9 is a flowchart showing the processing contents in the fourth embodiment.

FIG. 10 is a flowchart showing the processing contents in the fifth embodiment.

FIG. 11 is a flowchart showing the processing contents in the sixth embodiment.

FIG. 12 is a flowchart showing the processing contents in the seventh embodiment.

FIG. 13 is a diagram showing images at various processing stages in image processing.

FIG. 14 is a diagram showing the relationship between the irregularity of the width d of the bright-dark boundary region and the smoothness.

FIG. 15 is a characteristic diagram showing the relationship between the variance of the width d of the light-dark boundary region and the smoothness.

FIG. 16 is a diagram showing processing when a light-dark boundary region is inclined due to a three-dimensional curved surface.

FIG. 17 is a diagram showing signal waveforms and images at various processing stages in other image processing.

FIG. 18 is a diagram showing pixel positions in expansion / contraction processing.

FIG. 19 is a diagram showing a state in which the number of stripes reflected on the entire screen changes due to a curved surface.

FIG. 20 is a diagram showing images at various processing stages in other image processing.

FIG. 21 is a signal waveform diagram showing changes in brightness and contrast depending on the brightness of the coating color.

FIG. 22 is a diagram showing a waveform and a threshold value of a differentiation result due to changes in luminance and contrast according to the brightness of a coating color.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Illumination device 5 ... Image processing device 1a ... Light source 6 ... Host computer 1b ... Diffusion plate 7 ... Monitor 1c ... Stripe plate 8 ... Buffer amplifier 1d ... Background 9 ... A / D converter 2 ... Video camera 10 ... MPU 3 ... Surface to be inspected 11 ... Memory 4 ... Camera control unit 12 ... D / A converter

─────────────────────────────────────────────────── ─── Continuation of the front page (58) Fields surveyed (Int.Cl. ⁷ , DB name) G01B 11/00-11/30 G01N 21/88

Claims (10)

(57) [Claims] 1. A coated surface to be inspected is irradiated with light, a light-receiving image is created based on the reflected light from the surface to be inspected, and the smoothness on the surface to be inspected is detected based on this light-receiving image. In the coating film smoothness inspection apparatus, an illuminating means for displaying a predetermined light and dark pattern on the surface to be inspected, and an imaging means for converting a light-receiving image of the light and dark pattern obtained by imaging the surface to be inspected into image data of an electric signal, A bright / dark boundary extracting means for extracting a boundary area between a bright portion and a dark portion in the image of the light / dark pattern from the image data, and the image of the bright / dark boundary area obtained by the bright / dark boundary extracting means are binarized by a predetermined threshold value. And a judging unit for judging the smoothness of the surface to be inspected according to the degree of variation in the width of the light-dark boundary region in the binarized image of the binarizing unit. Membrane smoothness inspection device. 2. A thinning means for thinning the binarized image of the light-dark boundary region to extract the center line of the image, wherein the width of the light-dark boundary region in the direction perpendicular to the center line is the width of the light-dark boundary region. The coating film smoothness inspection apparatus according to claim 1, wherein the smoothness is determined as. 3. A coated surface to be inspected is irradiated with light, a light receiving image is created based on the reflected light from the surface to be inspected, and the smoothness on the surface to be inspected is detected based on this light receiving image. In the coating film smoothness inspection apparatus, an illuminating means for displaying a predetermined light and dark pattern on the surface to be inspected, and an imaging means for converting a light-receiving image of the light and dark pattern obtained by imaging the surface to be inspected into image data of an electric signal, A bright / dark boundary extracting means for extracting a boundary area between a bright portion and a dark portion in the image of the light / dark pattern from the image data, and the image of the bright / dark boundary area obtained by the bright / dark boundary extracting means are binarized by a predetermined threshold value. Binarization means, expansion / contraction means for expanding / contracting the light / dark boundary area binarized by the binarization means and then contracting it, and binarized by the binarization means. Image and expansion / recovery above Determination means for performing a logical operation with the processed image and determining the smoothness of the surface to be inspected on the basis of the number or total area of the regions corresponding to the disturbed portions extracted as a result. Coating smoothness inspection device. 4. A coated surface to be inspected is irradiated with light, a light receiving image is created based on the reflected light from the surface to be inspected, and the smoothness on the surface to be inspected is detected based on this light receiving image. In the coating film smoothness inspection apparatus, an illuminating means for displaying a predetermined light and dark pattern on the surface to be inspected, and an imaging means for converting a light-receiving image of the light and dark pattern obtained by imaging the surface to be inspected into image data of an electric signal, The light-dark boundary extracting means for extracting a light-dark boundary area in the light-dark pattern image from the image data, and the light-dark boundary area image obtained by the light-dark boundary extracting means are binarized by a predetermined threshold value. A first binarizing means; a smoothing means for smoothing the light-dark boundary area obtained by the light-dark boundary extracting means; and a second binarizing the smoothed image with a predetermined threshold value. Binarizing means of the above, and the first A logical operation is performed between the processing result of the binarizing means and the processing result of the second binarizing means, and the surface to be inspected is smoothed based on the number or total area of the regions corresponding to the disturbed portions extracted as a result. A coating film smoothness inspecting device comprising: a determination unit for determining the property. 5. A coated surface to be inspected is irradiated with light, a light receiving image is created based on the reflected light from the surface to be inspected, and the smoothness on the surface to be inspected is detected based on this light receiving image. In the coating film smoothness inspection apparatus, an illuminating means for displaying a predetermined light and dark pattern on the surface to be inspected, and an imaging means for converting a light-receiving image of the light and dark pattern obtained by imaging the surface to be inspected into image data of an electric signal, A light-dark boundary extracting means for extracting a boundary area between a light portion and a dark portion in the image of the light-dark pattern from the image data, a smoothing means for smoothing the light-dark boundary area obtained by the light-dark boundary extracting means, and the light-dark boundary. Difference detecting means for obtaining a difference between the bright / dark boundary region obtained by the extracting means and the processing result of the smoothing means; and binarizing means for binarizing the detection result of the difference detecting means with a predetermined threshold value. , The above binarization A coating film smoothness inspection apparatus comprising: a determination unit that determines the smoothness of a surface to be inspected based on the number or total area of regions corresponding to the disordered portion obtained as a processing result of the unit. 6. A coated surface to be inspected is irradiated with light, a light receiving image is created based on the reflected light from the surface to be inspected, and the smoothness on the surface to be inspected is detected based on this light receiving image. In the coating film smoothness inspection apparatus, an illuminating means for displaying a predetermined light and dark pattern on the surface to be inspected, and an imaging means for converting a light-receiving image of the light and dark pattern obtained by imaging the surface to be inspected into image data of an electric signal, Binarization means for binarizing the received light image with a predetermined threshold value, and expansion / contraction means for expanding / contracting the image binarized by the binarization means and then contracting the same. And a logical operation is performed on the image binarized by the binarizing unit and the image after the expansion / contraction processing, and based on the number or total area of regions extracted as a result of the disordered portion. Judgment for judging the smoothness of the surface to be inspected Coating smoothness inspection apparatus characterized by comprising a means. 7. A coated surface to be inspected is irradiated with light, a light-receiving image is created based on the reflected light from the surface to be inspected, and the smoothness on the surface to be inspected is detected based on this light-receiving image. In the coating film smoothness inspection apparatus, an illuminating means for displaying a predetermined light and dark pattern on the surface to be inspected, and an imaging means for converting a light-receiving image of the light and dark pattern obtained by imaging the surface to be inspected into image data of an electric signal, A first binarizing means for binarizing the received light image with a predetermined threshold value, a smoothing means for smoothing the received light image, and an image after the smoothing processing with a predetermined threshold value. A second binarizing means for binarizing, and a logical operation of the processing result of the first binarizing means and the processing result of the second binarizing means are performed, and the disturbed portion extracted as a result is Determine the smoothness of the surface to be inspected based on the number of corresponding areas or the total area. Coating smoothness inspection apparatus characterized by comprising: a constant-determining means. 8. A coated surface to be inspected is irradiated with light, a light receiving image is created based on the reflected light from the surface to be inspected, and the smoothness on the surface to be inspected is detected based on this light receiving image. In the coating film smoothness inspection apparatus, an illuminating means for displaying a predetermined light and dark pattern on the surface to be inspected, and an imaging means for converting a light-receiving image of the light and dark pattern obtained by imaging the surface to be inspected into image data of an electric signal, Smoothing means for smoothing the received light image, difference detection means for obtaining the difference between the received light image and the processing result of the smoothing means, and the detection result of the difference detection means are binarized with a predetermined threshold value. And a determining means for determining the smoothness of the surface to be inspected based on the number or the total area of the regions corresponding to the disturbed portion obtained as the processing result of the binarizing means. Coating smoothness inspection device characterized by . 9. The threshold value in the binarizing means is a value that is predetermined according to the color of the coating on the surface to be inspected. Coating smoothness inspection device. 10. The apparatus according to claim 1, further comprising means for automatically setting a threshold value in the binarizing means in accordance with a physical quantity relating to the brightness of the received light image.
The coating film smoothness inspection device according to any one of 1.