JP2016109532A5 - - Google Patents

Description

Surface evaluation method and surface evaluation apparatus

  The present invention relates to a technique for evaluating aspects of various products using an image processing technique.

  The surface of the product, for example, the outer peripheral surface of the piston ring, is processed as a smooth polished surface by a polishing process. Conventionally, the inspection of the outer peripheral surface of the piston ring has been performed visually or by image processing technology.

  For example, Patent Document 1 applies image processing technology to inspect the outer peripheral surface of a piston ring, but the inspection is not a simple method, and there is a problem with a complicated processing program and processing speed. is there.

  In Patent Document 2, a painted surface of an automobile or the like is an object to be inspected, a light / dark pattern is irradiated on the painted surface, a brightness difference of a light / dark image is obtained from a photographed image, and the unevenness height or depth of the painted surface is determined from this brightness difference. It is detected. This technique for inspecting a painted surface is an inspection of irregularities on a surface with a gentle undulation, and is not an evaluation of a smooth surface.

Japanese Patent Laid-Open No. 9-257438 JP 2004-191070 A

  Any of the above-mentioned techniques uses imaging and image processing techniques, but it is not a simple technique and cannot be applied as it is as an evaluation of an outer peripheral surface such as a piston ring, particularly a polished surface.

  Therefore, the problem to be solved by the present invention is that image processing technology can be applied to appropriately evaluate the surface of a product such as a piston ring, particularly the polished surface on the outer periphery of the piston ring, in relation to other good standard polished surfaces. Is to do so.

Based on the above problems, the present invention generates reflected light corresponding to light and dark on the surface by irradiating the light to be evaluated with illumination light from a light source via a filter of a light and dark pattern. The reflected light is photographed by the camera, the brightness difference between the bright image brightness and the dark image brightness is obtained from the photographed image by the image processing unit, and the threshold value corresponding to the value of the brightness difference and a good standard surface is obtained. And the quality of the surface is evaluated from the comparison result.

In addition, according to the present invention, the surface is determined to be defective when the expression ( luminance difference value <threshold value) is obtained from the comparison result by the image processing unit.

In particular, according to the present invention, an evaluation target surface is a polished surface on the outer periphery of a piston ring, and an average luminance difference is obtained from an average value of a plurality of bright image luminances and an average value of a plurality of dark image luminances by an image processing unit. Seeking.

  In the present invention, at the time of comparison by the image processing unit, at least one of the expressions (bright brightness value> bright lower limit brightness value) and (dark brightness value <dark upper limit brightness value) is compared. By weighting, the inspection is stabilized.

According to the present invention, the value of the brightness difference between the bright image brightness and the dark image brightness is compared with the threshold value corresponding to the good standard surface. Therefore, the evaluation target surface and the good standard surface of the good product are compared. The luminance difference between the two can be converted into a numerical value and can be evaluated efficiently, and an appropriate quality evaluation can be performed. As a result, the visual work can be automated and labor-saving, and the evaluation process can be simplified as compared with the conventional image processing technique.

Further, according to the present invention, when comparing the formulas ( brightness difference value <threshold value), the threshold value can be appropriately set according to the actual good standard surface, and therefore, erroneous evaluation is reduced.

In particular, according to the present invention, since the average luminance difference is used according to the polished surface on the outer periphery of the piston ring, the evaluation is stable without being affected by irregular luminance and the specific luminance such as noise. , Erroneous evaluation can be reliably prevented.

In addition, according to the present invention, it is possible to weight the comparison of at least one of the equation (bright luminance value> bright lower luminance value) and the equation (dark luminance value <dark upper luminance value). When the equation ( luminance difference value <threshold value) is satisfied, the necessity of calibration of the threshold value can be predicted from the evaluation result of the lower limit luminance and the upper limit luminance, and the inspection can be stabilized. .

It is a front view of the optical system of the surface evaluation apparatus. It is a top view of the positional relationship of a piston ring and a light-and-dark pattern in the surface evaluation apparatus. It is explanatory drawing of the picked-up image of a camera in the surface evaluation apparatus. It is a graph of the brightness | luminance change of a picked-up image.

  In FIG. 1, a surface evaluation apparatus 1 according to the present invention includes a light source 2, a filter 3, a camera 4, and an image processing unit 5, and a product, for example, an outer peripheral surface 7 of a piston ring 6 is an evaluation target. The surface 7 is a polished surface. When the polished surface is evaluated, the piston ring 6 is held by a suitable work holder (not shown), and rotates at a constant speed with the center line of the piston ring 6 as the center of rotation.

  The light source 2 irradiates the surface 7 of the piston ring 6 with uniform illumination light 8. The filter 3 is interposed between the light source 2 and the surface 7, and forms one or a plurality of, for example, one dark line in the generatrix direction of the surface 7. For this reason, the illumination light 8 passes through the filter 3 and is irradiated onto the surface 7, thereby forming a light / dark pattern on the surface 7. Since the surface 7 acts as a convex mirror, the light / dark pattern of the surface 7 by the illumination light 8 is reduced in the circumferential direction.

  As shown in FIG. 2, one dark line on the surface 7 is formed in the direction parallel to the generatrix direction of the surface 7 in the field of view of the camera 4, that is, in the width direction of the piston ring 6. In this way, in the process in which the illumination light 8 is irradiated onto the surface 7 via the filter 3 having the light / dark pattern, the bright light and dark light corresponding to the bright region of the filter 3 are changed due to the difference in the light transmittance of the filter 3. It is in a state divided into dark light corresponding to the line.

  The camera 4 is, for example, a CCD camera, and is arranged at a symmetrical position with respect to the light source 2 with the normal line of the surface 7 of the piston ring 6 as the axis of symmetry, and a light / dark pattern, that is, reflected light 9 corresponding to light on the surface 7. The reflected light 9 corresponding to the dark is incident, a photographed image as shown in FIG. 3 is photographed, and photographed image data is sent to the image processing unit 5. In the captured image of FIG. 3, the upper and lower portions correspond to the upper and lower positions in the photographing posture of the camera 4, and the central bright portion (gray portion) corresponds to the surface 7 of the piston ring 6. The central dark portion (black portion) in the horizontal direction corresponds to one dark line of the filter 3, and the left and right dark portions (black portions) in the left and right directions correspond to the background at the time of photographing.

The image processing unit 5 is constituted by a computer or the like that incorporates an image processing program according to the present invention. As shown in FIG. 3, a dark line is formed on the surface 7 of the piston ring 6 from the photographed image of the camera 4. Scanning in the transverse direction, that is, the circumferential direction of the surface 7 (in the direction of the arrow), as shown in FIG. A luminance difference between the luminance and the dark image luminance is obtained, and the value of the luminance difference is compared with a threshold value corresponding to a good standard surface to evaluate the quality of the polished state of the surface 7. In FIG. 4, the upper and lower parts correspond to the upper and lower parts of the captured image in FIG.

The image processing unit 5, the comparison, for example, expression of the threshold value corresponding to the value and the standard surface luminance difference - calculated value of the luminance difference from the calculation of (luminance difference good surface brightness difference of the surface 7), When the expression ( luminance difference value <threshold value) is satisfied, the surface 7 is evaluated as defective, and when the expression ( luminance difference value <threshold value) is not satisfied, the surface 7 is evaluated as good. The threshold value at this time is a standard value for a well-polished surface and fluctuates under the influence of the luminance of the light source 2, so the surface 7 of the good piston ring 6 is periodically inspected. Is set as a limit value for determining a defective product from the value of the luminance difference of the non-defective product surface 7.

When the surface 7 to be evaluated is a polished surface on the outer periphery of the piston ring 6, the polished surface may diffusely reflect the illumination light 8 by fine streaks and generate irregular reflected light 9. At this time, the photographed image does not form a clear boundary at the boundary position of the light / dark pattern, and includes a specific luminance value even in the bright range or the dark range. In such a situation, if necessary, the image processing unit 5 scans at a plurality of positions in the circumferential direction of the surface 7 (in the direction of the arrow in FIG. 3), and displays a luminance change graph ( A plurality of brightness sections are created, and an average brightness difference value is obtained from the average value of the plurality of bright image brightnesses and the average value of the plurality of dark image brightnesses from these brightness change graphs (brightness section). Such an average luminance difference is stable without being affected by specific luminance such as irregular boundaries and noise, and is effective in preventing erroneous evaluation.

  When the surface 7 to be evaluated is a polished surface on the outer periphery of the piston ring 6 and irregular reflected light 9 is generated by fine streaks on the polished surface, the generation of irregular reflected light 9 causes The streak can be evaluated as poor polishing. At the time of inspection of fine streaks, the image processing unit 5 recognizes irregular reflected light 9 from complicated changes in the luminance graph of the captured image. For example, the width direction of the surface 7 on the outer periphery of the piston ring 6 By dividing into several areas and determining each block, it is possible to detect partial polishing defects due to the presence of fine streaks on the polished surface.

The average luminance difference value can also be realized by forming a striped light / dark pattern on the filter 3 with a plurality of light / dark patterns, that is, two or more dark lines. In this case, the image processing unit 5 obtains an average luminance difference value from the average value of the plurality of bright image luminances and the average value of the plurality of dark image luminances, and then compares the two by the above formula. It will be. This reduces the adverse effects of fine streaks on the polished surface during the evaluation process. When the filter 3 has a plurality of bright and dark patterns, for example, a plurality of lines in the outer peripheral direction of the piston ring 6 are used. In this way, a line sensor camera that scans in the generatrix direction of the outer peripheral surface 7 of the piston ring 6 can be used as the camera 4, thereby enabling high-speed evaluation. However, when the line sensor camera is used , partial polishing defects cannot be evaluated on the entire outer periphery, and this is taken into consideration.

In the above example, the value of the luminance difference is compared with the standard threshold value. However, the evaluation of the surface 7 is performed by comparing the bright luminance value of the surface 7 with the lower luminance limit value of the light, or the dark luminance value. It is also possible to apply a weighting condition for comparison with the dark lower limit luminance value. By this condition weighting, the surface 7 has the formula (bright brightness value> bright lower limit brightness value), formula (dark brightness value <dark upper limit brightness value) or formula (bright brightness value> bright lower limit brightness value). Evaluation can be made from a combination of (dark brightness value <dark upper limit brightness value), and it can be determined that the inspection has been normally performed when these expressions are satisfied. When the expression ( luminance difference value <threshold value) is satisfied and the evaluation based on the upper and lower limit luminance values is not satisfied, it can be predicted that the light amount of the light source 2 is insufficient. Become.

  Although the present invention was developed with the outer peripheral surface 7 of the piston ring 6 as an evaluation object, the present invention is not limited to the piston ring 6 and can be used for evaluation of polished surfaces and painted surfaces of various products and inspection of quality.

DESCRIPTION OF SYMBOLS 1 Surface evaluation apparatus 2 Light source 3 Filter 4 Camera 5 Image processing unit 6 Piston ring 7 Evaluation object surface 8 Illumination light 9 Reflected light

Claims (8)

By illuminating the surface to be evaluated with illumination light from a light source via a light / dark pattern filter, reflected light corresponding to light and dark is generated on the surface, and the reflected light is photographed by the camera. The brightness difference between the bright image brightness and the dark image brightness is obtained, and the brightness difference value is compared with a threshold value corresponding to a good standard surface, and the quality of the surface is evaluated from the comparison result. Evaluation method of the characteristic surface. 2. The method for evaluating a surface according to claim 1, wherein the surface is evaluated as defective when an expression (value of luminance difference <threshold) is obtained from the comparison result. The average luminance difference is obtained from an average value of a plurality of bright image luminances and an average value of a plurality of dark image luminances, wherein the evaluation target surface is a polished surface on the outer periphery of the piston ring. Or the evaluation method of the surface of Claim 2.   At the time of comparison, it is possible to weight at least one of the comparison between the formula (bright luminance value> bright lower luminance value) and the formula (dark luminance value <dark upper luminance value). The surface evaluation method according to claim 1, wherein: Photographing a light source that irradiates the surface to be evaluated with illumination light, a light / dark pattern filter interposed between the light source and the surface, and reflected light corresponding to light and dark by the light irradiated to the surface via the filter The brightness difference between the brightness of the bright image and the dark image from the camera and the image captured by the camera is obtained, and the quality of the surface is determined from the comparison result between the brightness difference value and the threshold value corresponding to a good standard surface. An apparatus for evaluating a surface, comprising: an image processing unit to be evaluated. The image processing unit compares the brightness difference value with a threshold value corresponding to a good standard surface, and the surface is evaluated as defective when the expression ( luminance difference value <threshold value) is obtained from the comparison result. The surface evaluation apparatus according to claim 5, wherein: The evaluation target surface is a polished surface on the outer periphery of the piston ring, and an average luminance difference is obtained by an image processing unit from an average value of a plurality of bright image luminances and an average value of a plurality of dark image luminances. The surface evaluation apparatus according to claim 5 or 6.   At the time of comparison by the image processing unit, it is possible to weight at least one of the expression (light brightness value> light lower limit brightness value) and expression (dark brightness value <dark upper limit brightness value). The surface evaluation apparatus according to claim 5, wherein:

Images