JP2008185438A - Method and device for detecting defect on cylindrical surface - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To detect only a projecting defect, of the projecting defect and a recessed defect on a cylindrical surface of a measuring object. <P>SOLUTION: The columnar or cylindrical measuring object W is rotated, and parallel light 1a parallel to the center axis Wo of the measuring object W is radiated from the tangential direction on the cylindrical surface of the measuring object W. A part slightly shifted the shade side from the boundary line We between a part W<SB>a</SB>receiving light and a dark field section W<SB>b</SB>that does not receive light is imaged using a line sensor camera 2. The projecting defect Wi of the measuring object W is detected as a white defect image in the image of the dark field section W<SB>b</SB>imaged by the line sensor camera 2, but the recessed defect is not detected. The minimum height (h) of the detected projecting defect Wi can be set arbitrarily by changing the shift amount (x) of the imaged position (focal position) 2b of the line sensor camera 2 from the boundary line We. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a defect detection method and a defect detection apparatus for a cylindrical surface for detecting surface defects of a cylinder or a cylinder, such as various rubber roller shafts used in electrophotographic copying machines and printers.

  Generally, various rubber roller shafts used in OA equipment such as electrophotographic copying machines and printers are used by being directly inserted into holes of a copying machine main body or a housing of a toner cartridge. Alternatively, parts such as a bearing, a position regulating roller, and a gear are attached and used. Therefore, if there is a convex defect on the surface of the shaft body, there arises a problem that the shaft body cannot be inserted into a hole in the housing or the like, or a component such as a bearing cannot be attached. Also, when the rubber roller rotates, if there is a convex defect on the surface of the shaft serving as the sliding surface, it becomes a resistance to rotation and uneven rotation occurs, or wear of other parts is accelerated. Concave defects such as voids and pinholes caused by the shaft body base material do not become a problem, but convex defects can be affected even by scratches of several micrometers, so the surface of the shaft body before assembling the rubber roller. It is necessary to inspect the convex defects.

  In general, inspection of defects on the surfaces of these cylinders or cylinders has been carried out by visual inspection or palpation, but in recent years it has become possible to automate by improving the resolution of CCD cameras and increasing the speed of computers that perform image processing (Patent Documents). 1).

  However, since the conventional defect detection method detects surface irregularity defects, it is not possible to detect only convex defects that are problematic on the surface of the rubber roller shaft. For this reason, it is necessary to discard all those having irregularities on the surface as defective, or to reexamine those having irregularities by visual inspection or palpation, and to isolate only those having convex defects.

  As a method of detecting only convex scratches, there is a method of rotating a cylinder or a cylindrical body, which is an object to be measured, and reading a change in the position of the ridge line with a laser length measuring device or an area sensor camera. However, with this method, it is necessary to pick up changes in the position of the ridge line due to the shake of the object to be measured, and the measurement range in the axial direction is extremely narrow, so it is necessary to perform measurements while scanning the axial direction many times, so inspection takes time. It will take.

Therefore, there is a method of detecting defects on the cylindrical surface by irradiating a columnar or cylindrical body as an object to be measured with a regular pattern of light and evaluating the regularity of the pattern (see Patent Document 2). However, in order to detect a defect of about several μm, it is necessary to irradiate pattern light with very high accuracy. Furthermore, it is necessary to correct the cylindrical shape of the object to be measured and the shake when it is rotated, and to compare an image without a defect with an image with a defect, which complicates image processing.
Japanese Patent No. 3469714 JP 11-185040 A

  The present invention has been made in view of the above-mentioned unsolved problems of the prior art, and a cylindrical surface capable of detecting only a convex defect on the surface of a columnar or cylindrical body with high accuracy and efficiency. It is an object of the present invention to provide a defect detection method and a defect detection apparatus.

  The cylindrical surface defect detection method of the present invention irradiates parallel light parallel to the central axis in the tangential direction of the cylindrical surface while rotating the object having the cylindrical surface around the central axis. And imaging the non-irradiated portion of the cylindrical surface that is not exposed to the parallel light with a line sensor camera arranged parallel to the central axis, and the cylindrical surface based on the captured image. And a step of detecting a convex defect.

  By projecting the non-irradiated part in the vicinity of the boundary line between the irradiated part to which the parallel light hits and the non-irradiated part to which the parallel light does not hit, the convex shape that the parallel light hits in the non-irradiated part. Only detect defects. The height of the convex defect that can be detected can be arbitrarily set according to the shift amount of the imaging position of the line sensor camera with respect to the boundary line between the irradiation part and the non-irradiation part.

  The best mode for carrying out the present invention will be described with reference to the drawings.

As shown in FIG. 1, the parallel light 1a emitted from the light source 1 is irradiated in the tangential direction of the surface (cylindrical surface) of the object W to be measured. Then, the measurement object W is slightly shifted to the negative side (non-irradiation side) from the boundary line We between the portion (irradiated portion) W _a where the light hits and the dark field portion (non-irradiated portion) W _b where the light does not hit. A part is imaged using the line sensor camera 2 arranged in parallel to the central axis Wo of the workpiece W.

Examples of the light source 1 include a method of attaching a line type light guide and a line condenser lens to a commercially available halogen light source device or meral halide light source device. Moreover, you may use the line illuminating device using LED. In particular, it is preferable to use short-wavelength illumination with little diffraction so that the boundary line We between the portion W _a where the light to be measured W is irradiated and the dark field portion W _b where the light is not irradiated is clear. Examples of short wavelength illumination include a method of attaching a blue filter to a light source or using a blue LED or an ultraviolet light source.

Imaging position 2b is the focal position of the line sensor camera 2 lens 2a, as described above, slightly boundary line We of the dark field W _b which is part W _a and light the light of the workpiece W hits not hit The position is shifted to the negative side.

As shown in (a) of FIG. 2, when there is a convex defect Wi protruding from the tangent connecting the light source 1 and the workpiece W in a dark field portion W _b, because the light hits the convex defect Wi, In an image processing apparatus (not shown), the defect image 3 can be detected as shown in FIG.

  On the other hand, as shown in FIG. 3, the concave defect Wf does not protrude from the tangent line connecting the light source 1 and the object W to be measured, so that the parallel light 1 a does not strike and is not captured by the line sensor camera 2. Therefore, only the convex defect Wi can be detected.

  As shown in FIG. 1B, a convex defect is detected by detecting the amount of deviation between the positive and negative boundary line We and the imaging position 2b as x (mm), the radius of the object to be measured as R (mm), and so on. Assuming that the minimum height of Wi is h (mm), the shift amount x is obtained as follows by the similarity of triangles and the three-square theorem.

h: a = (h + R): (a + x) (1)
R ₂ + (a + x) ² = (h + R) ² (2)
Obtaining a from equation (1), substituting it into equation (2) and deleting a,
From equation (1)

（２）式に代入し、

Substituting into equation (2)

（３）式から以下の計算によって（４）式を求める。

Equation (4) is obtained from equation (3) by the following calculation.

  If the shift amount x is set to be equal to or less than the value obtained by the equation (4), light hits the apex of the convex defect Wi and can be detected by image processing. That is, the minimum height h of the convex defect Wi to be detected can be arbitrarily set by adjusting the shift amount x.

  An image obtained from the line sensor camera 2 is taken into the image processing apparatus as a developed image of at least one round and is subjected to image processing to detect the convex defect Wi. Examples of the image processing method include a method of performing binarization processing and detecting a bright portion of a certain level or more as the convex defect Wi.

  As a mechanism for gripping and rotating a columnar or cylindrical body that is the object to be measured W, for example, the center axis is gripped at both ends of the object to be measured with a pointed center axis or a reverse center axis recessed in a mortar shape. A method of rotating the workpiece W by rotating the. In addition, two places on the outer peripheral surface of the workpiece W are received by a guide composed of two rollers such as a bearing, and the workpiece W is rotated by rotating at least one roller, or a rotating disk or belt is measured. The method of rotating the to-be-measured object W by pressing on the outer peripheral surface of the thing W is mentioned.

The rotation speed (rpm) of the object W to be measured is R (mm) for the radius of the object W to be measured, L (mm) for the resolution of the line sensor camera 2 including the optical system such as the lens 2a, and the like. When the imaging interval is set to f (Hz), it is preferable to set the rotational speed r ₀ or less according to the following equation.

  The resolution L of the line sensor camera 2 is determined by (size per pixel of the image sensor of the line sensor camera) / (lens magnification).

  FIG. 4 shows a defect detection apparatus for detecting a convex defect of a metal shaft of an electrophotographic charging roller. The metal shaft body to be measured W is a shaft body of material SUM24L free-cutting steel, outer diameter φ6 mm (radius R = 3 mm), length 250 mm, and the surface of this shaft body is subjected to electroless nickel plating treatment. It has been subjected. As the light source 1, a commercially available 150 W halogen lamp light source device (trade name: MHF-G150LRC, manufactured by Moritex) equipped with a blue filter, a line-type light guide, and a line condenser lens is used. And it arrange | positioned in parallel with respect to the central axis of the to-be-measured object W so that it may irradiate from the tangential direction of the to-be-measured object surface, and was offset 3 mm with respect to the normal line of the to-be-measured object W.

  As a line sensor camera 2 for capturing images, a commercially available camera (TL-5150UFD brand name, manufactured by Takenaka System Equipment) is used, and a lens for a 35 mm single-lens reflex camera (Nikon brand name) with a focal length of 28 mm is used as a lens 2a. Reversed through. At this time, the photographing magnification was about 1.5 times, and the resolution of the line sensor camera 2 including the optical system was about 4.7 μm.

  The line sensor camera 2 and the light source 1 are arranged in a 90 ° direction, and the amount of deviation x of the imaging position 2b of the line sensor camera 2 with respect to the boundary line between the negative side and the positive side of the measured object W is a convex defect. X = 0.24 mm corresponding to the minimum height h = 10 μm was set.

  The mechanism for rotating the workpiece W includes a pair of driving rollers 11 and a pair of driven rollers 12 that are rotated by a motor 16 via a pulley 14, a belt 15, and a gear 13. The rollers 11 and 12 received positions of 20 mm from both ends of the object to be measured W, and the driving roller 11 was rotated to drive the object to be measured W. At this time, the rotation speed of the workpiece W was set to 70 rpm, and defect detection was performed.

  For the same object to be measured as in Example 1, the deviation amount x between the boundary line on the negative side and the positive side and the imaging position of the line sensor camera is set to x = 0.0 so that the minimum height h of the convex defect is 5 μm. Defect detection was performed under the same conditions as in Example 1 except that the thickness was set to 17 mm.

(Comparative example)
As a comparative example, as shown in FIG. 5, the line sensor camera 2 is arranged in the normal direction of the object W to be measured using the same light source 1, line sensor camera 2, and the like as those of the first and second embodiments. Then, the center of the light source was aligned with the imaging position 2 b of the line sensor camera 2 and further arranged at a position of 45 ° with respect to the line sensor camera 2. The mechanism for rotating the workpiece W and the number of rotations were set to be the same as those in the first embodiment.

  In both Example 1, Example 2, and Comparative Example, developed images slightly overlapped by one round of the object to be measured were synthesized from the images obtained from the line sensor camera, and these images were observed. At this time, the size of the obtained image is 20 million pixels of 5000 pixels in the axial direction of the object to be measured and 4000 pixels in the circumferential development direction. The obtained image was binarized using a commercially available personal computer.

  In Example 1, since the normal part enters the shadow of the object to be measured and does not receive light, it becomes a black image, and the convex defect of 10 μm or more protruding from the tangent line connecting the light source and the object to be measured is irradiated with light. As shown in (b), a white defect image was obtained. On the other hand, convex defects and concave defects having a height of less than 10 μm were not detected because they were behind the object to be measured. That is, only convex defects of 10 μm or more could be detected.

  In Example 2, it was possible to detect even a convex defect of 5 to 10 μm that could not be detected in Example 1, but the normal part and the concave defect entered the shadow of the object to be measured as in Example 1, and light was emitted. Since it did not hit, it was detected as a black image.

  In the comparative example, the normal part is exposed to light and becomes a white image. Then, as shown in FIG. 6A, when there is a convex defect Wi, the opposite side to the light source 1 is shaded, so that a black defect image 4 is detected as shown in FIG. 6B. On the other hand, as shown in FIG. 7A, the concave defect Wf is also detected as a black defect image 5 as shown in FIG. For this reason, in the comparative example, it was not possible to determine whether the defect was a convex defect or a concave defect from the obtained image.

  Table 1 summarizes the detection results of defects according to Example 1, Example 2, and Comparative Example.

  In Example 1 and Example 2, it is possible to detect only convex defects on the cylindrical surface of a cylindrical or cylindrical object to be measured. Furthermore, it is possible to arbitrarily set the minimum height of the convex defect to be detected by adjusting the imaging position of the line sensor camera.

BRIEF DESCRIPTION OF THE DRAWINGS It shows schematic structure of the defect detection apparatus by one Embodiment, (a) is a model elevation which shows the principal part, (b) is the elements on larger scale which expand and show a part of (a). is there. It is a figure explaining the defect detection method of the cylindrical surface by the apparatus of FIG. It is a figure explaining not detecting a concave defect. BRIEF DESCRIPTION OF THE DRAWINGS The defect detection apparatus by Example 1 is shown, (a) is the elevation view, (b) is a side view. It is a figure explaining a comparative example. It is a figure explaining the case where a convex defect is detected in a comparative example. It is a figure explaining the case where a concave defect is detected in a comparative example.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Light source 1a Parallel light 2 Line sensor camera 2a Lens 2b Image pick-up position 11 Roller for driving 12 Roller for follower 16 Motor

Claims (4)

While rotating an object to be measured having a cylindrical surface around a central axis, parallel light parallel to the central axis is irradiated in a tangential direction of the cylindrical surface, and the parallel light on the cylindrical surface is irradiated. Imaging a non-irradiated portion not hit by a line sensor camera arranged parallel to the central axis;
And detecting a convex defect on the cylindrical surface based on a captured image. A method for detecting a defect on a cylindrical surface, comprising: When the radius of the cylindrical surface is R (mm) and the minimum height of the convex defect to be detected is h (mm), the space between the non-irradiated part of the cylindrical surface and the irradiated part to which the parallel light hits. The cylindrical surface defect detection method according to claim 1, wherein a deviation amount x (mm) of the imaging position of the line sensor camera with respect to the boundary line satisfies a relationship represented by the following expression.

  A mechanism for gripping an object to be measured having a cylindrical surface and rotating it around a central axis; a light source that irradiates parallel light parallel to the central axis in a tangential direction of the cylindrical surface; A non-irradiated portion that is not exposed to the parallel light on the cylindrical surface, and has a line sensor camera arranged in parallel to the image sensor and an image processing device that processes an image captured from the line sensor camera. A defect detection apparatus, wherein an image is picked up by the line sensor camera. When the radius of the cylindrical surface is R (mm) and the minimum height of the convex defect to be detected is h (mm), the space between the non-irradiated part of the cylindrical surface and the irradiated part to which the parallel light hits. The defect detection apparatus according to claim 3, wherein a deviation amount x (mm) of the imaging position of the line sensor camera with respect to the boundary line satisfies a relationship represented by the following expression.

Images