JP2012063190A - Surface unevenness detection apparatus and surface unevenness detection method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a surface unevenness inspection apparatus for efficiently detecting surface unevenness of an object to be inspected.SOLUTION: The surface unevenness inspection apparatus includes a line sensor 2, an illumination device 1 for generating contrast patterns with a plurality of phases in parallel with the line sensor 1 and an arithmetic unit 5 for combining linear imaging data imaged by the line sensor 2 with a frame image and is configured so that contrast patterns with a plurality of phases having different irradiation angles are applied to the same imaging target line. The line sensor 2 acquires light from the contrast patterns with the plurality of phases having different irradiation angles reflected by the same imaging target line as a plurality of linear imaging data synchronously with the switching of the contrast patterns. The arithmetic unit 5 generates a plurality of screen images of each phase by combining the imaging data of the same phase out of the plurality of phases when combining the linear imaging data acquired by the line sensor 2 with a frame image, and applies the same image processing to a plurality of generated screen images.

Description

  The present invention relates to a surface unevenness inspection technique for inspecting unevenness on the surface of an object to be inspected, and in particular, a surface capable of inspecting the presence or absence of unevenness on a painted surface or metal surface as an object to be inspected with a simple structure and processing. The present invention relates to an unevenness detecting device.

  As conventional surface inspection techniques for inspecting the surface of an object to be inspected, for example, Japanese Patent Application Laid-Open No. 2002-257528 (Patent Document 1), Japanese Patent Application Laid-Open No. 2001-349716 (Patent Document 2), and Japanese Patent No. 4190636 ( There are those disclosed in Patent Document 3) and the like.

(A) Japanese Patent Application Laid-Open No. 2002-257528 (Patent Document 1) synthesizes a line image sequentially captured by a line sensor by applying light having a sinusoidal distribution to the surface of an object to be inspected, and a phase shift method. A three-dimensional shape measuring apparatus that reconstructs a three-dimensional image has been proposed. For example, a plurality of light sources are aligned so that the phase of the grating pattern on the surface of the inspection object is π / 2, and each light source is aligned for each line synchronization signal generated each time one line is imaged. A phase shift pattern is generated by sequentially turning on and off, and surface irregularities are detected by performing phase shift calculation as image processing.

(B) Japanese Patent Application Laid-Open No. 2001-349716 (Patent Document 2) uses grid-like illumination to illuminate the surface of a specimen with a clear boundary between a bright part and a dark part, and the boundary on the specimen surface. When there is an irregularity defect, a bright / dark reversal part is generated when the boundary passes, so the presence / absence of a concave / convex defect is determined by detecting this bright / dark reversal part. A surface roughness inspection apparatus has been proposed.

(C) Japanese Patent No. 4190636 (Patent Document 3) discloses a bright-field light source arranged so that the total reflected light of the light projected on the object to be inspected reaches the line sensor camera, and is projected onto the object to be inspected. Illuminated from different directions with two or more light sources for dark field arranged so that the total reflected light of the reflected light hardly reaches the line sensor camera, and provided with one line sensor camera as an imaging means, Each time the line sensor camera scans one line while the object to be inspected is moving, a single line sensor camera switches a plurality of different light sources by alternately switching between the light source for dark field and the light source for bright field. There has been proposed a surface inspection apparatus that obtains an image photographed by irradiation and uses the image for different purposes. For example, detection of dirt attached to the surface with a light source for bright field, detection of scratches with relatively little unevenness, detection of scratches with large unevenness with a light source for dark field, and detection of two purposes in one scan Like to do.

  In the above-described conventional technique, a non-inspected object is imaged by combining illumination having a grid-like slit and a line sensor, and surface inspection is performed by subsequent image processing.

However, each of the above conventional techniques has the following problems.
(A) In Japanese Patent Laid-Open No. 2002-257528 (Patent Document 1), the light of the slit illumination needs to have a distribution on a sine wave, and the calculation of the phase shift method is necessary. There is a problem that both processes are complicated.

(B) In Japanese Patent Laid-Open No. 2001-349716 (Patent Document 2), it is necessary to continuously track a light / dark pattern and detect a discontinuity “displacement”, and this image processing is also relatively complicated. There is a problem that it becomes a thing.

(C) In Japanese Patent No. 4190636 (Patent Document 3), a bright-field light source and a dark-field light source are arranged in different directions, and images obtained by these different light sources are used for different detection purposes. Therefore, there is a problem that image processing must be performed for each of images obtained by different light sources, and image processing and control are relatively increased.

  In view of these problems, the present invention efficiently performs the same image processing on a plurality of synthesized images obtained from a lighting device and a line sensor with a simple configuration that can be easily realized. An object of the present invention is to provide a surface unevenness inspection apparatus for detecting unevenness.

  The surface unevenness detecting apparatus according to the present invention has the following configuration in order to achieve the above object.

a) An imaging line sensor, an illumination device that generates a light / dark pattern (hereinafter, a plurality of light / dark patterns having a plurality of phases) in parallel with the imaging line sensor, and the imaging line sensor. Comprising an arithmetic unit that synthesizes linear imaging data into a frame image, and is configured to illuminate the same imaging target line with a plurality of phases of bright and dark patterns with different irradiation angles,
The imaging line sensor acquires light from a plurality of phases of bright and dark patterns with different irradiation angles reflected by the same imaging target line as a plurality of linear imaging data in synchronization with the switching of the bright and dark patterns. Then, when the linear imaging data acquired by the imaging line sensor is synthesized with a frame image, the arithmetic device synthesizes imaging data of the same phase among the plurality of phases for each phase. A plurality of surface images are generated, and the same image processing is performed on the generated plurality of surface images.

b) In addition, the illumination device is irradiated with a plurality of line-shaped LEDs, a filter having a plurality of striped patterns, and a plurality of phases of light having different phases from the plurality of line-shaped LEDs that have passed through the filter. A light and dark pattern parallel to the line sensor is generated on the screen.

  The same image processing is characterized by feature enhancement filter processing for detecting a luminance change on the image and binarization processing.

  Further, the same image processing is performed on the plurality of combined surface images, the image processing extracts a luminance change on the image, and a portion where the luminance change exists in the plurality of surface images is regarded as an uneven defect. It is characterized by being determined.

  The surface unevenness detection method according to the present invention includes an imaging line sensor, an illumination device that generates a multi-phase light and dark pattern in parallel with the imaging line sensor, and a linear imaging imaged by the imaging line sensor. A surface unevenness detection method using an arithmetic unit that synthesizes data with a frame image, illuminating a plurality of phases of bright and dark patterns with different irradiation angles on the same imaging target line, and the imaging line sensor The light from the plurality of bright and dark patterns having different irradiation angles reflected by the imaging target line is acquired as a plurality of linear imaging data in synchronization with the switching of the bright and dark patterns, and the imaging device performs the imaging. When combining the linear imaging data acquired by the line sensor with the frame image, the imaging data of the same phase among the plurality of phases is synthesized. Is characterized by performing the same image processing on a plurality of faces images together, it generated to generate a plurality of surfaces image for each phase by.

  According to the present invention, by adopting the configuration as described above, it is possible to realize a surface unevenness detecting device and a surface unevenness detecting method capable of detecting unevenness on the surface of an inspection object with a simple structure and processing.

It is a figure which shows one Example of the surface unevenness | corrugation detection apparatus which concerns on this invention. It is the figure which expanded and showed the part of the optical system of FIG. 1-A. The figure which shows the timing signal of the trigger signal sent to a camera from camera control, and the LED lighting signal at the time of selecting 3 phase (a) or 5 phase (b) sent to each linear LED lamp of an illuminating device from illumination control It is. It is a figure which shows the relationship between the uneven | corrugated | grooved part of the to-be-inspected surface, and the range of illumination. FIG. 1A is a diagram illustrating an embodiment in which a timing signal generated by a transmitter circuit or some other method is used in place of an encoder signal when an encoder cannot be attached due to the system configuration. is there. It is a figure which shows the mode of the image in the frame memory 6 when phase switch SW8a selects 3 phases. It is a figure which shows the light intensity in the irradiation surface (screen surface) in the case of irradiating the line-like LED lamp 1a for phase 1-the line-like LED lamp 1e for phase 5 in 5 lines in a line form. It is a figure which shows the light intensity | strength in a to-be-inspected object in the case of arranging the line-shaped LED lamp 1a for phase 1-the line-shaped LED lamp 1e for phase 5 in 5 lines in a line form, and irradiating. It is a figure which shows an example of the characteristic of the optical filter of a fringe pattern. It is a figure which shows the fine unevenness detection principle in this invention. It is an example of a state in which a striped bright / dark pattern is irradiated on the surface of an object to be inspected (object to be measured) ((a) when there is no unevenness on the surface, (b) when there is unevenness on the surface) . It is a figure for demonstrating the method of detecting the presence or absence of the unevenness | corrugation of a surface by imaging the striped light-dark pattern irradiated to the surface of the to-be-inspected object (measuring object) with a line sensor.

1-A and FIG. 1-B are diagrams showing an embodiment of a surface unevenness detecting apparatus for an inspection object according to the present invention.
In FIG. 1-A, 1 is an illumination device having a line-shaped LED lamp having a plurality of lines, 2 is a camera (line sensor), 3 is an object to be inspected (object to be measured), 4 is a transport device, 5 is a computer arithmetic device, 5a is an image processing unit, 6 is a frame memory, 7 is a camera control, 8 is an illumination control, SW8a is a changeover switch for selecting the phase of the line-shaped LED lamp constituting the illumination device 1, and 9 is a rotary switch. An encoder 10 indicates a timing controller.

  In the surface unevenness detecting device shown in FIG. 1A, when the rotating roller of the transport device 4 rotates in the direction of the arrow, the inspection object 3 on the belt moves toward the right side of the drawing. The rotary encoder 9 attached to the rotating roller rotates with the rotation of the roller and generates a pulse signal. The pulse signal is guided to the computer arithmetic unit 5 through the timing controller 10. FIG. 1B is an enlarged view of a portion of the optical system in FIG.

  The pulse signal guided to the computer arithmetic unit 5 is supplied to the camera 2 as a trigger signal (see trigger signal in FIG. 2) via the camera control 7.

  On the other hand, an LED lighting signal (see FIG. 2) is generated from the same trigger signal via the lighting control 8, and only one predetermined line among a plurality of line-shaped LED lamps of the lighting device is turned on.

  FIG. 2 shows a timing chart of the trigger signal sent from the camera control 7 to the camera 2 and the LED lighting signal when the three-phase or five-phase sent from the lighting control 8 to each line-shaped LED lamp of the lighting device is selected. FIG.

  FIG. 2A shows a case where the phase switch SW8a for lighting control selects the three phases, and the LED lamps 1a to 1c are repeatedly lit in order.

  FIG. 2B shows a case where the phase changeover switch SW8a selects 5 phases, and the line-shaped LED lamps 1a to 1e are repeatedly lit in order. Of course, when the phase changeover switch SW8a selects 4 phases, the line-shaped LED lamps 1a to 1d are repeatedly lit in order.

FIG. 3 is a diagram illustrating the uneven portion on the surface of the object to be inspected and the range of illumination.
In the same figure, 3 is an object to be inspected, 3a, 3b and 3c are uneven portions (defects) on the surface of the object to be inspected, and 3d is an illumination range.

  If the encoder 9 cannot be attached due to the system configuration, it is possible to use a timing signal externally generated by a transmission circuit or some other method instead of the signal of the encoder 9, as shown in FIG. is there.

  In FIG. 4, the illumination device 1, the camera (line sensor) 2, the inspection object (measurement object) 3, the computer arithmetic device 5, the image processing unit 5a, the frame memory 6, the camera control 7, the illumination control 8, and the changeover switch SW8a. May have the same configuration as that shown in FIG.

<Relationship between imported original image and phase image>
The camera 2 captures one line of the image signal into the frame memory 6 in the computer arithmetic device 5 for each trigger signal.

FIG. 5 shows a state of an image in the frame memory 6 when the phase changeover switch SW8a selects the three phases.
In the same figure, the original image 11 shown in (a) shows a state in which phase 1, phase 2, phase 3, phase 1, phase 2, phase 3,. , 11a to 11c shown in (b) to (d) of FIG.

When the value is defined as Img (x, y) as the position (x, y) of the original image and the position (i, j) of the image of each phase,
(A) When the phase changeover switch SW8a selects “three phases”
The position (x, y) of the original image,
The relationship between the position (i, j) of the phase 1 image (indicated by 11j0 in the figure) and its value a is
x = i, y = 3 * j + 0, a = Img (i, 3 * j + 0)
The relationship between the position (i, j) of the phase 2 image (indicated by 11j1 in the figure) and its value b is
x = i, y = 3 * j + 1, b = Img (i, 3 * j + 1)
The relationship between the position (i, j) of the phase 3 image (indicated by 11j2 in the figure) and its value c is
x = i, y = 3 * j + 2, c = Img (i, 3 * j + 2)
Indicated by

(B) When the phase switch SW8a selects “four phases”
The position (x, y) of the original image,
The relationship between the position (i, j) of the phase 1 image and its value a is
x = i, y = 4 * j + 0, a = Img (i, 4 * j + 0)
The relationship between the position (i, j) of the phase 2 image and its value b is
x = i, y = 4 * j + 1, b = Img (i, 4 * j + 1)
The relationship between the position (i, j) of the phase 3 image and its value c is
x = i, y = 4 * j + 2, c = Img (i, 4 * j + 2)
The relationship between the position (i, j) of the phase 4 image and its value d is
x = i, y = 4 * j + 3, d = Img (i, 4 * j + 3)
Indicated by

(C) When the phase switch SW8a selects “5 phase”
The position (x, y) of the original image,
The relationship between the position (i, j) of the phase 1 image and its value a is
x = i, y = 5 * j + 0, a = Img (i, 5 * j + 0)
The relationship between the position (i, j) of the phase 2 image and its value b is
x = i, y = 5 * j + 1, b = Img (i, 5 * j + 1)
The relationship between the position (i, j) of the phase 3 image and its value c is
x = i, y = 5 * j + 2, c = Img (i, 5 * j + 2)
The relationship between the position (i, j) of the phase 4 image and its value d is
x = i, y = 5 * j + 3, d = Img (i, 5 * j + 3)
The relationship between the position (i, j) of the phase 5 image and its value e is
x = i, y = 5 * j + 4, e = Img (i, 5 * j + 4)
Indicated by

《Explanation of lighting device》
FIG. 6 shows a line LED lamp 1a for phase 1, a line LED lamp 1b for phase 2, a line LED lamp 1c for phase 3, a line LED lamp 1d for phase 4, and a line LED lamp 1e for phase 5. The light intensity in the irradiation surface (screen surface) in the case of irradiating 5 lines in a line is shown.

  The screen 1g is irradiated with light from a linear LED lamp that has passed through the optical filter 1f, and the intensity (brightness) is uniform from the back to the front (x direction) and from right to left (y direction). As shown in FIG. 6, both sides are attached so as to form a trapezoidal wave that is weak on both sides and bright in the center.

  FIG. 6 shows the brightness in a state where the phase 3 line-shaped LED lamp 1c is lit. When the LED lamp 1a or 1e is lit, a wave of brightness moves like a one-dot chain line and a dotted line. .

  The optical filter 1f in the illuminating device 1 may be in close contact with the screen 1g or may be slightly separated.

  FIG. 7 shows a line LED lamp 1a for phase 1, a line LED lamp 1b for phase 2, a line LED lamp 1c for phase 3, a line LED lamp 1d for phase 4, and a line LED lamp 1e for phase 5. The light intensity on the surface of the object to be inspected in the case of irradiating with five lines arranged in a line is shown.

  By the light intensity pattern shown in FIG. 6 formed on the screen irradiated with the light of the LED lamp that has passed through the optical filter 1f, the light of the brightness pattern having the waveform shown in FIG. .

  FIG. 7 shows the brightness on the inspection object in a state where the phase 3 line-shaped LED lamp 1c is lit, and when the LED lamp 1a or 1e is lit, as indicated by a one-dot chain line or a dotted line. The brightness waveform moves to the illumination range.

  FIG. 8 is a diagram illustrating an example of the characteristics of a striped pattern optical filter. FIG. 8A illustrates an optical filter having a striped pattern, and FIG. 8B illustrates the transmittance thereof. As shown in the figure, in the optical filter having a fringe pattern, the transmittance of the portion indicated by hatching is low and the transmittance of the portion other than the hatching is high.

《Explanation of detection principle of micro unevenness》
FIG. 9 is a diagram for explaining the principle of detecting fine unevenness according to the present invention.
It is assumed that the camera and the screen are mounted symmetrically so as to have an angle θ across the normal, and the camera is focused on the point of interest P0.

If the point of interest P0 is an object to be specularly reflected by a plane, the camera should detect a value proportional to the brightness of the point S0 on the screen.
Assuming that the point of interest P0 is tilted by an angle α with respect to the plane, the normal line is also tilted by the angle α, so that the point S0 on the screen moves to a position Sα that is shifted by an angle 2α.

When the distance between the point of interest P0 and the point S0 on the screen is L, these relational expressions are
((Sα−S0) / L) = tan (2α)
And expand,
Sα = L * tan (2α) + S0
It can be expressed as

As an example,
When L = 100 mm, α = 1 °, S0 = 0 mm, Sα = 3.49 mm
When L = 100 mm, α = −1 °, S0 = 0 mm, Sα = −3.49 mm
When L = 100 mm, α = 2 °, S0 = 0 mm, Sα = 6.99 mm
When L = 100 mm, α = −2 °, S0 = 0 mm, Sα = −6.99 mm
It becomes.

  It is obvious that if the entire screen shines uniformly with the same brightness, the object to be inspected has fine irregularities, and even if Sα changes, the brightness recognized by the camera does not change.

  However, if there is a bright and dark stripe pattern on the screen and the brightness changes depending on the position, the brightness on the screen will differ depending on the position of the Sα point.

  If there is a fine unevenness at the position of the point of interest P0 on the surface of the object to be inspected and tilted by an angle α with respect to the plane, the camera 2 receives the brightness (brightness) of the point Sα on the screen. Based on the change in the intensity of the light received by the camera 2, it can be recognized that the surface of the object to be inspected is inclined (inclined), that is, has unevenness. Since the fine unevenness usually changes at a continuous angle, it is easy to detect by this method. When there is a portion that is inclined by an angle α with respect to the plane in the fine uneven portion, the brightness (light intensity) on the screen depending on the position of the Sα point is recognized by the camera 2 and the frame is formed as line-shaped image data. Captured in the memory 6.

<< Image composition and detection filter >>
The line-shaped image data in the captured frame memory is the above-described image data having the same phase (that is, phase 1 image data, phase 2 image data, phase 3 image data). The images are arranged so that one image is formed for each phase. For example, if the illumination is three-phase, three images are obtained, and if the illumination is five-phase, five images are obtained.

  Each phase image is the same reflection imaged at each phase. If there are discontinuous irregularities on the surface of the non-inspection object, as shown above, the reflection angle changes and the brightness changes accordingly. Will be.

FIG. 10 is an example of a state in which a striped light / dark pattern is irradiated on the surface of an object to be inspected (object to be measured).
FIG. 10A shows a case where there is no unevenness on the surface, and even if illumination / imaging is performed from the phase, the light / dark interface of the striped pattern is a straight line, and there are no discontinuities (lines) such as distortion.
FIG. 10B shows a case where the surface has unevenness, and as described in the explanation of the unevenness detection principle, a discontinuous point / line such as distortion corresponding to the change in the reflection angle is generated, and the unevenness is formed in that portion. Indicates that there is a defect.

  FIG. 11 is a diagram for explaining a method of detecting the presence / absence of unevenness on the surface by imaging a striped bright / dark pattern irradiated on the surface of the inspection object (object to be measured) with a line sensor.

  This figure shows a line image when the line-shaped light and dark pattern when the surface shown in FIG. 10B is uneven is scanned linearly in parallel with the light and dark pattern of the stripe.

  As shown in the figure, a striped light-dark pattern with discontinuities such as distortion when the surface is uneven is linearly parallel to the striped light-dark pattern with a line sensor, and the discontinuous points are In a line image that is an imaging result obtained by scanning across, a discontinuity point (line) of light or darkness or luminance is generated, whereby it is possible to detect the presence of unevenness on the surface of the inspection object.

  If the line images obtained in this way are illuminated and imaged with the same phase and collected and re-synthesized, the uneven defect position is reflected as a region with different brightness.

  In addition, in the case of illumination with a single phase, there has been a problem that it is impossible to detect a luminance change due to reflection that is an integral multiple of the interval between the striped repetitive patterns. In the present invention, however, a plurality of illuminations having different phases are used. This can be solved.

  For the detection of the irregularities, for example, a filter for enhancing the luminance change and the subsequent binarization process may be performed on the plurality of images to extract a specific portion of the luminance in each image. If binarization processing is performed, unevenness, that is, surface defects can be obtained as information on the presence or absence, and finally, this information is used to inspect non-defective products with uneven defects and defective products with uneven defects (OK / NG). Can be the result.

1: Illuminating device 1a to 1e: Line-shaped LED lamp 1f: Optical filter 1g: Screen 2: Camera (line sensor)
3: Inspected object (object to be measured)
3a to 3c: Concavities and convexities on the surface of the inspected object (concave and convex defects)
3d: Illumination range 4: Transport device 5: Computer arithmetic device 6: Frame memory 7: Camera control 8: Illumination control 9: Rotary encoder 10: Timing controller 11: Original images 11a to 11c: Images divided for each phase

JP 2002-257528 A JP 2001-349716 A Japanese Patent No. 4190636

Claims (5)

An imaging line sensor, an illuminating device that generates a light / dark pattern (hereinafter, a plurality of light / dark patterns having a plurality of phases) in parallel with the imaging line sensor, and a line imaged by the imaging line sensor Comprising an arithmetic unit that synthesizes image data in the form of a frame image,
The same imaging target line is configured to illuminate multiple phases of light and dark patterns with different irradiation angles,
The imaging line sensor acquires light from a plurality of phases of bright and dark patterns with different irradiation angles reflected by the same imaging target line as a plurality of linear imaging data in synchronization with the switching of the bright and dark patterns. And
When the linear imaging data acquired by the imaging line sensor is synthesized with a frame image, the arithmetic device synthesizes imaging data of the same phase among the plurality of phases, thereby combining a plurality of imaging data for each phase. A surface unevenness detecting device that generates a surface image and performs the same image processing on the plurality of generated surface images. The lighting device includes: a plurality of line-shaped LEDs; a filter having a plurality of striped patterns; and a screen on which light having a plurality of phases from the plurality of line-shaped LEDs that have passed through the filter is irradiated. Have
The surface unevenness detection apparatus according to claim 1, wherein a plurality of phases of light and dark patterns parallel to the line sensor are generated on the screen.   3. The surface unevenness detecting device according to claim 1, wherein the same image processing is feature enhancement filter processing for detecting a luminance change on the image and binarization processing.   The same image processing is performed on the plurality of combined surface images, the image processing extracts a luminance change on the image, and determines a portion where the luminance change exists in the plurality of surface images as an uneven defect. 4. The surface unevenness detecting device according to claim 1, wherein the surface unevenness detecting device is a device. An imaging line sensor, an illuminating device that generates a multi-phase light and dark pattern in parallel with the imaging line sensor, and an arithmetic unit that synthesizes linear imaging data captured by the imaging line sensor into a frame image The surface unevenness detection method used,
Illuminate light and dark patterns of multiple phases with different irradiation angles for the same imaging target line,
Light from a plurality of phases of bright and dark patterns with different irradiation angles reflected by the same imaging target line is acquired as a plurality of linear imaging data in synchronization with the switching of the light and dark patterns by the imaging line sensor. And
When the linear imaging data acquired by the imaging line sensor is synthesized with the frame image by the arithmetic device, a plurality of phase-specific imaging data are synthesized by synthesizing imaging data of the same phase among the plurality of phases. A method for detecting surface irregularities, wherein a surface image is generated and the same image processing is performed on the plurality of generated surface images.

Images