JP5418378B2 - Surface defect inspection apparatus, surface defect inspection method, and program - Google Patents

Description

  The present invention relates to a surface defect inspection apparatus, a surface defect inspection method, and a program.

  In the manufacturing process of strips such as steel sheets, it is necessary to detect defects that may impair the quality of the product at an early stage in the manufacturing stage and to prevent generation of defects in subsequent products by changing manufacturing conditions. . For this purpose, for example, in the production line, the inspection of the flaw is performed while moving the steel plate. Numerous inspection methods, such as electromagnetic and optical, have been developed as inspection methods for wrinkles. In particular, optical inspection methods can detect wrinkles without contact with steel plates, and can easily obtain wrinkle images. Widely used in

  A generally applied optical inspection method is based on image signals obtained by photographing the surface of a steel plate to be passed through a plurality of locations in the width direction of the steel plate being transported by an imaging device such as a CCD camera. Thus, wrinkles on the surface of the steel plate are detected (see, for example, Patent Document 1 below).

  There is also an optical cutting type shape measuring method in which a linear laser light source whose output is modulated and a delay integration type linear sensor are combined. With this measurement method, it is possible to measure the shape of a steel plate transported at high speed in the width direction and longitudinal direction, and from each light section image (stripe image) obtained from a delay-integration linear sensor, the fringe shift A technique for generating a shape image representing the shape of a steel sheet based on the above, and a luminance image representing a difference in roughness or color tone on the surface of the steel sheet from the density of the stripe image is disclosed. According to this optical section image type shape measurement method, it is possible to measure the entire length of the entire surface with high density in online shape measurement. Therefore, this shape measuring method is suitable for detection of uneven ridges on the steel sheet surface (see, for example, Patent Document 2 below).

JP-A-11-183399 JP 2005-30812 A

  By the way, in recent years, it has become more and more important to detect finer wrinkles from the viewpoint of quality inspection at the time of shipment and measures for preventing wrinkles by early detection of wrinkles. For this reason, there is a need to capture a higher resolution image.

  With regard to this point, the technique for detecting wrinkles on the surface of a steel sheet from the image brightness photographed by a CCD camera as disclosed in Patent Document 1 reveals slight irregular reflection differences and reflectance angles that occur on the surface of a minute uneven surface. There has been a problem that it is often difficult to detect the luminance change caused by the dependency.

  On the other hand, since the technology disclosed in Patent Document 2 can obtain a shape image representing the shape of the steel sheet and a luminance image representing the difference in roughness of the steel sheet surface, It can be detected. However, since there is an upper limit to the number of pixels of the delay integration type linear sensor, there is a problem that a single unit cannot pick up a high-resolution fringe image when inspecting a wide steel plate.

  Therefore, the present invention has been made in view of the above problems, and an object of the present invention is to capture a high-resolution fringe image when inspecting a wide steel plate and to detect minute surface defects. An object of the present invention is to provide a surface defect inspection apparatus, a surface defect inspection method, and a program that can be detected.

In order to solve the above problems, according to an aspect of the present invention, a linear laser beam is irradiated on a steel sheet transported on a transport line, and the linear laser light reflected by the steel sheet is imaged. A steel plate imaging device that generates a captured image of the laser beam, and image processing is performed on the captured image of the steel plate generated by the steel plate imaging device to determine whether or not a defect exists on the surface of the steel plate A plurality of laser irradiation apparatuses that irradiate the steel sheet so that the modulated linear laser beams do not overlap each other, and the plurality of laser irradiation apparatuses, respectively. When the linear laser beam is irradiated, each of the linear laser beams reflected by the steel plate is imaged, and a plurality of light cut images of the steel plate that are captured images of the laser beam are generated. The integral processing type imaging device, and the arithmetic processing unit respectively configures a fringe image from the light section images generated by the steel plate imaging device, and uses the fringe image, An image generation unit that generates a plurality of shape images that represent the uneven state of the surface and a plurality of luminance images that represent differences in roughness on the surface of the steel sheet, and the shape image, two of the defect detection processing unit for detecting surface defects on the surface, was closed, of the plurality of delayed integration type imaging apparatus for imaging a respective said line-shaped laser beam, adjacent to each other in the width direction of the steel sheet The imaging regions of the imaging devices are not overlapped with each other in the conveyance direction of the steel plate, and are overlapped with each other in the width direction of the steel plate, In the width direction of the steel sheet Range of areas in which the line-shaped laser light is irradiated, a surface defect inspection apparatus is a portion located substantially in the center with respective excluding a portion of the constant width from both ends of the plate width direction are provided .

  The arithmetic processing unit expands and contracts the shape image and the luminance image generated by the image generation unit in accordance with a conveyance speed of the steel plate in the conveyance line, and has a conveyance direction resolution regardless of the conveyance speed of the steel plate. It is preferable to further include an image expansion / contraction processing unit that generates an image that is constant.

  The defect detection processing unit, when the surface defect is detected in a region corresponding to the overlapped portion in the width direction of the steel plate among the shape images obtained from the delay integral imaging devices adjacent to each other, It is preferable to perform an integration process for integrating surface defects detected from shape images obtained from the delay integration type imaging devices adjacent to each other as one surface defect.

  The image expansion / contraction processing unit uses information on a point in time when the starting point of the steel plate is imaged and information on a conveyance speed for conveying the steel plate, and the conveyance direction resolution of the shape image and the luminance image is constant. An image may be generated.

In order to solve the above-mentioned problem, according to another aspect of the present invention, a laser beam for irradiating a plurality of modulated linear laser beams on a steel plate transported on a transport line so as not to overlap each other An irradiation step, and a light cut image that respectively images the linear laser light reflected by the steel plate with a plurality of delay integration imaging devices to generate a light cut image of the steel plate that is an image of the laser light A plurality of shape images representing a concavo-convex state of the surface of the steel sheet using the fringe images, each of which forms a fringe image from each light cut image generated in the generation step and the light cut image generation step; An image generation step for generating a plurality of luminance images each representing a difference in roughness on the surface of the steel sheet, and a defect for detecting a surface defect existing on the surface of the steel sheet using the shape image. Seen including a detection step, the respective imaging regions imaging device of, two adjacent in the width direction of the steel plate among the plurality of delayed integration type imaging apparatus for imaging a respective said line-shaped laser beam, the steel plate Are not superimposed on each other in the conveying direction of the steel sheet, and are superimposed on each other in the width direction of the steel sheet, and the range in the plate width direction of the steel sheet of each imaging region of the delay integration type imaging device is A surface defect inspection method is provided which is a portion located in a substantially central portion of a region irradiated with a linear laser beam except for a portion having a certain width from both ends in the plate width direction .

In order to solve the above-mentioned problem, according to still another aspect of the present invention, a steel sheet transported on a transport line is irradiated with a plurality of modulated linear laser beams so as not to overlap each other, the line-shaped laser beam reflected by the steel plates by imaging a plurality of delay integration type imaging device, and a light section image of the steel sheet is a captured image of the laser beam and generates respectively, the line Imaging regions of two imaging devices adjacent to each other in the width direction of the steel plate among a plurality of delay integration type imaging devices that capture each of the laser beams in a shape do not overlap each other in the conveyance direction of the steel plate And in the width direction of the steel plate, they overlap each other, and the range in the plate width direction of the steel plate of each imaging region of the delay integration type imaging device is irradiated with the linear laser light Out of the area, The portion at which the steel plate imaging apparatus capable of communicating with a computer located in a substantially central with each except the portion of constant width from both ends in the width direction, fringe from light section image generated respectively by the plurality of delay integration type imaging device Each image is configured, and using the fringe image, a plurality of shape images representing the uneven state of the surface of the steel sheet and a plurality of luminance images representing the difference in roughness on the surface of the steel sheet are generated. There is provided a program for realizing an image generation function to be performed and a defect detection function for detecting a surface defect existing on the surface of the steel sheet by using the shape image.

  As described above, according to the present invention, the image size of the shape image generated from the captured image of the steel plate is changed according to the conveyance speed of the steel plate and then used for detection of the surface defect of the steel plate, so that the width is wide. When inspecting a steel sheet, it becomes possible to obtain a high-resolution image and detect a minute surface defect.

It is the block diagram which showed the structure of the surface defect inspection apparatus which concerns on the 1st Embodiment of this invention. It is explanatory drawing which showed the steel plate imaging device which concerns on the same embodiment. It is explanatory drawing which showed the steel plate imaging device which concerns on the same embodiment. It is explanatory drawing which showed the steel plate imaging device which concerns on the same embodiment. It is explanatory drawing which showed an example of the fringe image which concerns on the embodiment. It is the block diagram which showed the structure of the image process part which concerns on the same embodiment. It is explanatory drawing for demonstrating the phase continuation process which concerns on the same embodiment. It is explanatory drawing for demonstrating the image expansion-contraction process which concerns on the same embodiment. It is explanatory drawing for demonstrating the image expansion-contraction process which concerns on the same embodiment. It is explanatory drawing for demonstrating the defect detection process which concerns on the same embodiment. It is the flowchart which showed the flow of the surface defect inspection method which concerns on the same embodiment. It is the block diagram which showed the hardware constitutions of the arithmetic processing unit which concerns on embodiment of this invention.

  Exemplary embodiments of the present invention will be described below in detail with reference to the accompanying drawings. In addition, in this specification and drawing, about the component which has the substantially same function structure, duplication description is abbreviate | omitted by attaching | subjecting the same code | symbol.

(the purpose)
Before describing the surface defect inspection apparatus and the surface defect inspection method according to the embodiment of the present invention, the object of the present invention will be briefly described first.

  In order to solve the problem that it is difficult to inspect a wide steel plate using a high-resolution image as described above, as a result of intensive studies by the inventors, a plurality of delays have been obtained. It was thought that this can be realized by taking a high-resolution light section image using an integral linear line sensor. However, in order to realize such a configuration, it is necessary to study the arrangement method of the linear laser light source and the delay integration type linear line sensor.

  For example, when a plurality of image pickup devices are provided for one illumination device, such as a wrinkle inspection device using a plurality of ordinary CCD cameras, that is, a single linear laser light source is used to provide a plurality of delay integration type linear devices. Consider a case in which a plurality of sensors are arranged along the width direction of the steel plate. When such a configuration is adopted, the following problems occur.

  That is, since the linear laser light source irradiates a fan-shaped plate-shaped light beam, it is necessary to increase the distance between the steel plate and the light source in order to widen the irradiation area, and the amount of light irradiated to the steel plate is insufficient. There is a problem. Further, if the irradiated laser beam diverges and becomes thicker than the photographing pixel size, the stripes formed by the laser lines are blurred, and there is a problem that a high-resolution image cannot be photographed.

  On the other hand, let us consider a case where one delay integration type linear sensor and one linear laser light source are each set as one set of imaging units, and a plurality of imaging units are arranged in a line along the steel plate width. In such a case, it is necessary to precisely match the irradiation positions of the respective linear lasers so that the adjacent imaging units are not adversely affected, and to synchronize the modulation of the linear lasers. It takes time to make adjustments, and there are problems such as high maintenance costs and equipment costs.

  Further, the linear laser light source is usually realized by forming the spot light irradiated from the point laser light source into a linear shape using a rod lens or a cylindrical lens. However, the light from a normally available laser light source has a non-uniform light intensity distribution in the cross section of the light beam, and there is a problem in that the light intensity increases or decreases in the vicinity of the side edge of the irradiated linear laser compared to the central portion. There was also a problem that the linear width widened.

  In addition, since the fringe image captured by the linear laser light source and the delay integration type linear sensor is always generated at a constant rate regardless of the line speed, the resolution in the longitudinal direction of the image obtained as a result of image processing is It will vary depending on the transport speed. Therefore, if the steel plate to be inspected is not transported at a constant speed, an image that expands or contracts in the longitudinal direction in accordance with the transport speed of the steel plate is generated, and there is a problem that an accurate coordinate position of the ridge cannot be obtained. It was.

  Therefore, the present inventors provide a surface defect inspection apparatus and a surface defect inspection method capable of solving the above-described problems and obtaining a high-resolution image when inspecting a wide steel plate. As a goal, we conducted intensive studies.

  As a result, it was possible to clarify the configuration of a suitable optical system when inspecting a wide steel plate, and the inventors have conceived a surface defect inspection apparatus and a surface defect inspection method as described below. Hereinafter, a surface defect inspection apparatus and a surface defect inspection method according to an embodiment of the present invention will be described in detail with reference to FIGS.

(First embodiment)
<Overall configuration of surface defect inspection apparatus>
First, the overall configuration of the surface defect inspection apparatus 10 according to the first embodiment of the present invention will be described with reference to FIG. FIG. 1 is an explanatory view showing the configuration of the surface defect inspection apparatus according to the present embodiment.

  The surface defect inspection apparatus 10 according to the present embodiment images the steel plate 1 conveyed on the conveyance line, and performs image processing on an image obtained as a result of the imaging, so that a surface defect exists on the surface of the steel plate 1. It is a device that checks whether or not.

  As shown in FIG. 1, the surface defect inspection apparatus 10 according to the present embodiment includes a steel plate imaging apparatus 100 and an arithmetic processing apparatus 200. Further, a speed detector 300 that detects the rotational speed of the roller 2 (and hence the transport speed of the steel sheet 1) is installed on the roller 2 that transports the steel sheet 1 along the longitudinal direction of the steel sheet. Information representing the speed detection result by 300 is output to the arithmetic processing unit 200.

  The steel plate imaging device 100 is a device that sequentially images the steel plate 1, generates a light section image at each position in the longitudinal direction of the steel plate, and outputs it to the arithmetic processing device 200. In this steel plate imaging apparatus 100, the imaging timing of the steel sheet is controlled by the arithmetic processing unit 200.

  Further, the arithmetic processing device 200 generates a fringe image as described later using the light section image generated by the steel plate imaging device 100, and performs image processing as described below on the fringe image. Thus, the apparatus detects a surface defect that may exist on the surface of the steel plate 1.

  The speed detector 300 is installed on the roller 2 used for transporting the steel plate 1 and is a device that detects the rotational speed of the roller 2 (and hence the transport speed of the steel plate 1). An example of the speed detector 300 is a PLG (Pulse Logic Generator). The PLG is a detector that is installed on the rotating shaft of the roller 2 and outputs a pulse when the roller 2 rotates. Therefore, the rotational speed of the roller 2 can be specified by using the number of output pulses.

  In the following description of the present specification, the case where PLG is used as the speed detector 300 will be described. However, the speed detector according to the present invention is not limited to the PLG.

[Configuration of Steel Plate Imaging Device 100]
Next, the configuration of the steel plate imaging apparatus 100 according to the present embodiment will be described in detail with reference to FIGS. 2-4 is explanatory drawing which showed the steel plate imaging device 100 which concerns on this embodiment. Here, FIG. 2 is a perspective view showing the positional relationship between the steel plate imaging device 100 and the steel plate 1. 3 is a top view when the steel plate imaging device 100 and the steel plate 1 are viewed from above the steel plate 1, and FIG. 4 is a case when the steel plate imaging device 100 and the steel plate 1 are viewed from the side of the steel plate 1. FIG.

  As shown in FIGS. 2 to 4, the steel plate imaging apparatus 100 according to this embodiment includes a plurality of laser light sources 101, rod lenses 103, and delay integration imaging apparatuses 105. As apparent from FIGS. 2 to 4, the laser irradiation device including the laser light source 101 and the rod lens 103 and the delay integration type imaging device 105 form a set, and the delay integration type imaging device 105 is compatible. The reflected light of the linear laser light irradiated from the laser irradiation apparatus to be imaged.

  As the laser light source 101, for example, a CW laser light source that continuously performs laser oscillation can be used. The wavelength of light oscillated by the laser light source 101 is preferably a wavelength belonging to the visible light band of about 400 nm to 800 nm, for example. The laser light source 101 oscillates laser light based on an irradiation timing control signal sent from the arithmetic processing unit 200 described later.

  The rod lens 103 is a lens that spreads the laser light emitted from the laser light source 101 in a fan shape along the width direction of the steel plate 1 that is the inspection object (the direction perpendicular to the paper surface of FIG. 1). As a result, the laser light emitted from the laser light source 101 becomes a linear laser light and is irradiated onto the steel plate 1. Here, as shown in FIG. 3, the irradiation range of the linear laser light is assumed to be a range indicated by a thick line in the drawing. In the steel plate imaging apparatus 100 according to the present embodiment, a lens other than a rod lens such as a cylindrical lens or a Powell lens may be used as long as the laser beam can be fanned out.

  Here, as shown in FIGS. 2 and 4, the laser light source 101 and the rod lens 103 are arranged so that the linear laser light is obliquely incident on the steel plate 1 (vertical component incident angle: θ). Yes.

  In the portion irradiated with the linear laser beam on the surface of the steel plate 1 that is the inspection object, a linear bright portion is formed along the width direction of the steel plate 1. Moreover, since the steel plate 1 is conveyed along the longitudinal direction of the steel plate 1, when viewed from the steel plate 1, the linear bright part is also moved along the longitudinal direction of the steel plate 1. The reflected light (linear reflected image) from the linear bright part propagates to the delay integration type imaging device 105 and is imaged by the delay integration type imaging device 105.

  A delay integration type camera 105, which is an example of a delay integration type (Time Delay Integration: TDI) imaging device, captures a linear reflection image (generated by a corresponding laser irradiation device) of the steel plate 1 being conveyed. The TDI camera 105 includes a two-dimensional light receiving surface on which a large number of photoelectric conversion elements are arranged in a matrix. When the linear reflection image of the steel plate 1 enters the photoelectric conversion element with a width corresponding to one column through the lens of the TDI camera 105, each photoelectric conversion element of the TDI camera 105 photoelectrically converts the charge accumulated therein. The data is transferred to the photoelectric conversion element located in the same row as the element and located in the next column. The timing of this transfer is the same for all photoelectric conversion elements, and is controlled by a camera shift pulse signal sent from the arithmetic processing unit 200. That is, each time a camera shift pulse signal is input, each photoelectric conversion element transfers the charge one column at a time. When the camera shift pulse signal is input, the photoelectric conversion element located in the last column reads out the accumulated charge and outputs it to the arithmetic processing unit 200. As a result, a light cut image corresponding to the linear reflection image is output to the arithmetic processing device 200.

  Here, since the steel plate 1 is moving along the longitudinal direction of the steel plate, a laser beam is irradiated onto the steel plate 1 from the laser light source 101, and a linear reflection image of the steel plate 1 is captured for a certain time using the TDI camera 105. Then, the light cutting image in each position of the longitudinal direction of the steel plate 1 can be obtained sequentially. By arranging each light-cut image obtained in this manner in the horizontal direction in the vertical state, an entire image of the region of the steel sheet 1 captured by one TDI camera 105 can be obtained. In addition, an image of the entire steel plate 1 can be generated using an image acquired from each TDI camera 105.

  In general, in the TDI camera 105, when light is incident on each photoelectric conversion element while charges are being transferred, a charge corresponding to the intensity of the incident light is added. However, in the shape measuring apparatus 10 according to the present embodiment, as described above, a linear reflection image having a width corresponding to one column is incident on the photoelectric conversion elements, so that charges are added to the photoelectric conversion elements during the transfer of charges. Little happens. An optical bandpass filter that transmits only the wavelength of the laser beam may be provided in front of the TDI camera 105.

  Further, since the linear laser light is periodically modulated and the intensity of the linear laser light changes with time, the charges accumulated in the photoelectric conversion elements in the column direction in each row on the light receiving surface of the TDI camera 105. The distribution of the quantity (that is, the received light intensity) also changes periodically. For this reason, an image obtained by sequentially arranging the light cut images output from the TDI camera 105 in the horizontal direction in the vertical direction is the density of each light cut image (that is, the horizontal direction of the image (that is, (Corresponding to intensity) is a fringe image that periodically changes.

  Here, the arrangement of the plurality of laser light sources 101 (including the rod lens 103) and the plurality of TDI cameras 105 will be described with reference to FIGS.

  Since the laser light is coherent, when a plurality of laser light sources 101 are arranged and the light emitted from each laser light source 101 is superimposed, interference is caused by the superimposed laser light, and mutual interference occurs. It will have an adverse effect. Therefore, in order to eliminate such adverse effects, in the steel plate imaging apparatus 100 according to the present embodiment, as shown in FIGS. 2 to 4, the laser light sources emitted from the laser light sources 101 are not overlapped with each other as shown in FIGS. 101 and a rod lens 103 are installed. Here, the projection angle ψ of the linear laser light emitted from each rod lens 103 can be set to about 60 degrees, for example.

  The TDI camera 105 captures the linear reflected light of the linear laser light emitted from the corresponding laser light source 101 on the steel plate 1, but as illustrated in FIGS. 2 to 4, the imaging range of the TDI camera 105. LAs are set so that they do not overlap each other (ie, the camera fields of view do not overlap). As a result, the plurality of TDI cameras 105 capture an image of an independent location without receiving disturbance due to the linear laser applied to the adjacent imaging range.

  Here, the relationship of the imaging range LA of the TDI camera 105 will be described in more detail with reference to FIG.

  As shown in FIG. 3, the imaging range LA of each TDI camera 105 is narrower than the irradiation range of the corresponding laser light source 101 (the range indicated by the thick line in FIG. 3), and the vicinity of the side edge of the linear laser light is Not included. Since the side edge of the linear laser beam may have a broadened line width or a decrease in brightness, it is possible to take an image of a substantially central portion and not to image the vicinity of the side edge. A clear image can be taken. The region in the vicinity of the side end portion that is not imaged by the TDI camera 105 can be, for example, about 10 to 15% of the irradiation width of the linear laser beam from the side end of the linear laser beam inward. In this case, the length in the width direction of the imaging range LA of the TDI camera 105 is about 70 to 80% of the irradiation width of the linear laser beam.

  As shown in FIG. 3, the imaging range LA of each TDI camera 105 does not overlap in the conveyance direction of the steel plate 1 but partially overlaps in the width direction of the steel plate 1. That is, if the length of the imaging range LA in the conveyance direction is h and the offset value of the imaging range LA shown in FIG. The overlap margin OV in the width direction of the imaging range LA can be set as appropriate. For example, the size of surface defects that may occur on the surface of the steel plate 1 to be inspected is clarified in advance by statistical processing or the like. The length that can cover the standard surface defect obtained as a result of the statistical processing can be set as the length of the overlap margin OV. A specific example of such an overlap allowance OV can be about 10 mm to 20 mm.

  Further, since the width direction of the steel plate 1 is imaged using the plurality of laser light sources 101, it is not necessary to set the projection angle ψ of the laser light source 101 to an extremely large value, and the laser light source 101 is installed in the vicinity of the steel plate 1. Is possible. Therefore, it is possible to prevent the linear beam profile from deteriorating. In addition, the entire width direction of the steel sheet 1 can be shared by a plurality of TDI cameras 105 without imaging the vicinity of the side end where the line width may be broadened or the luminance may be lowered. Even with an output laser, sufficient luminance can be obtained.

  FIG. 5 shows an example of a fringe image generated as described above. Here, the stripe is a light section image corresponding to one period of density change. In such a fringe image, the vertical direction, that is, the direction parallel to the fringe corresponds to the width direction of the steel plate 1 that is the inspection object, and the horizontal direction, that is, the direction orthogonal to the fringe is the steel plate 1 that is the inspection object. Corresponds to the longitudinal direction. If the ratio of the camera shift frequency of the TDI camera 105 to the modulation frequency of the laser light is M: 1, M light-cut images, that is, M pixels in the horizontal direction form one stripe.

  Here, since the linear laser light is incident on the surface of the steel plate 1 as an inspection object from an oblique direction (vertical component incident angle: θ), for example, when a concave portion exists in the steel plate 1, the reflection point of the linear laser light is It will shift to the right. As a result, the position of the light section image on the photoelectric conversion element of the TDI camera 105 is also shifted rightward, that is, in the column direction. For this reason, in the fringe image, the light cut image corresponding to the linear laser light reflected by the concave portion is output earlier in time than the light cut image corresponding to the linear laser light reflected by the flat portion other than the concave portion. Will be. Therefore, in the two-dimensional image obtained by sequentially arranging the one-dimensional images output from the TDI camera 105 in the horizontal direction, the concave portions present in the steel plate 1 can be recognized as stripe deviation. For example, in FIG. 5, the portion where the stripes are bent corresponds to a concave portion present in the inspection object.

  The vertical component incident angle θ of the linear laser beam on the steel plate 1 is determined to an appropriate value in consideration of the surface roughness of the inspection object. For example, if the surface roughness of the object to be inspected is in a mirror state, a fringe image composed of stripes having luminance can be obtained by reducing the vertical component incident angle θ. On the contrary, when the surface roughness of the inspection object is high, the vertical component incident angle θ is increased. When applied to a general steel plate surface, if the angle is 45 degrees, the stripes can obtain sufficient luminance. Further, by setting the vertical component incident angle to 45 degrees, the depth change amount of the steel plate 1 as the inspection object becomes equal to the movement amount of the stripe, and the depth of the concave portion existing in the steel plate 1 can be easily determined from the movement amount of the stripe. Convenient information can be obtained.

  Below, the specific structure is enumerated about each apparatus which the steel plate imaging device 100 which concerns on this embodiment has. Such a configuration is merely an example, and the steel plate imaging apparatus 100 according to the present invention is not limited to the following specific examples.

-Laser irradiation device Laser light is irradiated at a vertical component incident angle θ = 45 degrees and a projection angle ψ = 60 degrees. Distance to steel plate 1 = 520 mm, irradiation width of linear laser beam = 600 mm
O TDI camera 2048 bits × 96 bits, shooting area LA of each camera = 410 mm × 48 mm (shooting resolution 0.2 mm × 0.2 mm). The overlapping range OV in the steel sheet width direction is OV = 16 mm, and the entire TDI camera has an imaging area of 1198 mm. The ratio between the TDI camera's camera shift frequency (= 18.7 kHz) and the laser beam modulation frequency (= 4675 Hz) is M: 1 ( = 4: 1), and the TDI camera captures a fringe image regardless of the conveying speed of the steel plate. That is, the vertical resolution of the fringe image varies depending on the conveyance speed.

[Overall configuration of arithmetic processing unit]
The configuration of the steel plate imaging device 100 has been described above. Next, returning to FIG. 1 again, the configuration of the arithmetic processing apparatus 200 provided in the surface defect inspection apparatus 10 according to the present embodiment will be described in detail.

  The arithmetic processing apparatus 200 according to the present embodiment mainly includes a timing signal generation unit 201, an image processing unit 203, a display unit 205, and a storage unit 207, for example, as illustrated in FIG.

  The timing signal generation unit 201 is realized by, for example, a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), a communication device, and the like. The timing signal generator 201 generates a sine waveform signal having a predetermined frequency ω, and sends the generated sine waveform signal to the laser light source 101. Since the laser light source 101 is capable of continuously changing the oscillation intensity by an irradiation timing control signal input from the outside, the sine wave signal is received by receiving the sine waveform signal sent from the timing signal generator 201. It is possible to oscillate laser light whose output changes with the waveform. That is, the timing signal generator 201 can periodically modulate the laser light emitted from the laser light source 101 by sending the generated sine waveform to the laser light source 101.

  In addition, the timing signal generation unit 201 generates a rectangular waveform having a frequency M times the frequency ω as a camera shift pulse signal, and sends the generated camera shift pulse signal to the TDI camera 105.

  As described above, the timing signal generation unit 201 can be said to be a drive control unit that controls the driving of the laser light source 101 and the TDI camera 105 provided in the steel plate imaging apparatus 100.

  The image processing unit 203 is realized by, for example, a CPU, a ROM, a RAM, a communication device, and the like. The image processing unit 203 uses the steel plate acquired from the speed detector 300 for the fringe image generated using the imaging data acquired from the steel plate imaging device 100 (more specifically, the TDI camera 105 of the steel plate imaging device 100). The image processing described below is performed using the information on the conveyance speed of 1 to detect defects present on the surface of the steel sheet 1 that is the inspection object. When the image processing on the striped image corresponding to the steel plate 1 is completed, the image processing unit 203 transmits information about the obtained inspection result of the steel plate 1 to the display unit 205.

  The image processing unit 203 acquires images from a plurality of TDI cameras 105 provided in the steel plate imaging apparatus 100 and performs image processing as described below. In advance, information indicating in what order and in which order a plurality of TDI cameras 105 are installed (that is, what position on the steel plate the image acquired from a certain TDI camera 105 is captured) is set. It is assumed that

  The image processing unit 203 will be described in detail later again.

  The display unit 205 is realized by, for example, a CPU, a ROM, a RAM, an output device, and the like. The display unit 205 displays the inspection result of the steel plate 1 that is the inspection object transmitted from the image processing unit 203 on an output device such as a display included in the arithmetic processing device 200. Thereby, the user of the surface defect inspection apparatus 10 can grasp the inspection result regarding the surface defect of the inspection object (steel plate 1) being conveyed on the spot.

  The storage unit 207 is an example of a storage device included in the arithmetic processing device 200. The storage unit 207 appropriately records various parameters, the progress of processing, or various databases that need to be saved when the arithmetic processing apparatus 200 according to the present embodiment performs some processing. The The storage unit 207 can be freely read and written by the timing signal generation unit 201, the image processing unit 203, the display unit 205, and the like.

[About the image processing unit]
Next, the image processing unit 203 included in the arithmetic processing apparatus 200 according to the present embodiment will be described in detail with reference to FIG. FIG. 6 is a block diagram illustrating a configuration of an image processing unit included in the arithmetic processing apparatus according to the present embodiment.

  As illustrated in FIG. 6, the image processing unit 203 according to the present embodiment mainly includes an image generation unit 209, an image expansion / contraction processing unit 229, and a defect detection processing unit 231.

  The image generation unit 209 is realized by, for example, a CPU, a ROM, a RAM, a communication device, and the like. The image generation unit 209 generates a plurality of shape images representing the uneven state of the surface of the steel plate 1 using a fringe image formed from the light section image generated by the steel plate imaging device 100.

  The image generation unit 209 includes an A / D conversion unit 211, a pre-filter unit 213, an orthogonal sine wave generation unit 215, low-pass filter units 217 and 219, a phase calculation unit 221, an amplitude calculation unit 223, a fringe defect determination unit 225, and a phase A continuous processing unit 227 is further provided.

  The A / D conversion unit 211 is realized by, for example, a CPU, a ROM, a RAM, and the like. The A / D conversion unit 211 performs A / D conversion on each light section image output from the TDI camera 105 and outputs it as digital multivalued image data. Such digital multivalued image data is stored in an image memory provided in the storage unit 207 or the like. By arranging these digital multivalued image data in order, a striped image as shown in FIG. 5 is formed. For example, when a 2048 bit × 96 bit camera is used as the TDI camera 105, the A / D conversion unit 211 forms a fringe image as one frame image of 2048 lines for each 2048 pulses of the camera shift frequency timing signal. .

  From such a stripe image (or digital multi-valued image data), data representing the density distribution of the stripe image along the horizontal direction is generated at each position in the vertical direction. Hereinafter, the data representing the density distribution of the stripe image along the horizontal direction will be referred to as “slice stripe image data”. The slice fringe image data at each position in the vertical direction is sequentially output from an image memory provided in the storage unit 207 or the like.

  The prefilter unit 213 is realized by, for example, a CPU, a ROM, a RAM, and the like. The pre-filter unit 213 performs predetermined filter processing on each slice stripe image data, thereby removing noise from each slice stripe image data and making the stripe state clear. Note that the filtering process by the prefilter unit 213 is not necessarily performed. For example, the filtering process may be performed only when a lot of fine noise is generated in the striped image.

When the noise removal processing on the slice fringe image data is completed, the prefilter unit 213 outputs the slice fringe image data I _j (k) at each position j (j = 0, 1, 2,...) In the vertical direction. . Here, k (k = 0, 1, 2,...) Is a parameter representing the position in the horizontal direction. At this time, it is assumed that the slice fringe image data I _j (k) at the position j in the vertical direction changes sinusoidally. That is, I _j (k) is expressed as shown in Equation 101 below.

・・・（式１０１）

... (Formula 101)

  Here, in Equation 101 above, A (j, k) represents the amplitude of the slice stripe image data at the pixel position (j, k), and φ (j, k) represents the slice at the pixel position (j, k). This represents the phase shift of the fringe image data.

Here, the influence of the fringe shift generated in the fringe image due to the dent on the surface of the steel plate 1 appears as a phase shift φ. Further, although the amplitude of the linear laser beam is constant, when the surface of the steel plate 1 is dirty or there are pattern-like wrinkles such as scale wrinkles, the amplitude is at the pixel position corresponding to the position. May fluctuate. Therefore, in the above equation 101, the amplitude A is expressed in a form depending on the pixel position (j, k). In the above equation 101, the reason why 1 is added after the cos term is to satisfy the condition that the slice fringe image data (density value) I _j (k) does not become negative. Therefore, the slice fringe image data I _j (k) changes between 0 and 2A.

  The orthogonal sine wave generation unit 215 is realized by, for example, a CPU, a ROM, a RAM, and the like. The orthogonal sine wave generation unit 215 generates two orthogonal reference sine wave data sin (2πk / M) and cos (2πk / M) generated in advance in the storage unit 207 or the like. Hereinafter, the former sine wave data is referred to as reference sin data, and the latter sine wave data is referred to as reference cos data.

These two types of reference sine wave data are respectively multiplied by the slice fringe image data I _j (k) output from the prefilter unit 213. By this multiplication processing, two outputs I _aj (k) and I _bj (k) are generated. The details of I _aj (k) and I _bj (k) are as shown in Equations 102 and 103 below.

・・・（式１０２）

・・・（式１０３）

... (Formula 102)

... (Formula 103)

Here, the output data I _aj (k) expressed by the equation 102 is input to a low-pass filter unit 217 described later, and the output data I _bj (k) expressed by the equation 103 is input by a low-pass filter unit described later. 219 is input.

The low-pass filter units 217 and 219 are realized by, for example, a CPU, a ROM, a RAM, and the like. The low-pass filter units 217 and 219 perform predetermined filter processing on the outputs I _aj (k) and I _bj (k) obtained by the above-described multiplication processing, thereby removing the fringe frequency component and its harmonic components. . By such processing, it is possible to extract a component including only the phase shift φ from the input data I _aj (k), I _bj (k). Here, the low-pass filter unit 217 is a processing unit that performs a predetermined filter process on the input data I _aj (k), and the low-pass filter unit 219 applies the input data I _bj (k). And a processing unit that performs predetermined filter processing.

_Assuming that the output from the low-pass filter unit 217 is LPF (I _aj (k)) and the output from the low-pass filter unit 219 is LPF (I _bj (k)), these are expressed by the following equations 104 and 105: expressed.

・・・（式１０４）

・・・（式１０５）

... (Formula 104)

... (Formula 105)

The low pass filter unit 217 outputs the data LPF (I _aj (k)) obtained by the filtering process to a phase calculation unit 221 and an amplitude calculation unit 223 described later. The low-pass filter unit 219 outputs the data LPF (I _bj (k)) obtained by the filtering process to a phase calculation unit 221 and an amplitude calculation unit 223 described later.

The phase calculation unit 221 is realized by, for example, a CPU, a ROM, a RAM, and the like. The phase calculation unit 221 calculates a phase shift φ (j, k) at each pixel position (j, k) based on the results output from the low-pass filter units 217 and 219. The phase shift φ (j, k) can be calculated by the following Expression 107 to Expression 109 according to the values of LPF (I _aj (k)) and LPF (I _bj (k)).

Here, the following expression 107 is an expression used by the phase calculation unit 221 when LPF (I _aj (k)) ≧ 0. Further, the following formula 108 is a formula used by the phase calculation unit 221 when LPF (I _aj (k)) <0 and LPF (I _bj (k)) <0. Further, the following formula 109 is a formula used by the phase calculation unit 221 when LPF (I _aj (k)) <0 and LPF (I _bj (k)) ≧ 0.

・・・（式１０７）

・・・（式１０８）

・・・（式１０９）

... (Formula 107)

... (Formula 108)

... (Formula 109)

The phase calculation unit 221 sets the range of the inverse trigonometric function (arctan) to −π / 2 to + π / 2 in the above formulas 107 to 109, and LPF (I _aj (k)), LPF (I _bj (k )) Is used to obtain the phase shift φ in the range of −π to + π. Here, the phase shift obtained in this range is represented by φ ′ again. In this case, the phase shift φ ′ obtained by the above formulas 107 to 109 has a periodic relationship with the dent (depth) of the surface of the steel sheet 1, and the depth at which the phase shift φ ′ takes a certain value. There are multiple sizes. Therefore, when such a phase shift φ ′ is used, accurate information on the surface shape of the steel sheet 1 cannot be obtained. For this reason, it is necessary to obtain a phase shift φ that is proportional to the dent (depth) of the surface of the steel sheet 1 from the phase shift φ ′. The process of obtaining the phase shift φ proportional to the depth is performed by the phase continuation processing unit 227 described later.

  Therefore, the phase calculation unit 221 outputs information on the calculated phase shift φ ′ to the phase continuation processing unit 227 described later. An image can be generated by illustrating the phase shift φ ′ (j, k). An image generated based on such a phase shift φ ′ is hereinafter referred to as a phase image. There are various methods for illustrating the phase shift φ ′ (j, k). For example, the image becomes white when the phase shift φ ′ is + π, and the image is displayed when the phase shift φ ′ is −π. It can be illustrated as a grayscale image that becomes black.

  The amplitude calculation unit 223 is realized by a CPU, a ROM, a RAM, and the like, for example. The amplitude calculation unit 223 calculates the amplitude A (j, k) at each pixel position (j, k) based on the results output from the low-pass filter units 217 and 219. The amplitude calculation unit 223 calculates the amplitude A (j, k) by the following expression 110.

・・・（式１１０）

... (Formula 110)

  The amplitude calculation unit 223 outputs information on the amplitude A calculated in this way to the fringe defect determination unit 225 and the image expansion / contraction processing unit 229 described later. An image can be generated by illustrating the amplitude A (j, k). Such an image generated based on the amplitude A is hereinafter referred to as an amplitude image. There are various methods for illustrating the amplitude A (j, k). For example, the amplitude A (j, k) can be illustrated as a grayscale image in which the image becomes black as the amplitude A is small.

  Further, in the arithmetic processing device 200 according to the present embodiment, the amplitude image that can be generated in this way is used as a luminance image that is an image representing the difference in roughness on the surface of the steel plate.

  By the way, when the surface of the steel plate 1 is soiled with oil, the region of the striped image corresponding to the soiled region may be crushed in black. In such a region, since the amplitude A is small and the phase shift φ ′ changes abruptly between adjacent pixel positions, the phase shift φ ′ calculated in the corresponding region is not reliable. In such a dirty area, a lot of noise is generated and the image becomes rough. Therefore, it is desirable to remove such an unreliable region of the phase image from the detection target.

  Such an unreliable region in the phase image can be obtained from the amplitude image. That is, an unreliable phase region can be obtained by specifying a region having an extremely small amplitude. For example, in an amplitude image, a region corresponding to an unreliable region of the phase image should be black compared to other regions, and such an amplitude image is used to identify a region whose shape should be excluded from the measurement target. It is possible to use.

  The fringe defect determination unit 225 is realized by a CPU, a ROM, a RAM, and the like, for example. The fringe defect determination unit 225 determines an unreliable region in the phase image as described above based on the amplitude image. Specifically, the fringe defect determination unit 225 binarizes the amplitude image using a predetermined threshold. As this threshold value, for example, a small value is set according to the dirt on the surface. Further, contraction processing or the like is performed on the binary image as necessary. Then, the fringe defect determination unit 225 determines an area having a value smaller than the threshold based on the binary image thus obtained, and extracts the area as an unreliable area (deletion area) in the phase image.

  The phase continuation processing unit 227 is realized by a CPU, a ROM, a RAM, and the like, for example. For example, as illustrated in FIG. 7, the phase continuation processing unit 227 detects discontinuous points of the phase shift φ ′ based on the phase image calculated by the phase calculation unit 221, and the phase shift φ ′. The phase shift φ ′ is corrected so as to be smoothly connected.

  As described above, since the value range of the phase shift φ ′ calculated by the phase calculator 221 is −π to + π, the phase shift φ ′ is not −π and + π, as shown in FIG. It will be continuous. For example, in a phase image represented as a grayscale image, a portion that changes from white (or black) to black (or white) corresponds to a discontinuous point of the phase shift φ ′. If such a phase image is used as it is, it is difficult to recognize the surface shape of the steel plate 1. Accordingly, it is necessary to correct the phase shift φ ′ so that the phase shift φ ′ is smoothly connected at the discontinuous points of the phase shift φ ′. Such correction (phase jump correction) is a process for obtaining a unique phase shift φ proportional to the dent (depth) of the surface of the steel sheet 1 from the phase shift φ ′ defined in the range of 2π.

  Specifically, the phase continuation processing unit 227 first masks a region corresponding to the defect region obtained by the fringe defect determination unit 225 in the phase image calculated by the phase calculation unit 221. By such a mask process, a region other than the masked region becomes a target of phase skip correction.

  Next, the phase continuation processing unit 227 detects a discontinuous point of the phase shift φ ′ and corrects the phase shift φ ′ at the discontinuous point. It is difficult to determine whether or not the phase shift φ ′ is discontinuous even if only one pixel is referred to, and it is preferable to determine by referring to adjacent pixels. Therefore, the phase continuation processing unit 227 examines the phase image along the horizontal direction at each position in the vertical direction of the phase image, and compares the phase shift φ ′ between adjacent pixels. When the phase shift φ ′ differs greatly between the adjacent pixels, it is determined that the phase shift φ ′ is discontinuous between the pixels, and the phase shift φ ′ is corrected. Here, since the depth at the surface of the steel plate 1 does not change abruptly, it is considered that the phase shift φ ′ is largely different because the phase shift φ ′ changes by ± 2π. Therefore, the phase continuation processing unit 227 examines pixels in which the phase shift φ ′ is significantly different from the phase shift φ ′ of the adjacent pixels, and smoothly connects these phase shifts φ ′. You can do it.

  For example, when the phase shift φ ′ is close to + π at a certain pixel position and the phase shift φ ′ is close to −π at the pixel position on the right side of the pixel position, the phase continuation processing unit 227 recognizes that the phase shift φ ′ is changed by + 2π at the pixel position on the right side. The phase continuation processing unit 227 corrects the phase shift φ ′ by adding + 2π to the phase shift φ ′ at the pixel position on the right side. Further, when the phase shift φ ′ is a value close to −π at a certain pixel position and the phase shift φ ′ is a value close to + π at the pixel position to the right of the pixel position, the phase continuation processing unit 227 recognizes that the phase shift φ ′ is changed by −2π at the pixel position on the right side. Then, the phase continuation processing unit 227 corrects the phase shift φ ′ by adding −2π to the phase shift φ ′ at the pixel position on the right side.

  The phase continuation processing unit 227 examines adjacent pixels along the horizontal direction at each position in the vertical direction by the method described above, corrects the phase shift φ ′, and then performs vertical correction at each position in the horizontal direction. The adjacent pixels along the direction are examined, and the phase shift φ ′ is corrected in the same manner. The phase shift at each pixel position after correction is a unique phase shift φ proportional to the dent (depth) of the surface of the steel plate 1.

  Next, the phase continuation processing unit 227 newly creates a phase image based on the corrected phase shift φ. This new phase image accurately represents the surface shape of the steel plate 1. This new phase image is hereinafter referred to as a shape image.

  The phase continuation processing unit 227 outputs the shape image whose phase shift is corrected to the image expansion / contraction processing unit 227 described later.

  The image expansion / contraction processing unit 229 is realized by, for example, a CPU, a ROM, a RAM, and the like. The image expansion / contraction processing unit 229 expands / contracts the shape image and the luminance image generated by the image generation unit 209 according to the conveyance speed of the steel plate 1 in the conveyance line, and considers the conveyance speed of the steel plate in the image size of these images. Change to the selected image size.

  In addition, the image expansion / contraction processing unit 229 includes a camera shift frequency timing signal acquired from the timing signal generation unit 201, an online PLG signal (hereinafter also referred to as a PLG signal) having a predetermined pulse rate acquired from the speed detector 300, and the like. It also has a counter function for counting. In this embodiment, a PLG signal having a pulse rate of 0.1 mm is used. The image expansion / contraction processing unit 229 stores the above-described camera shift frequency timing signal and the PLG signal using the RAM, the storage unit 207, and the like. In the following, it is assumed that the camera shift frequency timing signal is stored in the storage area CH1, and the PLG signal is stored in the storage area CH2.

  In addition, a welding point signal (not shown) indicating the leading position of the steel sheet 1 being conveyed on the conveyance line is input to the image expansion / contraction processing unit 229 as a trigger from an external device or the like.

  The image expansion / contraction processing unit 229 uses the time point when the starting point of the steel plate is imaged (that is, the time point when the welding point signal is input) and the conveyance speed (PLG signal) acquired from the speed detector 300, and the shape image and The image size is changed so that the resolution in the conveyance direction of the luminance image is constant. In order to make the resolution in the conveyance direction constant, the image expansion / contraction processing unit 229 initializes the count of the number of pulses of the PLG signal to zero at the time when the starting point of the steel plate is imaged, and handles the number of pulses from zero to a predetermined threshold value. The image size is changed so that the stripe image line becomes one shape image or luminance image.

  Hereinafter, the image size changing process performed by the image expansion / contraction processing unit 229 will be described in detail with reference to FIGS. 8 and 9. In the following example, a shape image will be described as an example, but it goes without saying that the same processing can be performed on a luminance image.

  The image expansion / contraction processing unit 229 performs the following image processing in parallel on the 2048-line striped image captured by each TDI camera 105 in accordance with information obtained from the storage areas CH1 and CH2. That is, the image expansion / contraction processing unit 229 is caused by a change in the conveyance speed of the steel plate 1 each time a 512-line shape image and a luminance image frame generated from the 2048-line stripe image are input from the image generation unit 209. The expansion / contraction processing is performed so that the vertical resolution is constant.

  More specifically, the image expansion / contraction processing unit 229 stacks the shape image (or luminance image) subjected to the expansion / contraction processing on an image processing frame buffer provided in the RAM, the storage unit 207, or the like. Here, the image processing frame is an image frame that has been stretched and contracted to a constant vertical resolution, which is used when performing defect detection sequentially in post-processing (that is, defect detection processing). As soon as the 2048-line depth image processing frame generated from the shape image and the 2048-line luminance image processing frame generated from the luminance image are generated, the image expansion / contraction processing unit 229 converts the generated image processing frame to It outputs to the defect detection process part 231 mentioned later.

  Hereinafter, how the shape image generated from the striped image based on the output data of one TDI camera 105 is processed will be described in detail, but the luminance image is similarly processed at the same time.

  The value PR shown in FIG. 8 is a PLG count value corresponding to the vertical size of the image processing frame. On the other hand, the count value of CH2 corresponds to the distance that the steel plate is transported on the transport line. Therefore, the image expansion / contraction processing unit 229 performs processing while paying attention to the relationship between the count value of CH2 and the parameter PR, so that the length in the longitudinal direction of the steel sheet existing in one image frame can be made constant. it can. Specifically, this parameter PR is a value defined by the number of shape image processing frame lines (2048) × the shape image vertical resolution setting value / PLG pulse rate. In the present embodiment, since the steel plate 1 is imaged with a vertical resolution of 0.2 mm by the TDI camera 105, the shape image vertical resolution setting value of the shape image obtained after the expansion / contraction process is set to 0.2 mm. Therefore, the parameter PR is 4096.

  When the welding point signal is input to the image expansion / contraction processing unit 229, the image expansion / contraction processing unit 229 reads the count number WP of CH1 by referring to the RAM, the storage unit 207, and the like (the upper graph in FIG. 8). The value is stored in a storage area for storing the previous reading value (hereinafter also referred to as a previous reading value memory). The image expansion / contraction processing unit 229 sets the WP / 4 line as the starting point of the expansion / contraction image. The reason why WP is divided by 4 is that, in the present embodiment, the ratio of the camera shift frequency to the modulation frequency of the laser light is M: 1 = 4: 1.

  In addition, the image expansion / contraction processing unit 229 clears the CH2 previous reading value memory and the CH2 count number. Further, the image expansion / contraction processing unit 229 may detect the welding point from the shape image instead of the welding point signal. When detecting the welding point from the shape image, the number of lines at the welding point position of the detected shape image is stored in the previous reading value memory of CH1, and at the same time, the previous reading value memory of CH2 and the CH2 count number are cleared.

  Here, consider a case where the count value of CH1 reaches 2048 as shown in the upper graph of FIG. In such a case, it means that the size of the image data stored in the storage area CH1 has reached the capacity of the storage area CH1. When the CH1 count value reaches 2048, the image expansion / contraction processing unit 229 calculates a CH1 increment value ΔK1 obtained by subtracting the previous reading value WP from 2048 that is the current reading value, and similarly, calculates the CH2 increment value ΔP1. calculate. At the same time, the image expansion / contraction processing unit 229 clears the CH1 count number and the CH1 previous reading value memory.

  Here, PLG pulse rate × ΔP1 calculated from the CH2 increment value ΔP1 shown in the lower graph of FIG. 8 is equal to the length in the longitudinal direction in which the steel plate 1 is imaged. Accordingly, the shape image corresponding to the period until the count value of CH2 increases from 0 to ΔP1, that is, the shape image corresponding to the (CH1 increment value ΔK1 / 4) line of the generated shape image, A shape image with a certain vertical resolution can be obtained by expanding and contracting the image so as to coincide with the captured length in the longitudinal direction. Therefore, as shown in FIG. 9, the image expansion / contraction processing unit 229 performs the vertical direction of “shape image 1” to be equal to the value calculated from (CH1 increment value ΔK1 / 4 × shape image vertical resolution setting value). Is expanded and contracted to generate “expanded image 1”. That is, the image expansion / contraction processing unit 229 calculates a ratio between (PLG pulse rate × ΔP1) and (CH1 increment value ΔK1 / 4 × shape image vertical resolution setting value), and “shape image 1” based on the calculated ratio. Stretch the vertical direction. Thereafter, the image expansion / contraction processing unit 229 stacks the generated “expanded image 1” in an image processing frame buffer provided in the RAM, the storage unit 207, or the like.

  Thereafter, similarly, the count number of CH1 and the count number of CH2 are added. Next, when the CH1 count reaches 2048 as shown in the upper graph of FIG. 8, the image expansion / contraction processing unit 229 first determines whether the CH2 count reaches PR. When the count number of CH2 does not reach PR, it means that the image is not stacked in the image processing frame buffer until the vertical size of the image processing frame reaches a predetermined size. , CH1 count values corresponding to values between 0 and 2048 are stacked in the image processing frame buffer. Specifically, the image expansion / contraction processing unit 229 calculates the CH1 increment value ΔK2 (= 2048) obtained by subtracting the previous reading value (= 0), reads the CH2 count number, and subtracts the previous reading value of CH2. An increment value ΔP2 is calculated. A value represented by the calculated CH2 increment value ΔP2 × PLG pulse rate is the length in the longitudinal direction in which the steel plate 1 is imaged. At the same time, the image expansion / contraction processing unit 229 clears the CH1 count number and the CH1 previous reading value memory.

  Similarly to the method described above, the image expansion / contraction processing unit 229 has a value calculated from (CH1 increment value ΔK2 / 4 × shape image vertical resolution setting value) as a value calculated from PLG pulse rate × ΔP2. The vertical direction of “shape image 2” is expanded and contracted so as to be equal, and “expanded image 2” is generated. Thereafter, as shown in FIG. 9, the image expansion / contraction processing unit 229 stacks the generated “expanded image 2” in the image processing frame buffer.

  Thereafter, the count number of CH1 and the count number of CH2 are added in the same manner, but as shown in the upper graph of FIG. 8, before the CH1 count number reaches 2048, the CH2 count number becomes PR. (= 4096) may also occur. In such a case, the capacity of the storage area CH2 is full, and the data necessary for generating one image with a fixed resolution used in the defect detection process is stored in the buffer. Represents. In such a case, the image expansion / contraction processing unit 229 reads the CH2 count number, calculates the CH2 increment value ΔP3a obtained by subtracting the previous CH2 reading value, and simultaneously clears the CH2 count number and the CH2 previous reading value memory. Thereafter, the image expansion / contraction processing unit 229 reads the count number of CH1 at this time, and calculates the CH1 increment value ΔK3a obtained by subtracting the previous reading value of CH1.

  Since PR is a PLG count value corresponding to the vertical size of the image processing frame, that the CH2 count number has reached PR (= 4096) means that the number of image processing frame lines has reached 2048 lines (maximum number of lines). Therefore, the image expansion / contraction processing unit 229 needs to expand / contract the shape image for (ΔK3a / 4) lines from the top of “shape image 3” and stack it in the image processing buffer. Accordingly, the image expansion / contraction processing unit 229 sets the “shape so that the value calculated from (CH1 increment value ΔK3a / 4 × shape image vertical resolution setting value) is equal to the value calculated from (PLG pulse rate × ΔP3a)”. The image 3 "is divided and the vertical direction is expanded and contracted to generate an" expanded image 3a ". After that, as shown in FIG. 9, the image expansion / contraction processing unit 229 stacks the generated “expandable image 3a” in the image processing frame buffer.

  Further, since the count number is cleared after the count number of CH2 reaches PR, the image expansion / contraction processing unit 229 reads the previous reading value of CH1 (that is, the reading value of CH1 when CH2 reaches PR) / 4. Set the line to the starting point of the stretchable image. That is, the image expansion / contraction processing unit 229 accumulates data for generating a second fixed-resolution image used in the defect detection process in the image processing frame buffer from this point.

  Similarly, the count number of CH1 and the count number of CH2 are added in the same manner. Since the count number of CH2 has just been cleared, it is considered that the situation where the count number of CH1 reaches 2048 arrives first. As shown in the upper graph of FIG. 8, when the count number of CH1 reaches 2048, the image expansion / contraction processing unit 229 refers to the count number of CH2 and confirms that the count number of CH2 has not reached PR. To do. Subsequently, the image expansion / contraction processing unit 229 performs processing for stacking shape images corresponding to the CH1 count value from the previous reading value to 2048 in the image processing frame buffer. Specifically, the image expansion / contraction processing unit 229 calculates the CH1 increment value ΔK3b obtained by subtracting the previous reading value, reads the CH2 count number, and calculates the CH2 increment value ΔP3b obtained by subtracting the previous reading value of CH2. A value represented by the calculated CH2 increment value ΔP3b × PLG pulse rate is the length in the longitudinal direction in which the steel plate 1 is imaged. At the same time, the image expansion / contraction processing unit 229 clears the CH1 count number and the CH1 previous reading value memory.

  Similarly to the method described above, the image expansion / contraction processing unit 229 uses a value calculated from (CH1 increment value ΔK3b / 4 × shape image vertical resolution setting value) as a value calculated from PLG pulse rate × ΔP3b. The vertical direction of the remaining part of “shape image 3” is expanded and contracted so as to be equal to generate “expanded image 3b”. Thereafter, as shown in FIG. 9, the image expansion / contraction processing unit 229 stacks the generated “expandable image 3b” in the image processing frame buffer.

  Thereafter, the count number of CH1 and the count number of CH2 are added in the same manner, but as shown in the upper graph of FIG. 8, before the CH1 count number reaches 2048, the CH2 count number becomes PR. It is assumed that (= 4096) is reached. In this case, the image expansion / contraction processing unit 229 reads the CH2 count number, calculates the CH2 increment value ΔP4a obtained by subtracting the previous CH2 reading value, and simultaneously clears the CH2 count number and the CH2 previous reading value memory. Thereafter, the image expansion / contraction processing unit 229 reads the count number of CH1 at this time, and calculates a CH1 increment value ΔK4a obtained by subtracting the previous reading value of CH1.

  Here, the image expansion / contraction processing unit 229 expands / contracts the shape image of (ΔK4a / 4) lines from the top of “shape image 4”, and stacks it in the image processing buffer. At this time, the image expansion / contraction processing unit 229 sets the value calculated from (CH1 increment value ΔK4a / 4 × shape image vertical resolution setting value) to be equal to the value calculated from (PLG pulse rate × ΔP4a). After dividing the shape image 4 ”, the vertical direction is expanded and contracted to generate the“ expanded image 4a ”. After that, as shown in FIG. 9, the image expansion / contraction processing unit 229 stacks the generated “expanded image 4a” in the image processing frame buffer.

  Thereafter, the image expansion / contraction processing unit 229 repeats the same procedure, whereby an image processing frame having a constant vertical resolution (for example, a vertical pixel size of 0.2 mm in the present embodiment) that does not depend on a change in the conveyance speed of the steel plate 1. Can be obtained for the entire length of the steel plate. In addition, the image expansion / contraction processing unit 229 performs the same process for images acquired from other TDI cameras 105, and integrates the images (images after the expansion / contraction processing) to obtain a shape image of the entire steel plate. Can be obtained. Hereinafter, the image processing frame of the shape image after the image expansion / contraction process is performed is referred to as a “depth image”.

  There are many algorithms for image expansion / contraction such as the nearest neighbor method, the bilinear method, and the bicubic method. In the image expansion / contraction processing unit 229 according to the present embodiment, an appropriate method may be selected from these algorithms in consideration of both the processing speed and the quality of the obtained image.

  With the method described above, the image expansion / contraction processing unit 229 according to the present embodiment can generate a shape image and a luminance image with a constant vertical resolution regardless of the conveyance speed of the steel plate. The image expansion / contraction processing unit 229 outputs the shape image and the luminance image after the expansion / contraction processing generated in this way to the defect detection processing unit 231 described later.

  The defect detection processing unit 231 is realized by a CPU, a ROM, a RAM, and the like, for example. The defect detection processing unit 231 calculates the depth d of the dent present on the surface of the steel plate 1 that is the inspection target based on the shape image after the expansion / contraction processing output from the image expansion / contraction processing unit 229, and the inspection target Identify the shape of the object. Thereafter, when the shape of the steel plate 1 that is the inspection target is specified, the defect detection processing unit 231 determines whether there is a surface defect on the surface of the steel plate 1.

  Here, when the imaging resolution in the column direction of the photoelectric conversion elements in the TDI camera 105 is s (mm / pixel) and the vertical component incident angle of the linear laser light is θ, the reflection point of the linear laser light is in the longitudinal direction. The shifted distance h = d · tan θ corresponds to h / s pixels in the striped image. When the ratio between the camera shift frequency of the TDI camera 105 and the modulation frequency of the laser beam is M: 1, M pixels in the horizontal direction form one stripe in the stripe image. That is, when the fringes are shifted by M pixels, the phase shift is 2π. Therefore, the phase shift Δφ in the fringe image data when the reflection point of the linear laser beam L is shifted by a distance h in the longitudinal direction is expressed by the following formula 111 from the relationship of M / 2π = (h / s) / Δφ. It becomes like this.

  d = {M · s / (2π · tan θ)} · Δφ (Formula 111)

  Therefore, the defect detection processing unit 231 includes a set value of the steel plate imaging apparatus 100 such as the imaging resolution of the TDI camera 105 and the vertical component incident angle θ of the linear laser beam, and a frequency ratio M acquired from the timing signal generation unit 201, Using the phase φ obtained from the shape image output from the image expansion / contraction processing unit 229 and the above formula 111, the depth d of the dent existing on the surface of the inspection object (steel plate 1) can be calculated.

  Strictly speaking, when a normal lens is used, the imaging resolution s changes according to the depth d, and thus correction is necessary. However, as in the case of measuring the dent of the steel plate, the lens working distance is not affected. When the depth change is very small, the change in the imaging resolution s can be ignored in practice. If a telecentric lens is used, the imaging resolution s can be made constant regardless of the depth d.

  When the calculation of the depth of the dent existing on the surface of the steel sheet is completed, the defect detection processing unit 231 specifies the position of the surface defect (flaw) existing on the surface of the steel sheet based on the obtained information. The position (X coordinate) in the width direction of the steel plate can be specified from the pixel position of the corresponding image. Further, the generation position (Y coordinate) in the longitudinal direction of the eyelid can be calculated from the number of frames and the number of lines of the depth image by setting WP to 0 (the origin of the Y coordinate).

  Here, for example, as shown in FIG. 3, in the depth image calculated from the TDI camera 105 installed by shifting by the offset D in the upstream direction of the steel sheet conveyance, the WP position corresponds to the position earlier by D. Therefore, it is necessary to subtract only D when calculating the Y coordinate.

  The defect detection processing unit 231 outputs information on the surface defect (wrinkle) detected in this way to the display unit 205 or outputs a form.

  In addition, when detecting wrinkles, there may be a case where wrinkles exist in the OV portion of the imaging region in the depth image obtained from each TDI camera 105. Since the wrinkles applied to the overlap margin OV are wrinkles detected also in the depth image obtained from the adjacent TDI camera 105, the defect detection processing unit 231 integrates the wrinkles with respect to such wrinkles. Perform the process.

  Specifically, when the defect detection processing unit 231 detects a wrinkle in the overlap allowance OV, an identifier (overlap flag) indicating that the flaw exists in the overlap allowance OV is added to the information related to the overlap allowance OV. To do. Thereafter, the defect detection processing unit 231 uses the (X, Y) coordinates obtained from the respective depth images for the wrinkles to which the overlap flag has been given at the time of display or form output as one wrinkle. Perform the integration process so that it is output.

Hereinafter, a specific example of the integration processing of the bag will be described with reference to FIG.
In FIG. 10, coordinates (X1, Y1) and coordinates (X2, Y2) are the coordinates of the center of the circumscribed rectangle of の 1 and 疵 2, respectively. In addition, the width and length of the circumscribed rectangle of the collar are m1 and n1 in the collar 1, and m2 and n2 in the collar 2.

  Since the heel shown in FIG. 10A exists in the overlap margin OV, the defect detection processing unit 231 combines the heel 1 and the heel 2 and handles them as one heel. At this time, in the coordinate system on the steel plate where the left edge of the steel plate tip (welding point position) is the origin, the center coordinates of the circumscribed rectangle of 疵 1 are (X1-E, Y1), and the center coordinates of the circumscribed rectangle of 疵 2 Becomes (X2 + WE−OV, Y2 + D). FIG. 10A is a view in which the steel plate is conveyed from the bottom to the top. In the drawing, W is the photographing field width of the camera, and is common to all cameras.

  The defect detection processing unit 231 combines the ridges 1 and 2 shown in FIG. 10A to form a ridge having a width M and a length N as shown in FIG. Here, the defect detection processing unit 231 calculates the width M of the wrinkles after combining using the following expression 112, and calculates the length N of the wrinkles after combining using the following expression 113. Further, the defect detection processing unit 231 calculates the center coordinates (coordinates on the steel plate) of the bounding rectangles of the wrinkles after the combination from (X1−E + M / 2−m1 / 2, Y2 + D + N / 2−n1 / 2).

M = m1 + m2-OV (Formula 112)
N = 1/2 (n1 + n2) + | Y1-Y2 | (Formula 113)

  As described above, the image processing unit 203 according to the present embodiment can grasp the surface shape of the steel sheet 1 accurately and easily by using the shape image (depth image) after the expansion / contraction process. it can.

  Although it is possible to easily grasp the uneven state of the entire surface of the steel plate 1 by using the depth image, for example, it is desired to know only the wrinkles of the uneven shape while ignoring the inclination of the steel plate 1. In some cases. In such a case, the defect detection processing unit 231 can detect the irregular-shaped wrinkles generated on the surface of the steel sheet 1 based on the depth image output from the phase continuation processing unit 229 as follows. Is possible.

  Specifically, the defect detection processing unit 231 first extracts the distribution of the phase shift φ along the vertical direction at each position in the horizontal direction from the shape image after the expansion / contraction process. Then, the defect detection processing unit 231 performs, for example, least square approximation on the extracted phase shift φ distribution, and calculates an approximate curve for the phase shift φ distribution along the vertical direction. Thereafter, the defect detection processing unit 231 subtracts the approximate curve from the extracted distribution curve of the phase shift φ along the vertical direction. The subtraction result obtained in this way includes only information related to defects.

  The defect detection processing unit 231 performs such processing at all lateral positions. By expressing the result thus obtained as an image, the defect detection processing unit 231 can obtain a defect image in which only the irregular shape wrinkles are extracted. Thereafter, the defect detection processing unit 231 performs processing such as binarization and labeling on the obtained defect image to detect defects.

  Here, the case where the subtraction process is performed on the distribution of the phase shift φ along the vertical direction at each position in the horizontal direction when obtaining the defect image by the defect detection processing unit 231 has been described. In addition, a subtraction process may be performed on the phase shift φ along the horizontal direction at each position in the vertical direction.

  As described above, the arithmetic processing device 200 according to the present embodiment is based on the shape image processing frame and the luminance image processing frame having a constant vertical resolution according to the conveyance speed of the steel plate, in which the welding points (starting points of the steel plate coils) are aligned. Thus, it is possible to perform the wrinkle detection process. For this reason, it is possible to accurately calculate defect detection position information indicating a position where a defect is detected.

  Heretofore, an example of the function of the arithmetic processing apparatus 200 according to the present embodiment has been shown. Each component described above may be configured using a general-purpose member or circuit, or may be configured by hardware specialized for the function of each component. In addition, the CPU or the like may perform all functions of each component. Therefore, it is possible to appropriately change the configuration to be used according to the technical level at the time of carrying out the present embodiment.

  A computer program for realizing each function of the arithmetic processing apparatus according to the present embodiment as described above can be produced and installed in a personal computer or the like. In addition, a computer-readable recording medium storing such a computer program can be provided. The recording medium is, for example, a magnetic disk, an optical disk, a magneto-optical disk, a flash memory, or the like. Further, the above computer program may be distributed via a network, for example, without using a recording medium.

<Flow of surface defect inspection method>
Next, the flow of the surface defect inspection method according to the present embodiment will be briefly described with reference to FIG. FIG. 11 is a flowchart showing the flow of the surface defect inspection method according to the present embodiment.

  First, the surface defect inspection apparatus 10 captures the steel sheet 1 irradiated with the linear laser light with the TDI camera 105 of the steel sheet imaging apparatus 100 and generates a fringe image (step S101). Thereafter, the image processing unit 203 (more specifically, the image generation unit 209) of the arithmetic processing device 200 calculates the amplitude and phase based on the signal of the fringe image (step S103). Here, it is preferable that the fringe defect determination unit 225 of the image processing unit 203 determines the presence or absence of the fringe defect using the obtained amplitude image (step S105).

  Thereafter, the phase continuation processing unit 227 performs phase continuation processing on the phase image generated by the phase calculation unit 221 (step S107), and generates a shape image representing the uneven state of the surface of the steel sheet.

  Subsequently, the image expansion / contraction processing unit 229 performs image expansion / contraction processing on the shape image and the luminance image (amplitude image) to expand and contract the shape image and the luminance image according to the conveyance speed of the steel plate in the conveyance line (step S109). Thereby, the arithmetic processing apparatus 200 which concerns on this embodiment can obtain the depth image with which the vertical resolution of an image is constant irrespective of the conveyance speed of a steel plate.

  Next, the defect detection processing unit 231 uses the depth image generated by the image expansion / contraction processing unit 229 to inspect whether or not a defect exists on the surface of the steel plate (step S111). In addition, when a surface defect is detected in the overlapping range of the imaging regions, the defect detection processing unit 231 performs the flaw integration process as described above on the surface defect.

  By processing in this flow, it becomes possible to accurately and easily grasp the surface shape of the steel sheet, and it is possible to detect minute irregularities and pattern-like wrinkles simultaneously with high accuracy and high speed. it can.

(About hardware configuration)
Next, the hardware configuration of the arithmetic processing apparatus 200 according to the embodiment of the present invention will be described in detail with reference to FIG. FIG. 12 is a block diagram for explaining a hardware configuration of the arithmetic processing device 200 according to the embodiment of the present invention.

  The arithmetic processing device 200 mainly includes a CPU 901, a ROM 903, and a RAM 905. The arithmetic processing device 200 further includes a bus 907, an input device 909, an output device 911, a storage device 913, a drive 915, a connection port 917, and a communication device 919.

  The CPU 901 functions as an arithmetic device and a control device, and controls all or a part of the operation in the arithmetic processing device 200 according to various programs recorded in the ROM 903, the RAM 905, the storage device 913, or the removable recording medium 921. The ROM 903 stores programs used by the CPU 901, calculation parameters, and the like. The RAM 905 primarily stores programs used by the CPU 901, parameters that change as appropriate during execution of the programs, and the like. These are connected to each other by a bus 907 constituted by an internal bus such as a CPU bus.

  The bus 907 is connected to an external bus such as a PCI (Peripheral Component Interconnect / Interface) bus via a bridge.

  The input device 909 is an operation unit operated by the user, such as a mouse, a keyboard, a touch panel, a button, a switch, and a lever. The input device 909 may be, for example, remote control means (so-called remote control) using infrared rays or other radio waves, or may be an external connection device 923 such as a PDA corresponding to the operation of the arithmetic processing device 200. May be. Furthermore, the input device 909 includes, for example, an input control circuit that generates an input signal based on information input by a user using the operation unit and outputs the input signal to the CPU 901. The user of the arithmetic processing device 200 can input various data and instruct processing operations to the arithmetic processing device 200 by operating the input device 909.

  The output device 911 is configured by a device that can notify the user of the acquired information visually or audibly. Examples of such devices include CRT display devices, liquid crystal display devices, plasma display devices, EL display devices and display devices such as lamps, audio output devices such as speakers and headphones, printer devices, mobile phones, and facsimiles. The output device 911 outputs results obtained by various processes performed by the arithmetic processing device 200, for example. Specifically, the display device displays results obtained by various processes performed by the arithmetic processing device 200 as text or images. On the other hand, the audio output device converts an audio signal composed of reproduced audio data, acoustic data, and the like into an analog signal and outputs the analog signal.

  The storage device 913 is a data storage device configured as an example of a storage unit of the arithmetic processing device 200. The storage device 913 includes, for example, a magnetic storage device such as an HDD (Hard Disk Drive), a semiconductor storage device, an optical storage device, or a magneto-optical storage device. The storage device 913 stores programs executed by the CPU 901, various data, various data acquired from the outside, and the like.

  The drive 915 is a recording medium reader / writer, and is built in or externally attached to the arithmetic processing unit 200. The drive 915 reads information recorded on a removable recording medium 921 such as a mounted magnetic disk, optical disk, magneto-optical disk, or semiconductor memory, and outputs the information to the RAM 905. The drive 915 can also write a record on a removable recording medium 921 such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory. The removable recording medium 921 is, for example, a CD medium, a DVD medium, a Blu-ray medium, or the like. The removable recording medium 921 may be a CompactFlash (registered trademark) (CompactFlash: CF), a flash memory, an SD memory card (Secure Digital memory card), or the like. Further, the removable recording medium 921 may be, for example, an IC card (Integrated Circuit card) on which a non-contact IC chip is mounted, an electronic device, or the like.

  The connection port 917 is a port for directly connecting a device to the arithmetic processing device 200. Examples of the connection port 917 include a USB (Universal Serial Bus) port, an IEEE 1394 port, a SCSI (Small Computer System Interface) port, and an RS-232C port. By connecting the external connection device 923 to the connection port 917, the arithmetic processing apparatus 200 acquires various data directly from the external connection device 923 or provides various data to the external connection device 923.

  The communication device 919 is a communication interface configured with, for example, a communication device for connecting to the communication network 925. The communication device 919 is, for example, a communication card for wired or wireless LAN (Local Area Network), Bluetooth (registered trademark), or WUSB (Wireless USB). The communication device 919 may be a router for optical communication, a router for ADSL (Asymmetric Digital Subscriber Line), or a modem for various communication. The communication device 919 can transmit and receive signals and the like according to a predetermined protocol such as TCP / IP, for example, with the Internet and other communication devices. The communication network 925 connected to the communication device 919 is configured by a wired or wireless network, and may be, for example, the Internet, a home LAN, infrared communication, radio wave communication, satellite communication, or the like. .

  Heretofore, an example of the hardware configuration capable of realizing the function of the arithmetic processing device 200 according to the embodiment of the present invention has been shown. Each component described above may be configured using a general-purpose member, or may be configured by hardware specialized for the function of each component. Therefore, it is possible to change the hardware configuration to be used as appropriate according to the technical level at the time of carrying out this embodiment.

  In this way, by using the surface defect detection method according to the present embodiment, it is possible to simultaneously generate minute uneven ridges and pattern ridges with high accuracy and high speed over the entire length and width in real time during the conveyance of the steel sheet. It is possible to detect the wrinkle occurrence position accurately. This makes it possible to guarantee the quality of the steel sheet and can greatly contribute to the productivity and yield improvement of the steel sheet manufacturing.

  The preferred embodiments of the present invention have been described in detail above with reference to the accompanying drawings, but the present invention is not limited to such examples. It is obvious that a person having ordinary knowledge in the technical field to which the present invention pertains can come up with various changes or modifications within the scope of the technical idea described in the claims. Of course, it is understood that these also belong to the technical scope of the present invention.

  For example, in the above-described embodiment, the arithmetic processing device 200 collectively controls a plurality of laser irradiation devices and delay integration type imaging devices provided in the steel plate imaging device 100, and is obtained from a plurality of delay integration type imaging devices. Although the case where images are processed together has been described, the present invention is not limited to such an example. For example, the arithmetic processing device 200 may be provided for each imaging unit including one laser irradiation device and one delay integration imaging device.

DESCRIPTION OF SYMBOLS 1 Steel plate 2 Roller 10 Surface defect inspection apparatus 100 Steel plate imaging device 101 Laser light source 103 Rod lens 105 Delay integration type imaging device (TDI camera)
DESCRIPTION OF SYMBOLS 200 Arithmetic processor 201 Timing signal generation part 203 Image processing part 205 Display part 207 Storage part 209 Image generation part 211 A / D conversion part 213 Prefilter part 215 Orthogonal sine wave generation part 217,219 Low-pass filter part 221 Phase calculation part 223 Amplitude calculation unit 225 Stripe defect determination unit 227 Phase continuation processing unit 229 Image expansion / contraction processing unit 231 Defect detection processing unit

Claims (6)

A steel plate imaging device that irradiates a steel sheet conveyed on a conveying line with linear laser light, images the linear laser light reflected by the steel plate, and generates an image of the laser light;
An arithmetic processing device that performs image processing on a captured image of the steel plate generated by the steel plate imaging device and determines whether or not a defect exists on the surface of the steel plate;
With
The steel plate imaging device
A plurality of laser irradiation devices for irradiating the steel sheet so as not to superimpose modulated linear laser beams on each other;
When the linear laser light is irradiated from each of the plurality of laser irradiation devices, each of the linear laser light reflected by the steel plate is imaged, and the light of the steel plate that is an image captured by the laser light A plurality of delay integration type imaging devices that respectively generate cut images;
Have
The arithmetic processing unit includes:
A fringe image is formed from each of the light cut images generated by the steel plate imaging device, and using the fringe image, a plurality of shape images representing the concavo-convex state of the surface of the steel plate, and a rough surface on the surface of the steel plate A plurality of luminance images representing the difference in degrees, respectively,
Using the shape image, a defect detection processing unit that detects surface defects present on the surface of the steel plate,
I have a,
Of the plurality of delay integration type imaging devices that capture each of the linear laser beams, the imaging regions of the two imaging devices adjacent to each other in the width direction of the steel plate overlap each other in the transport direction of the steel plate. And overlapping with each other in the width direction of the steel sheet,
The range in the plate width direction of the steel plate of each imaging region of the delay integration type imaging device is a constant width from both ends in the plate width direction in the region irradiated with the linear laser beam. A surface defect inspection apparatus characterized by being a portion located substantially in the center excluding the portion . The arithmetic processing unit includes:
The shape image and the luminance image generated by the image generation unit are expanded or contracted according to the conveyance speed of the steel plate in the conveyance line, and an image with a constant conveyance direction resolution is generated regardless of the conveyance speed of the steel plate. The surface defect inspection apparatus according to claim 1, further comprising an image expansion / contraction processing unit. The defect detection processing unit
When the surface defect is detected in a region corresponding to the overlapped portion in the width direction of the steel plate among the shape images obtained from the delay integration imaging devices adjacent to each other, the delay integration imaging adjacent to each other. The surface defect inspection apparatus according to claim 2 , wherein an integration process for integrating surface defects detected from shape images obtained from the respective apparatuses as one surface defect is performed. The image expansion / contraction processing unit
Generating an image in which the resolution in the conveyance direction of the shape image and the luminance image is constant, using information about the time when the starting point of the steel plate is imaged and information about the conveyance speed for conveying the steel plate The surface defect inspection apparatus according to claim 3 , wherein: A laser beam irradiation step for irradiating a plurality of modulated linear laser beams so as not to overlap each other on a steel plate transported on a transport line;
An optical cut image generation step of imaging the linear laser light reflected by the steel plate with each of a plurality of delay integration imaging devices to generate an optical cut image of the steel plate that is an image of the laser light, and
A stripe image is formed from each of the light cut images generated in the light cut image generation step, and a plurality of shape images representing the uneven state of the surface of the steel sheet using the stripe image, and the surface of the steel plate A plurality of luminance images each representing a difference in roughness at the image generation step,
Using the shape image, a defect detection step for detecting surface defects present on the surface of the steel sheet;
Only including,
Of the plurality of delay integration type imaging devices that capture each of the linear laser beams, the imaging regions of the two imaging devices adjacent to each other in the width direction of the steel plate overlap each other in the transport direction of the steel plate. And overlapping with each other in the width direction of the steel sheet,
The range in the plate width direction of the steel plate of each imaging region of the delay integration type imaging device is a constant width from both ends in the plate width direction in the region irradiated with the linear laser beam. A method for inspecting a surface defect, wherein the surface defect inspection method is a portion located substantially in the center excluding the portion . A plurality of modulated linear laser beams are irradiated onto a steel plate transported on a transport line so as not to overlap each other, and the linear laser beams reflected by the steel plate are irradiated by a plurality of delay integration imaging devices. Each of the plurality of delay-integration imaging devices that captures each of the linear laser beams to generate a light-cut image of the steel plate that is a captured image of the laser light. The imaging regions of the two imaging devices adjacent to each other in the width direction of the steel sheet do not overlap with each other in the conveyance direction of the steel sheet, and overlap with each other in the width direction of the steel sheet. The range in the plate width direction of the steel plate of each of the imaging regions is substantially the same as the region where the linear laser light is irradiated, except for portions of a certain width from both ends in the plate width direction. Located in the center The steel sheet imaging apparatus capable of communicating with a computer is a partial,
A fringe image is formed from each of the light cut images generated by the plurality of delay integration imaging devices, and a plurality of shape images representing the unevenness state of the surface of the steel sheet using the fringe image, and the steel sheet An image generation function for generating a plurality of luminance images each representing a difference in roughness on the surface;
Using the shape image, a defect detection function for detecting surface defects present on the surface of the steel plate,
A program to realize

Images