JP4873100B2 - Surface inspection device - Google Patents

Description

  The present invention relates to an apparatus for inspecting a surface defect of an inspection object, and more particularly to a surface inspection apparatus suitable for inspecting a minute point defect generated on a surface of a steel plate or the like.

  Conventionally, ring illumination has been widely used as illumination used when inspecting or observing the surface of an inspection object in order to uniformly irradiate the surface of the inspection object. However, when the object to be inspected is irradiated with ring illumination, especially when the surface of the object is highly specular, the light emission part of the ring illumination is reflected on the surface of the object, resulting in uneven brightness and small dots. There was a problem that it became difficult to see the defect.

  As a countermeasure, for example, Patent Document 1 discloses a technique of inserting a light diffusing plate between a light irradiation unit and an inspection object in ring illumination. Patent Document 2 discloses a technique in which light emitted from a light emitting portion is arranged at a predetermined angle outward in the circumference, and a diffuse reflection hood having an inner surface as a diffuse reflection surface is provided on the ring illumination end surface portion. It is disclosed.

  Specifically, there is a ring illumination device 100 as shown in FIG. In this ring illumination device 100, in order to avoid halation and suppress the occurrence of uneven brightness, a light diffusion plate 102 is disposed in front of the light emitting unit 101 to diffuse the light emitted from the light emitting unit 101. The surface of the inspection object 103 was irradiated. The light emitted from the light emitting unit 101 is diffused by the diffusion plate 102 and irradiated onto the inspection object 103, and the reflected light 101 a, 101 b, 101 c,... Is imaged on the area sensor camera 105 through the lens 104. It has become.

JP 2007-57421 A Japanese Patent Application Laid-Open No. 6-235821 Japanese Patent No. 3585214 Japanese Patent Laid-Open No. 2007-3243

  However, when the conventional ring illumination device 100 including the light diffusing plate 102 shown in FIG. 18 is used, a brightness pattern of bright and dark spots is generated on the surface having a fine concavo-convex structure such as metal. As a result, the SN ratio of defect detection is reduced. This is because the reflected light includes regular reflected light 101a, forward diffuse reflected light 101b, backward diffuse reflected light 101c, and the like. FIG. 19 is a diagram schematically showing a light reflection intensity distribution 203 when light 201 is incident on a rough surface 103 such as a steel plate. The light reflection intensity becomes maximum in the regular reflection direction 202. As can be seen from FIG. 19, since the reflection intensity of the regular reflection light 101a and the reflection light in the vicinity of the regular reflection greatly fluctuates due to slight angle fluctuations, the inspection object 103 has a bright and dark surface on the surface having a fine uneven structure such as metal. A luminance pattern of points is generated, which becomes ground noise and causes a decrease in the SN ratio of defect detection. Accordingly, part of the diffused and irradiated light enters the area sensor camera 105 as a specularly reflected light component, and the formation noise intensity due to the fine unevenness on the surface of the steel sheet is increased, and a minute amount is included in the formation noise. The point-like defects are buried, and it becomes difficult to separate the ground noise from the minute point-like defects.

  That is, although the techniques disclosed in Patent Document 1 and Patent Document 2 have the effect of suppressing halation, there are no minute point-like defects depending on the surface state of the inspection object or the material of the inspection object. However, there is a case where false detection that there is a minute point-like defect occurs due to formation noise.

  The present invention has been made in view of the above, and an object of the present invention is to provide a surface inspection apparatus capable of inspecting minute point defects with high accuracy.

  In order to solve the above-described problems and achieve the object, a surface inspection apparatus according to the present invention includes a ring-shaped light emitting portion and a concentric circle shape between the light emitting portion and the inspection object. And a ring illumination device having a light shielding plate having an optical opening having a diameter smaller than the inner diameter of the light emitting portion, and a center line of the opening of the light shielding plate, and through the opening An imaging unit that images the surface of the inspection object, and the imaging region on the surface of the inspection object that is captured by the imaging unit includes the light shielding plate of the light emitted from the light emitting unit. Only the light diffracted at the edge of the opening is irradiated, and the distance between the light emitting portion and the surface of the inspection object is such that the average luminance level in the imaging region on the object surface is a predetermined level or more, and The brightness level difference in the imaging area is set to be within a predetermined range. And features.

  Further, in the surface inspection apparatus according to the present invention, in the above invention, the inspection object is a strip-shaped material transported along a longitudinal direction, and a continuous generation length of a continuity defect set unique to the strip-shaped material. While the section having a certain length which is equal to or less than the minimum length is being conveyed, the ring illumination device and the imaging unit are reciprocated at least once in the entire width of the band-shaped material, and the section having the certain length is conveyed. A traverse portion that repeats the same operation each time is provided, and a continuity defect generated on the surface of the belt-shaped material is inspected.

  In the surface inspection apparatus according to the present invention, in the above invention, a reciprocation between the transport distance detection unit that detects the transport distance of the belt-shaped material and the traverse unit based on the transport distance information detected by the transport distance detection unit. And an imaging timing control unit that controls an imaging timing so that the imaging unit images the belt-like material substantially over the entire width.

  Further, the surface inspection apparatus according to the present invention is the above-described invention, wherein the surface inspection apparatus is adjacent to each other so as to be parallel to each other along the longitudinal direction of the strip-shaped material, and has a length substantially the same as the visual field in the width direction of the imaging unit. When a plurality of elongated track areas having a width are set, the imaging unit images the adjacent partial areas belonging to the respective track areas or the partial areas spaced apart from each other in the longitudinal direction. To do.

  In the surface inspection apparatus according to the present invention, in the above invention, the image processing unit that extracts the defect harmfulness in each image captured by the imaging unit as numerical data, and the each extracted by the image processing unit. And a defect distribution calculating unit that calculates a two-dimensional defect occurrence state on the surface of the belt-shaped material based on the numerical defect data of the image and the imaging position thereof.

  The surface inspection apparatus according to the present invention may further include a defect map display unit that displays the defect distribution status calculated by the defect distribution calculation unit on the two-dimensional development view of the surface of the belt-shaped material in the above invention. Features.

  In the surface inspection apparatus according to the present invention as set forth in the invention described above, the defect map display unit divides the surface of the belt-like material into rectangular meshes, and changes the display color or display mark of the defect harmfulness of each mesh. It is characterized by displaying.

  In the surface inspection apparatus according to the present invention as set forth in the invention described above, the inspection object is a steel plate, and inspects minute spot-like defects in an imaging region on the steel plate.

  In the surface inspection apparatus according to the present invention as set forth in the invention described above, the strip material is a pickled steel plate, and the continuity defect is a scale residue.

  According to the present invention, an optical light that is concentric with the light emitting part and smaller in diameter than the inner diameter of the light emitting part between the ring-shaped light emitting part and the light emitting part and the detection target. A ring illuminating device having a light shielding plate having an opening, and an imaging unit that is disposed on a center line of the opening of the light shielding plate and images the surface of the inspection object through the opening. The imaging region on the surface of the inspection object to be imaged by the imaging unit is irradiated only with light diffracted at the edge of the opening of the light shielding plate among the light emitted from the light emitting unit, and the light emission The distance between the portion and the surface of the inspection object is such that the average luminance level in the imaging area on the surface of the inspection object is equal to or higher than a predetermined level, and the luminance level difference in the inspection imaging area is within a predetermined range. Therefore, the brightness in the imaging area is flattened and the ground noise is Is suppressed, a fine dot-like defects can be accurately inspected.

1 is a schematic sectional view showing a schematic configuration of a surface inspection apparatus according to Embodiment 1 of the present invention. FIG. 2 is an explanatory diagram for explaining the distribution of illumination light on the surface of the steel sheet, which is an inspection object, and the imaging region by the imaging unit. FIG. 3 is an explanatory diagram for explaining the geometric arrangement of the ring illumination device. FIG. 4 is an explanatory diagram for explaining the distance between the surface of the steel plate and the light emitting portion. FIG. 5 is a diagram illustrating a change in luminance distribution in the imaging region depending on the distance between the surface of the steel plate and the light emitting portion. FIG. 6 is a schematic diagram showing a schematic configuration of a surface inspection apparatus according to Embodiment 2 of the present invention. FIG. 7 is a block diagram showing an example of a surface inspection apparatus according to Embodiment 2 of the present invention. FIG. 8: is sectional drawing which shows the relationship between the traverse part and strip steel plate in the surface inspection apparatus concerning Embodiment 2 of this invention, and shows the state which cut | disconnected the strip steel plate in the traverse direction. FIG. 9 is a plan view showing an example of the form of occurrence of continuity defects that occur on the surface of the strip steel plate. FIG. 10 is a plan view for explaining a situation in which the inspection frequency of the surface of the strip steel plate increases in a region where the conveyance speed has changed to a low speed. FIG. 11 shows a case where the movement of the imaging unit in the width direction is completed while the strip steel plate moves by a predetermined distance ΔL, and illustrates that the inspection frequency of the strip steel plate surface does not increase in a region where the conveyance speed has changed to a low speed. It is a top view of a strip steel plate. FIG. 12 is a plan view of a strip-shaped steel plate showing an imaging region by enlarging the broken line portion and the vicinity thereof in FIG. 11. FIG. 13 is a plan view showing an embodiment in which imaging is performed while moving the imaging unit in the width direction while the belt-shaped steel plate is moved by a predetermined distance ΔL, and returning without imaging in the width direction. FIG. 14 illustrates an imaging area in which each track area is imaged in an embodiment in which imaging is performed while moving the imaging unit in the width direction while the steel plate is moved by a predetermined distance ΔL, and returning without imaging in the width direction. It is a top view which shows the example which does not continue mutually. FIG. 15 is a plan view showing an example in which the occurrence distribution state of defects calculated by the surface inspection apparatus according to the second embodiment is displayed on a two-dimensional map. FIG. 16 is an example of the traverse portion used in the embodiment of the present invention, and is a cross-sectional view showing a state in which the strip steel plate is cut in the traverse direction of an example in which two detection heads are installed at a predetermined distance in the width direction. . FIG. 17 is a view showing an example in which the density of the mesh of the defect map is changed in three levels according to the size of the defect parameter calculated according to the embodiment of the present invention. FIG. 18 is an explanatory diagram in the case of using a ring illumination provided with a conventional light diffusion plate. FIG. 19 is an explanatory diagram showing a light reflection intensity distribution pattern on a rough surface such as a steel plate, and an explanatory diagram showing that the reflection intensity greatly fluctuates due to slight angle fluctuations of specular reflection light and reflected light near the specular reflection. It is.

  Hereinafter, a surface inspection apparatus according to an embodiment of the present invention will be described with reference to the drawings.

(Embodiment 1)
In the first embodiment, a minute dot defect on the surface of an inspection object can be inspected with high accuracy by using a generation mechanism of diffracted light. FIG. 1 is a schematic cross-sectional view showing a schematic configuration of a surface inspection apparatus 1 according to Embodiment 1 of the present invention. FIG. 2 is an explanatory diagram for explaining the distribution of illumination light on the surface of the steel plate 2 that is the detection target and the imaging region A by the imaging unit 4. Furthermore, FIG. 3 is explanatory drawing explaining the geometric arrangement | positioning of a ring illuminating device.

  As shown in FIG. 1, the surface inspection apparatus 1 includes a ring illumination device 3, an imaging unit 4, and an image processing device 5. The image processing device 5 is connected to the imaging unit 4. The ring illumination device 3 includes a ring-shaped light emitting portion 3A and a light shielding plate 3B. The imaging unit 4 includes an area sensor camera 4A and a lens 4B. The light shielding plate 3B is disposed between the light emitting portion 3A and the steel plate 2. The light emitting unit 3A and the light shielding plate 3B may be configured as a ring illumination device 3 integrally formed. Furthermore, the ring illumination device 3 and the imaging unit 4 may be integrally configured.

  The imaging unit 4 is disposed above the light emitting unit 3A so that the optical axis C4 coincides with the central axis of the opening of the light emitting unit 3A. Further, the area sensor camera 4A of the imaging unit 4 is set so that the reflected light of the diffracted light reflected by the surface of the steel plate 2 is imaged by the lens 4B.

  The light shielding plate 3B has a circular optical opening having a diameter smaller than the inner diameter of the opening of the light emitting portion 3A. The opening of the light shielding plate 3B and the ring-shaped light emitting portion 3A are concentric and have the same central axis C. Further, the center axis C coincides with the optical axis C4 of the imaging unit 4. The material of the light shielding plate 3B is not limited as long as it is optically opaque, but for example, an aluminum material whose surface is black anodized can be used.

  As shown in FIG. 3, the illumination light emitted from the light emitting portion 3A is irradiated toward the surface of the steel plate 2 with a predetermined divergence angle, but a part of the illumination light is blocked by the light shielding plate 3B. As the installation condition of the light shielding plate 3 </ b> B, the spread light emitted from the light emitting unit 3 </ b> A is not directly irradiated into the imaging region A on the surface of the steel plate 2. Then, the imaging region A is irradiated with light diffracted at the inner edge of the light shielding plate 3B.

That is, in FIG. 3, it arrange | positions so that following Formula may be satisfy | filled.
H · tanθ-R> r

  Here, R is the distance from the central axis C of the light emitting part 3A to the light emitting position of the light emitting part 3A. R is the distance from the central axis C of the light emitting portion 3A to the edge of the imaging region A. H represents the distance from the light emission position of the light emission part 3A to the surface of the steel plate 2. θ is an angle between the optical axis direction of the direct light blocked by the light shielding plate 3B and the central axis C.

  Next, the reason for using the ring illumination and the light shielding plate 3B will be described. The reason why the ring-shaped illumination is used as the illumination is to cancel shadows caused by fine irregularities on the surface of the steel plate 2 by irradiating the surface of the steel plate 2 symmetrically from all directions. In particular, when a minute defect is inspected on the surface of an inspection object having a poor surface property such as a pickled steel sheet, there is an effect of suppressing formation noise. Further, as shown in FIGS. 1 and 2, the light shielding plate 3 </ b> B is provided so that the specularly reflected light component included in a part of the direct light emitted from the ring illumination device 3 is captured on the imaging region A on the surface of the steel plate 2. This is to prevent it from entering. That is, the direct light irradiation region B (hatched portion in FIG. 2) that is irradiated on the surface of the steel plate 2 without being shielded by the light shielding plate 3B includes a specular reflection component, but the imaging region A is compared with the direct light irradiation region B. By setting the inner side to be inside, only the diffracted light from the light shielding plate 3B is irradiated into the imaging region A.

  As a result, when the ring illumination device 3 provided with the light shielding plate 3B is used, the imaging unit 4 images the imaging region A, and the imaging unit 4 receives the diffracted light generated in the portion blocked by the light shielding plate 3B. Only the reflected light from the light beam is reflected, and the regular reflected light is not imaged on the imaging unit 4. Therefore, even if it is a rough surface with poor surface properties such as the steel plate 2, formation noise can be suppressed and minute point defects can be detected with a high SN ratio.

  Here, as shown in FIGS. 4 and 5, the luminance distribution in the imaging region A centering on the optical axis C4 of the imaging unit 4 varies depending on the distance H from the light emitting unit 3A to the surface of the inspection object 2. To do. As shown in FIG. 5A, when the distance H is smaller than the appropriate predetermined value Dth, direct light from the light emitting portion 3A is reflected around the imaging region A, and the luminance of this peripheral portion becomes extremely high. . On the other hand, the brightness near the optical axis C4 is low. As a result, the luminance difference in the imaging area A becomes large, the luminance becomes non-uniform, and a highly accurate defect inspection cannot be performed.

  On the other hand, as shown in FIG. 5C, when the distance H is larger than the appropriate predetermined value Dth, the luminance in the imaging region A becomes uniform, but the intensity of the diffracted light into the imaging region A becomes weak. The average luminance level in the imaging area A is lowered. As a result, particularly when inspecting the surface of a high-gloss steel sheet or the like, illumination with extremely high brightness is required, and both the introduction cost and running cost of the inspection apparatus are expensive.

  Therefore, as shown in FIG. 5B, by setting the distance H to an appropriate predetermined value Dth, a flat luminance distribution can be obtained while maintaining a high luminance, and a highly accurate surface inspection is performed. Can do. Here, since the appropriate predetermined value Dth varies depending on the size and emission directivity of the ring illumination device 3 and the light reflection characteristics of the surface of the inspection object, the predetermined value Dth is determined in advance according to the inspection object. It is preferable to keep it. The predetermined value Dth is set as a value when the average luminance level in the imaging area A is equal to or higher than the predetermined level and the luminance level difference in the imaging area A is within a predetermined range, for example, within ± 10%. Is preferred.

  Note that the processing contents in the image processing apparatus 5 can use known ones. For example, after pre-processing such as shading correction, pixels that exceed a predetermined threshold level are extracted by binarization or multi-value processing, and after image concatenation processing, isolated point removal, and labeling processing, image feature amounts are calculated. The defect may be extracted and determined.

  Further, when the strip-shaped steel plate 2 conveyed in line is used as an inspection object, it is preferable to use a xenon / strobe light source or the like as the light source of the light emitting portion 3A in order to capture the strip-shaped steel plate 2 without blur. Further, as the light emitting portion 3A, a configuration in which LEDs are arranged in a donut shape, a configuration in which an optical fiber bundle connected to an illumination light source is distributed in a donut shape, or the like can be used.

  In Embodiment 1, only the light diffracted at the opening edge of the light shielding plate 3B among the light emitted from the light emitting portion 3A is irradiated onto the imaging region A, and the light emitting portion 3A and the surface of the inspection object Is a luminance distribution such that the average luminance level in the imaging region A is equal to or greater than the predetermined luminance level, and the luminance level difference in the imaging region A is within a predetermined range, for example, within ± 10%. Is flattened, the brightness non-uniformity is eliminated in the imaging region A, and a minute dot-like defect can be inspected with high accuracy.

(Embodiment 2)
In this embodiment, by using the ring illumination device 3 and the imaging unit 4 described above, a minute defect that occurs in a conveying direction on the surface of a strip-shaped steel plate (band-shaped steel plate) 2 that is line-transferred is simplified. A surface inspection apparatus to be inspected will be described.

By the way, as an example of inspecting the continuity defect generated on the surface of the strip-shaped steel plate 2a conveyed by the conveyance line, for example, there is a step of inspecting a scale residual defect generated on the surface of the pickled steel plate. In the line where pickling is performed, scales (FeO, Fe ₃ O ₄ , Fe ₂ O _3, etc.) on the surface are removed by passing the strip steel plate 2 a through a liquid tank containing a strong acid such as sulfuric acid. In this line, when the conveying speed of the steel sheet is too high, or when the surface of the steel sheet is so-called rough, the scale may rarely remain on the rough surface.

  Such scale remaining defects are formed by densely forming minute spot-like defects having a diameter of about 0.05 to 0.3 mm, and are difficult to visually recognize during the conveyance of the strip-shaped steel sheet 2a. For this reason, automatically inspecting the occurrence of the remaining scale with a surface inspection device is extremely important for ensuring the surface quality of the pickled steel sheet.

  As a feature of the generation form of the remaining scale defect, it is generated continuously at a position in the same width direction over a predetermined length along the conveying direction of the steel plate. Further, not only the remaining scale defects but also the continuous defects generated on the surface of the belt-like material often have a empirically known range of continuous generation length inherent to the production line. The second embodiment relates to an inspection apparatus suitable for such continuity defect inspection.

  As a surface inspection apparatus in a conventional pickled steel plate production line, linear illumination is arranged along the width direction of the steel plate (direction orthogonal to the conveying direction), and along the width direction of the steel plate at a position facing the steel plate. An apparatus in which several to a dozen or so line sensor cameras are installed over the entire width of a steel plate is well known so that the entire width can be inspected. In this surface inspection apparatus, camera output signals obtained by the respective line sensor cameras are connected in the conveying direction of the steel sheet, and surface defect portions are extracted by image processing.

  However, when inspecting the above minute spot defects generated on the surface of pickled steel sheet transported at high speed, the surface inspection apparatus using the line sensor camera has insufficient camera resolution in the transport direction, and is sufficient. There is a problem that a satisfactory defect inspection performance cannot be obtained. For example, in order to detect a point defect having a diameter of 0.05 mm, the camera resolution needs to be approximately 0.025 mm or less. However, in a line sensor camera with a video rate of 40 MHz using 4096 elements, which is normally used, the camera resolution in the transport direction on the pickling line with a transport speed of 2 m / sec is only about 0.2 mm, and the required resolution is about 1 There are not enough digits.

  On the other hand, as a means for improving the camera resolution in the transport direction, an apparatus that captures the steel plate surface as a still image using a high-definition area sensor camera using strobe illumination or the like can be considered. However, when the apparatus using the strobe illumination is applied to the inspection of the belt-shaped material to be transported, there is a problem of how to control the imaging timing in the transport direction. For example, Patent Document 3 discloses a surface inspection apparatus that transports an inspection target at a constant speed and performs camera imaging at equal time intervals so that the imaging field of view of the camera is constant in the transport direction of the surface of the inspection target. It is disclosed. In Patent Document 4, camera imaging is performed at equal time intervals, the conveyance distance of the inspection object is constantly measured, the conveyance distance moved between the imaging timings of each image is obtained, and this conveyance is performed in each image. There is disclosed a surface inspection apparatus in which only a part corresponding to a distance is an effective area and image processing is performed only on this area.

  The inspection apparatus disclosed in the above-mentioned Patent Document 3 cannot be applied to a line in which the conveyance speed constantly changes, such as a pickled steel plate production line. Further, in the apparatus disclosed in Patent Document 4, since the image processing range is changed in each image according to the conveyance speed, the image processing becomes complicated and the processing time becomes long, and real-time processing is performed on the high-speed conveyance line. Is difficult. Further, problems in the case of inspecting a scale remaining defect with a conventional inspection apparatus including the apparatuses disclosed in Patent Documents 3 and 4 include those described in the following (1) and (2).

(1) Since it is necessary to set the camera resolution in the width direction of the band-shaped material to half or less of the fine point defect size, a large number of cameras must be installed in the width direction, and the inspection apparatus becomes very expensive. Also, a great deal of labor is required for camera adjustment and maintenance.

(2) In the case of a scale remaining defect formed by a cluster of minute point-like defects, if the amount of features such as the size, brightness, and shape of each point-like defect are calculated and processed one by one, a huge load is imposed on image processing. And real-time processing becomes difficult.

  Therefore, in the second embodiment, minute spot-like defects can be inspected with high precision, and even an inspection target such as a scale remaining defect formed by clustering minute spot-like defects is continuously generated in the transport direction. By utilizing this feature, a surface inspection apparatus capable of processing in real time with a simple configuration is obtained.

  FIG. 6 is a schematic diagram showing a schematic configuration of a surface inspection apparatus according to Embodiment 2 of the present invention. FIG. 7 is a block diagram showing an example of this surface inspection apparatus. This Embodiment 2 demonstrates as a surface inspection apparatus with respect to the strip | belt-shaped steel plate 2a conveyed with the production line of a pickled steel plate.

[Schematic configuration of surface inspection equipment]
As shown in FIG. 6, the surface inspection apparatus 11 images the surface of the strip-shaped steel plate 2a to be conveyed by the ring illumination device 3 and the imaging unit 4 equipped with the light shielding plate described in the first embodiment. The ring illumination device 3 and the imaging unit 4 are fixedly arranged in the detection head 1a. Further, the surface inspection apparatus 11 includes a traverse unit 6 that moves the detection head 1a in the width direction of the strip steel plate 2a, an image processing unit 7, a transport distance detection unit 8, an imaging timing control unit 9, and a defect distribution calculation. Unit 10 and defect map display unit 12.

[Detection head]
The detection head 1a is mounted with the ring illumination 3 and the imaging unit 4 described in the first embodiment. In addition, the imaging part 4 can detect a minute point-like defect also in a high-speed conveyance line like the scale remaining inspection of a pickled steel plate by using a high-definition area sensor camera. It is preferable that the imaging unit 4 has a resolution that is approximately half or less than the minimum size of the defect to be inspected.

  The imaging unit 4 has at least one full width of the strip steel plate 2a while a section having a constant length equal to or less than the minimum continuous length of continuous defects set in the strip steel plate 2a is conveyed. While reciprocating, it sets so that the strip | belt-shaped steel plate 2a may be imaged over the full width in the meantime.

[Traverse part]
As shown in FIG. 8, the traverse unit 6 has a function of reciprocating the ring illumination device 3 and the imaging unit 4 provided in the detection head 1a along the width direction (traverse direction) Y of the strip steel plate 2a. Specifically, the traverse portion 6 includes a guide rail 61 installed so as to straddle a line conveying the strip-shaped steel plate 2a in the width direction Y, and a detection head 1a traveling on the guide rail 61 in the width direction of the strip-shaped steel plate 2a. And a hydraulic cylinder 62 that reciprocates in Y direction.

  As shown in FIG. 6, the traverse unit 6 is driven by a control signal from the imaging timing control unit 9. And as shown in FIG. 8, the traverse part 6 has an inspection range dimension longer than the width dimension w of the strip steel plate 2a so that the optical axis (shown by a broken line) of the imaging unit 4 completely traverses the strip steel plate 2a. The detection head 1a is driven to reciprocate with a W stroke. In addition, it is preferable that this inspection range dimension W also includes the meandering allowance of the strip | belt-shaped steel plate 2a conveyed in a line.

  The traverse portion 6 is detected while a section having a constant length equal to or shorter than the minimum length d (see FIG. 9) of the continuous occurrence length DL of the continuity defect D that is uniquely set in the strip-shaped steel plate 2a is conveyed. The head 1a is set to repeat the same operation each time the head 1a is reciprocated by the stroke W and a section having a certain length is conveyed.

[Conveyance distance detector]
The transport distance detection unit 8 detects the transport distance of the strip steel plate 2 a and outputs the transport distance information of the strip steel plate 2 a to the imaging timing control unit 9. A known rotary encoder or the like can be used as the transport distance detection unit 8.

[Imaging timing controller]
As illustrated in FIG. 7, the imaging timing control unit 9 includes a conveyance direction imaging timing control unit 91, a width direction timing control unit 92, and an imaging position storage unit 93.

  As the imaging timing control unit 9 as a whole, the conveyance distance information of the strip steel plate 2a is input from the conveyance distance detection unit 8 to the conveyance direction imaging timing control unit 91, and the width direction position information of the imaging unit 4 from the traverse unit 6 is the width direction. Each is input to the imaging timing control unit 92. The imaging timing control unit 9 outputs to the imaging unit 4 an imaging trigger signal that is synchronized with the conveyance distance and the traverse (crossing in the width direction Y) in synchronization with these inputs. Details of this will be described later. In addition, the imaging position storage unit 93 stores the imaging position of each image on the strip steel plate 2 a and outputs it to the defect distribution calculation unit 10.

[Image processing unit]
As illustrated in FIG. 7, the image processing unit 7 includes an image input unit 71 to which image data is input from the imaging unit 4, an image data storage unit 72, a defect extraction unit 73, and a defect parameter calculation unit 74. I have.

  Specifically, in the image processing unit 7, captured image data is input from the imaging unit 4 to the image input unit 71. The captured image data input to the image input unit 71 is stored in the image data storage unit 72. The defect extraction unit 73 extracts defects from the captured image data stored in the image data storage unit 72, calculates defect parameters that reflect the hazard degree of defects in the defect parameter calculation unit 74, and calculates a defect distribution calculation unit. 10 is output.

  Here, for the defect extraction, a well-known method, for example, a method of extracting a pixel exceeding a predetermined threshold value as a defective portion after performing preprocessing such as shading correction can be used. The calculation of the defect parameter is for representing the harmfulness of the defect as numerical data for each image. This numerical data includes, for example, the average luminance in the image, the number of pixels exceeding the threshold value, or the number of pixels weighted by the image luminance that exceeds the threshold value, and the correlation with the harmfulness of the defect. Try to choose the one with the highest. Note that the image processing means 7 is provided with an image data storage unit 72 so that images can be confirmed in real time or later so that all captured images or defective images can be stored. preferable.

  In the second embodiment, since the degree of harmfulness is evaluated as a numerical parameter for each image, not for each individual defect, minute dot-like defects are formed in a cluster, particularly like the remainder of the scale. It is extremely effective in evaluating the hazard level of continuous defects. In addition, when trying to evaluate the feature amount of each point-like defect with respect to continuity defects such as the remaining scale, a huge load is applied to the image processing unit and the inspection apparatus is stably operated. I can't do that. In addition, since the degree of harm increases as the density of minute point-like defects increases with the scale remaining, it is more appropriate to evaluate the entire image rather than evaluating each individual defect. It is possible to evaluate the degree of defect toxicity.

[Defect distribution calculator]
As shown in FIG. 6 and FIG. 7, the defect distribution calculation unit 10 conveys defect parameters indicating the degree of defect harm from the image processing unit 7 to the collected images and on the steel plate 2 from which the images are collected. The direction X and the width direction Y position are respectively input from the imaging timing control unit 9. The defect distribution calculation unit 10 calculates the two-dimensional defect occurrence state on the surface of the strip steel plate 2a from these pieces of information.

[Defect map display area]
The defect map display unit 12 displays the defect distribution status calculated by the defect distribution calculation unit 10 on a two-dimensional development view of the surface of the strip steel plate 2a.

[Imaging timing of the imaging unit]
Next, the imaging timing of the imaging unit 4 will be described. Continuity defects D such as scale residue, (1) the occurrence position in the width direction may be local or across the entire width, (2) as shown in FIG. It has a characteristic that it continuously occurs in the transport direction of the strip steel plate 2a for a predetermined length d or more, that is, for a continuous length of at least the continuous generation length DL of the continuity defect set unique to the strip steel plate 2a.

  Based on this finding, the present invention is based on this knowledge, if the inspection is substantially interrupted in the conveying direction if the entire width direction of the surface of the strip steel plate 2a is inspected with a pitch equal to or less than the minimum length d of the continuous defect D, Attention was paid to the fact that the remaining scale (continuity defect D) can be inspected over the entire length of the strip steel plate 2a. In addition, the present invention is intended to reduce the number of cameras by adopting an apparatus configuration that traverses the imaging unit 4 in the width direction.

  That is, the imaging unit 4 is imaged in the section having the constant length while the section having the constant length equal to or less than the minimum length d of the continuous occurrence length DL of the continuity defect D in the strip-shaped steel plate 2a is conveyed. The entire partial area is substantially the entire width of the strip-shaped steel plate 2a, and the sections having a certain length are relatively arranged so that the imaging positions correspond with a pitch having a certain length.

  For example, when the imaging unit 4 is inspected by traversing in the width direction at a constant speed, as shown in FIG. 10, the inspection frequency in the transport direction X is in an area where the transport speed has changed to a low speed (transport speed decrease portion E). To increase. As a result, in this low-speed portion, excessive image acquisition is performed more than necessary, and an unnecessary load is applied to the inspection apparatus, and the image memory is wasted. In addition, since the inspection frequency in the transport direction X changes according to the transport speed, it is difficult to uniformly evaluate the distribution of residual scale generation in the entire steel plate 2 in the transport direction X.

  Furthermore, when the conveyance speed of the strip steel plate 2a becomes extremely slow due to abnormal operation, or when the strip steel plate 2a is temporarily stopped, the same portion of the surface of the strip steel plate 2a will be inspected repeatedly. There is a problem that the same defect is counted repeatedly.

  Therefore, as shown in FIG. 11, every time the strip-shaped steel plate 2a moves by a predetermined distance ΔL, crossing in the width direction of the imaging unit 4 is started, and the entire width of the strip-shaped steel plate 2a is inspected only once. As a result, it is possible to keep the inspection frequency in the conveyance direction constant even when the conveyance speed becomes low or stops. In the example shown in FIG. 11, the imaging unit 4 is moved from one side in the width direction of the strip-shaped steel plate 2 a to the other side so as to capture substantially the entire width, and from the other side in the width direction to one side. In this example, the reciprocating operation of moving to take an image and returning to one side is performed three times during the conveyance length of the minimum length d of the continuity defect D.

  In the second embodiment, the image pickup unit 4 is made substantially full width while being transported for a certain length not more than the minimum length d of the continuous generation length DL of the continuity defect D of the strip steel plate 2a. The image may be moved so as to be imaged over at least one reciprocation. By repeating such an operation, the inspection is performed at a conveyance pitch that is equal to or less than the minimum length d of the continuous generation length DL of the continuity defect D. Therefore, the continuity defect D that is equal to or greater than the minimum length d is substantially full width. It is possible to reliably detect in any imaging region in the inspection.

  FIG. 12 is an enlarged view of the broken line portion of FIG. Squares in the figure indicate an imaging region (partial region) A of the imaging unit 4. The imaging timing control unit 9 receives a signal every time the strip steel plate 2 a is conveyed by the distance ΔL from the conveyance distance detection unit 8, and outputs a traverse start signal of the imaging unit 4 to the traverse unit 6. When the traverse unit 6 receives this signal, the traverse unit 6 starts the Y-traverse in the width direction of the imaging unit 4 and outputs an imaging start signal to the imaging unit 4 every time it traverses the predetermined distance ΔW. T1 to tN shown in FIG. 12 are adjacent to each other so as to be parallel to each other along the conveyance direction X of the strip-shaped steel plate 2a, and have a width ΔW having substantially the same length as the width direction visual field of the imaging unit 4. A plurality of elongate track areas are set for convenience. In the example shown in FIG. 12, the partial areas A in the track areas t1 to tN are continuous in the width direction.

  Then, when the image capturing by the imaging unit 4 finishes capturing the entire width of the strip steel plate 2a, the traverse unit 6 stops the traverse and waits until the next traverse start signal comes from the imaging timing control unit 9. During this time, the imaging unit 4 collects images G11, G12,..., G1N over the entire width of the strip steel plate 2a. When the next traverse start signal is received, the traverse unit 6 starts the traverse again, and outputs an imaging start signal to the imaging unit 4 according to the traverse distance. As a result, the imaging unit 4 collects images G2N, G2N-1, ..., G21.

  As described above, the case where the imaging unit 4 inspects the full width once in each of the forward and backward traverses has been described. However, as shown in FIG. Such a configuration may be adopted. Further, when the traverse speed can be regarded as substantially constant, the imaging unit 4 is continuously imaged at a time interval corresponding to the time to traverse the distance Δw in the width direction, and the imaging start signal for each image acquisition is omitted. You may do it.

  The upper limit value of the traverse speed is a speed at which no inspection omission occurs in the width direction, and this is determined by the width-direction visual field length of the imaging unit 4 and the maximum allowable repeated imaging rate of the imaging unit 4. As shown in FIG. 12, it is preferable that the partial area A slightly overlaps the adjacent track area in the width direction. The lower limit value of the traverse speed is determined by the traverse stroke and the maximum transport speed, and the minimum length d of the continuous occurrence length DL of the continuity defect D to be inspected. That is, for example, in the case of FIG. 11, the imaging unit 4 performs three reciprocating operations, and is set such that 6 · ΔL ≦ d. Further, when a stroke as shown in FIG. 13 is performed, it may be set such that ΔL ≦ d. In other words, while the belt-shaped steel plate 2a is transported by a distance shorter than the minimum length d of the continuous generation length DL of the continuity defect D that is inherently set in the belt-shaped steel plate 2a, the imaging unit 4 is strip-shaped at least once. What is necessary is just to set so that it may move over the full width of the steel plate 2a.

  In the example shown in FIGS. 12 and 13, the partial areas A are continuous. However, as shown in FIG. 14, the partial areas A between the adjacent track areas t are in the transport direction (longitudinal direction). It does not have to be continuous. Also in this case, since the partial area A is positioned once in each adjacent track area t, imaging can be performed over substantially the entire width, and the continuity defect D can be reliably detected. . Note that ΔLt shown in FIG. 14 is the distance traveled by the strip steel plate 2a while the imaging unit 4 traverses from one side of the strip steel plate 2a to the other side, and ΔLr is the return of the imaging unit 4 from the other side to the one side. This is the distance traveled by the strip steel plate 2a. ΔL obtained by adding ΔLt and ΔLr may be set to a length equal to or less than the minimum length d of the continuity defect D.

  The stroke length of the traverse can be made variable according to the change in the plate width of the strip steel plate 2a to be inspected in order to reduce useless image collection. It may be a fixed length regardless of the width. In this case, the stroke length is a length obtained by adding a meandering margin to the maximum value of the assumed strip-shaped steel plate 2a width.

  By controlling the imaging timing as described above, the image processing unit 7 only needs to perform the same routine processing on each image, which simplifies image processing and places an excessive load on the inspection system even in the high-speed conveyance line. Therefore, stable and highly reliable inspection can be performed.

[Specific examples of defect distribution calculation and defect map]
Next, specific examples of defect distribution calculation and defect map display will be described. .., G1N; G2N, G2N-1,..., G21; G31, G32,..., G3N; On the other hand, one defect parameter is calculated and assigned. As a result, the defect occurrence distribution in the width direction Y and the conveyance direction X on the surface of the strip-shaped steel plate 2a can be quantitatively grasped as the numerical value of the defect parameter. The defect map display unit 12 displays the occurrence distribution state on a two-dimensional map as in the display example shown in FIG.

  In the display example of FIG. 15, the display is actually performed in a grid pattern so that the coordinate position of an image obtained by zigzag scanning on the surface of the strip steel plate 2 a can be easily seen. The leftmost column shows the defect information of the image G1i (i = 1, 2,... N), and the second column shows the defect information of the image G2i (i = 1, 2,... N). Indicates that the defect parameter is larger than the predetermined threshold value Th1, and the hatched mesh indicates that the defect parameter is larger than the predetermined threshold value Th2 (Th2 <Th1) and smaller than Th1.

  Thus, by dividing the surface of the strip-shaped steel plate 2a into rectangular meshes and coating the color and pattern of each mesh according to the range of the defect parameter, the distribution of the continuity defect D can be grasped at a glance. When the map is displayed, the map may be indicated by gray scale shading, color separation by color, or display mark. Indications such as “X” for extremely harmful meshes, “△” for moderately harmful meshes, “○” for mildly harmful meshes, and blank for harmless meshes. Can be done.

  The mesh dimensions described above do not necessarily need to match the image size (camera field of view size). For example, two images in the transport direction and three images in the width direction may be assigned to one mesh. Good. In this case, the mesh sizes in FIG. 15 are ΔX = 2ΔL and ΔY = 3ΔW. The process of assigning a plurality of images to one mesh is performed by the defect distribution calculation unit 10. In this case, for example, the average value or the maximum value of the defect parameters of the corresponding plurality of images is calculated as the defect defect degree of the mesh.

  As described above, the second embodiment of the present invention has been described. According to the second embodiment, the imaging unit 4 is based on the conveyance distance information detected by the conveyance distance detection unit 8 and the width direction position information of the imaging unit 4. Since the imaging timing is controlled so that a partial region is imaged over the entire width of the strip steel plate 2a, the surface of the strip steel plate 2a can be inspected at a constant distance even if the transport speed changes, and the transport It is possible to inspect the continuity defect D continuously generated in the direction without omission. At this time, complicated image processing is not necessary, and it is not necessary to image an excessive area, so that image processing is not burdened, and high-speed and highly reliable inspection can be performed even on a high-speed conveyance line. .

  In the second embodiment, the number of cameras used can be reduced by moving the high-definition imaging unit 4 in the width direction Y of the strip steel plate 2a by the traverse unit 6. In the second embodiment, the number of the imaging unit 4 can be made one.

  Further, according to the surface inspection apparatus 11 according to the second embodiment, since the harmfulness of the defect is determined not by individual defects but by numerical parameters extracted from each captured image, micro defects are clustered. When inspecting a continuity defect that is formed in such a manner, the load of image processing can be reduced, and high-speed and stable image processing becomes possible.

  Furthermore, according to the surface inspection apparatus 11 according to the above-described second embodiment, the diffracted light generated by the light shielding unit 3B is set so that the reflected light in the imaging region of the strip steel plate 2a enters the imaging unit 4. By providing the ring illumination device 3 that has been made, it is possible to prevent the generation of brightness patterns of bright and dark spots from the surface of the object to be inspected, and it is possible to accurately inspect minute spot-like defects by suppressing formation noise. Depending on the inspection target, the detection head 1a is not limited to the one that uses the diffracted light shown in the first embodiment, but can also emit normal diffuse illumination light.

  As shown in FIG. 8, the ring illumination device 3 is traversed by being integrated with the imaging unit 4, and thus it is desirable to use a light and small illumination.

(Other embodiments)
As mentioned above, although Embodiment 1 and 2 of this invention were demonstrated, it should not be understood that the description and drawing which make a part of disclosure of said embodiment limit this invention. From this disclosure, various alternative embodiments, examples and operational techniques will be apparent to those skilled in the art.

  For example, in the above embodiment, the case where the present invention is applied to the inspection of the remaining scale defect of the pickled steel sheet is described, but the inspection object of the present invention is not limited to this, and the cold rolled steel sheet or the surface The present invention can also be applied to other steel plates such as treated steel plates, or production lines such as non-ferrous metals such as aluminum, paper, films, and plastics. In addition, the present invention can also be applied to inspection of other surface defects having characteristics that occur continuously in the conveying direction, such as a thread and a periodic defect.

  Moreover, in said embodiment, while the area which has the fixed length below the minimum length d of the continuous generation length DL of the continuous defect D in the strip | belt-shaped steel plate 2a is conveyed, the imaging part 4 has width | variety. The image is continuously captured when traversing from one side of the direction to the other side or from the other side to the one side. For example, half of the width direction is imaged on the forward path and the other half is imaged on the return path. It suffices if the entire captured partial area is imaged over substantially the entire width.

  In the above-described embodiment, a xenon / strobe light source or the like is used as the light source of the ring illumination device 3, but an LED may be used as the light emitting unit 3A.

  Furthermore, in the above embodiment, the traverse unit 6 includes the guide rail 61 and the hydraulic cylinder 62, and the detection head 1a is driven in the width direction by the hydraulic cylinder. However, the imaging unit 4 and the ring illumination device are substantially used. If it is the structure which moves 3 in the width direction synchronizing, it will not be limited to this.

(Example)
Hereinafter, an embodiment in which the present invention is applied to a scale remaining inspection of a pickled steel sheet will be described. In this embodiment, an area sensor camera 4 having a resolution of 0.03 mm and a ring illumination device 3 are fixed to the detection head 1a, and as shown in FIG. 16, two detection heads 1a are installed with a width direction of 840mm apart. It was. With the guide rail 61, the two detection heads 1a are simultaneously traversed in the width direction by 840 mm. Thereby, it was made possible to inspect the full width of the strip steel plate 2a which is a pickled steel plate having a maximum plate width of 1600 mm. A ring-shaped light-shielding plate (not shown) having an opening having a slightly smaller diameter than the light emitting part was installed on the front surface of the light emitting part (not shown) of the ring illumination device 3. The geometrical arrangement conditions shown in FIG. 3 are R = 38 mm, r = 21 mm, H = 100 mm, and θ = 35 °.

  As the transport distance detection unit 8, the image processing unit 7, the defect distribution calculation unit 10, and the imaging timing control unit 9 using a rotary encoder are all performed by a personal computer equipped with an image capture board, a digital input / output board, and an encoder board. It was. A large liquid crystal monitor was used as the defect map display unit 12.

  The imaging unit 4 is traversed once every time the belt-shaped steel plate 2a is transported by 10 m. This is based on the knowledge that the scale remaining defect D is continuously generated in the transport direction by 30 m or more.

  The inspection results are divided into three levels according to the size of the defect parameter and displayed by changing the density of the defect map mesh. An example of the inspection result according to this embodiment is shown in FIG. It was confirmed from FIG. 17 that the generation distribution of the remaining scale in the transport direction X and the width direction Y can be clearly seen.

  In this way, since the color or mark corresponding to the numerical parameter is displayed on the two-dimensional map of the surface of the strip steel plate 2a, the defect occurrence distribution on the material surface can be grasped quantitatively at a glance, and the defect occurrence Time can be dealt with quickly and accurately.

DESCRIPTION OF SYMBOLS 1,11 Surface inspection apparatus 1a Detection head 2 Steel plate 2a Strip | belt-shaped steel plate 3 Ring illumination apparatus 3A Light-emitting part 3B Light-shielding plate 4 Imaging part 4A Area sensor camera 4B Lens 6 Traverse part 7 Image processing part 8 Conveyance distance detection part 9 Imaging timing control Unit 10 Defect distribution calculation unit 12 Defect map display unit 61 Guide rail 62 Hydraulic cylinder 71 Image input unit 72 Image data storage unit 73 Defect extraction unit 74 Defect parameter calculation unit 91 Transport direction imaging timing control unit 92 Width direction imaging timing control unit 93 Imaging position storage unit A Imaging area (partial area)
B Direct light irradiation area C Center axis of ring illumination device C4 Optical axis of imaging unit D Continuity defect H Distance

Claims (9)

A light shielding part having a ring-shaped light emitting part and an optical opening concentrically with the light emitting part and having a diameter smaller than the inner diameter of the light emitting part between the light emitting part and the inspection object A ring illumination device having a plate;
An imaging unit that is disposed on the center line of the opening of the light shielding plate and images the surface of the inspection object through the opening;
With
The imaging region on the surface of the inspection object imaged by the imaging unit is irradiated only with light diffracted at the edge of the opening of the light shielding plate out of the light emitted from the light emitting unit, and the light emitting unit And the surface of the object to be inspected so that the average luminance level in the imaging region on the surface of the object is not less than a predetermined level and the difference in luminance level in the imaging region is within the predetermined range. A surface inspection apparatus characterized by being set. The inspection object is a band-shaped material conveyed along the longitudinal direction,
While the section having a certain length that is equal to or less than the minimum continuous occurrence length of the continuous defect set in the band-shaped material is conveyed, the ring illumination device and the imaging unit are set to have the full width of the band-shaped material. A traverse unit that reciprocates at least once and repeats the same operation every time the section having the predetermined length is conveyed,
The surface inspection apparatus according to claim 1, wherein continuity defects generated on the surface of the belt-shaped material are inspected. A transport distance detector for detecting the transport distance of the belt-shaped material;
The reciprocating operation of the traverse unit based on the conveyance distance information detected by the conveyance distance detection unit, and an imaging timing control unit that controls the imaging timing so that the imaging unit images the belt-like material substantially over the entire width. When,
The surface inspection apparatus according to claim 2, further comprising:   When a plurality of elongate track regions are set adjacent to each other so as to be parallel to each other along the longitudinal direction of the strip-shaped material and having a width of approximately the same length as the visual field in the width direction of the imaging unit, The surface inspection apparatus according to claim 2, wherein the imaging unit images the adjacent partial areas belonging to the respective track areas or the partial areas separated from each other in the longitudinal direction. An image processing unit that extracts the degree of harmfulness of defects in each image captured by the imaging unit as numerical data; and
A defect distribution calculating unit that calculates the two-dimensional defect occurrence status of the surface of the belt-like material based on the defect numerical data of each image extracted by the image processing unit and its imaging position;
The surface inspection apparatus according to any one of claims 2 to 4, further comprising:   The surface inspection apparatus according to claim 5, further comprising a defect map display unit that displays the defect distribution status calculated by the defect distribution calculation unit on a two-dimensional development view of the surface of the band-shaped material.   The surface inspection apparatus according to claim 6, wherein the defect map display unit divides the surface of the belt-like material into rectangular meshes, and displays the defect harmfulness of each mesh by changing a display color or a display mark. .   The surface inspection apparatus according to claim 1, wherein the inspection object is a steel plate, and inspects minute point-like defects in an imaging region on the steel plate.   The surface inspection apparatus according to claim 2, wherein the belt-shaped material is a pickled steel sheet, and the continuity defect is a scale residue.

Images