JP2010117322A - Surface flaw inspection device, method and program for surface flaw inspection - Google Patents

Abstract

A surface flaw inspection apparatus, a surface flaw inspection method, and a program capable of detecting a surface flaw of a belt-like belt to be conveyed more accurately and reliably in real time.
A two-dimensional imaging device for imaging a band-shaped light irradiation area LA on a surface of a band-shaped body irradiated with a band-shaped light L1 on the surface of the belt-shaped body to be conveyed, opposite to the incident of the band-shaped light, and the two-dimensional And a control device 30 for controlling the imaging device, wherein the two-dimensional imaging device controls the reflected light specified at two or more different light receiving angles β among the reflected light from the belt-shaped light irradiation region. An image is captured with an exposure time instructed from the apparatus, and the control apparatus generates a frame image by arranging and synthesizing the pixel rows partially read out from the two-dimensional imaging apparatus in synchronization with the transport speed of the belt. By determining an exposure time suitable for each of the light receiving angles from the image, wrinkle detection is accurately performed in real time at a plurality of light receiving angles in accordance with the generated wrinkles.
[Selection] Figure 1

Description

  The present invention relates to a surface wrinkle inspection device, a surface wrinkle inspection method, and a program, for example, a metal plate or the like, and is particularly suitable for wrinkle inspection in which various wrinkles on the surface of a band are inspected using a photographed image. Technology.

  For example, in a manufacturing process of a strip such as a steel plate that is one of metal plates, defects that may impair product quality are detected early in the manufacturing stage, and the manufacturing conditions of the manufacturing process or the upper process are changed. It is necessary to prevent the occurrence of defects in subsequent products. For this purpose, for example, the inspection of the flaw is performed while moving the steel plate in the production line. Numerous inspection methods, such as electromagnetic and optical methods, have been developed as inspection methods for wrinkles. Among them, optical inspection methods can detect wrinkles on the surface without touching the steel sheet, and the wrinkle image can be detected at high speed. Widely used because it can be easily obtained.

  The optical inspection method illuminates a portion to be inspected on a surface of a steel plate or the like to be passed, and based on an image signal obtained by continuously photographing with an imaging device such as a CCD camera,疵 is detected. In general, in such an optical surface defect inspection method, there is no or little difference between the incident angle of the illumination on the steel sheet and the reflection angle (for example, about 0 ° to 5 °) in order to perform bright field imaging. And a reflection / reflection optical system (hereinafter, 20 ° to 70 °) in which the angle difference between the incident angle and the reflection angle of the illumination on the steel plate is large in order to perform dark field imaging In many cases, two systems of optical systems are used (referred to as "regular reflection optical system" or "diffuse reflection optical system"). It is known that the specular reflection optical system is effective for detecting relatively large uneven ridges on the surface, and the scattering reflection optical system is effective for detecting colored foreign matters and dirt defects on the surface of the steel sheet. .

  However, in recent years, from the viewpoint of implementing quality inspections at the time of shipment of products such as steel sheets and early prevention of wrinkles by early detection of wrinkles, compared to the formation that is the normal part of the surface of steel sheets, It is becoming more and more important to detect even slight differences such as color tone and unevenness.

  In the surface inspection method having two systems of specular reflection optical system and irregular reflection optical system, the imaging device is fixed, that is, the light receiving angle of the imaging device is uniquely fixed within the light receiving angle range that satisfies the above two systems of optical systems. Has been. Therefore, in this surface wrinkle inspection method, it is often difficult to detect wrinkles with slight differences such as color tone and unevenness in imaging at a fixed light receiving angle. For example, in the case of minute uneven ridges such as pressing rods, and dirt that adheres thinly to the steel plate surface, a slight irregular reflection that occurs on the surface of the ridge, the angle dependence of the reflectivity, or the surface properties of the formation, etc. A change in luminance can be obtained only at the light receiving angle, and if the light receiving angle does not coincide with the light receiving angle of the imaging device with high accuracy, it may be difficult to detect the minute wrinkles.

  Therefore, as a device for inspecting surface defects on a steel plate or the like by changing the light receiving angle of the imaging device, the light receiving unit of the two-dimensional CCD camera is used as a pixel array of a two-dimensional CCD element as in the technique disclosed in Patent Document 1. In the width direction of the steel plate, obtained by scanning the pixel row at two or more successive times so as to face the flat steel plate so that the scanning direction substantially coincides with the width direction of the steel plate. An apparatus that combines two or more images of a long specific portion to create inspection images with different light receiving angles is known.

JP 2005-274325 A JP 2006-292593 A “Basics and Applications of CCD / CMOS Image Sensor”, Kazuya Yonemoto, CQ Publishing Co., 2003

However, in the above-described conventional technology, the following problems to be solved may occur.
In the technique disclosed in Patent Document 1, a specific portion that is long in the width direction of the steel sheet moves on the production line by conveyance. This utilizes the fact that the light receiving angle of a certain imaging device changes. A plurality of images at each of a plurality of light receiving angles using a two-dimensional CCD camera so that an image including a specific portion of a steel plate overlaps in an imaging region on a production line captured by the two-dimensional CCD camera. And synthesize the images. Accordingly, in order to obtain two or more different steel plate images having a plurality of predetermined light receiving angles set in advance, the specific portion must be photographed at least twice at a predetermined light receiving angle in the photographing region. is there. In other words, it is always necessary to capture a two-dimensional image including a specific portion with two or more imaging times. For this reason, efficient image capturing is difficult, and it often takes time to generate a composite image for detecting wrinkles.

  In addition, when using a two-dimensional CCD having a large number of pixels in order to obtain a light receiving angle with a large amount of change, it is necessary to further increase the frame rate of the two-dimensional CCD camera, and it is applied to a line that carries at high speed. Have difficulty. Furthermore, when performing the process of synthesizing the above image with the image processing apparatus in the subsequent process from the two-dimensional image data of the frame unit obtained by photographing the specific part of the steel plate at least twice in the imaging region, The image data including the image data of the unnecessary light reception angle is transmitted to the image processing apparatus in the subsequent process, and only the image data (for one pixel column) of the specific portion imaged at the predetermined light reception angle is image data for each frame. Must be extracted from. For this reason, processing of two-dimensional image data in units of frames necessitates data transmission time and image composition processing time. Therefore, with the technique disclosed in Patent Document 1, it is difficult to detect wrinkles in real time when a steel plate is passed, and there is a concern that the cost of the apparatus for performing high-speed image processing increases.

  Furthermore, generally, when the light receiving angle is different, the amount of light reflected on the surface is also greatly different, and the amount of reflected light of the irregular reflection optical system is generally much smaller than that of the regular reflection optical system. However, a plurality of images photographed at each of a plurality of light receiving angles obtained by the technique disclosed in Patent Document 1 are synthesized from images photographed at the same time by a CCD camera. Therefore, the exposure time is inevitably the same at each light receiving angle, and the image synthesized as the regular reflection optical system becomes very bright, while the image synthesized as the irregular reflection optical system becomes very dark. Therefore, significant brightness information is often lost due to whiteout or blackout, and the detection / recognition accuracy of the buttocks often decreases.

  Therefore, the present invention has been made in view of the above problems, and an object of the present invention is to detect the surface flaw of the belt-like belt to be conveyed more accurately and reliably in real time.

In order to solve the above-described problem, according to one aspect of the present invention, a band-shaped light is a region where a band-shaped light is irradiated to the entire width direction of the surface of a band-shaped body conveyed in the longitudinal direction and the band-shaped light is irradiated. A surface wrinkle inspection device that images the reflected light of the belt-shaped light from the irradiation region and inspects the wrinkles on the surface of the belt-shaped body,
An illumination device that irradiates the strip light so that the vertical incident angle of the strip light with respect to the strip is equal to a predetermined value set in advance at all points of the strip light irradiation region;
A plurality of different reflected lights from a plurality of partial regions that are disposed opposite to the illumination device across the belt-shaped light irradiation region, and that are different in position in the moving direction of the belt-shaped body in the belt-shaped light irradiation region. A two-dimensional imaging device capable of column unit exposure and partial readout, which images at an exposure time for each vertical direction reflection angle and outputs a plurality of column unit images;
The number of columns in which at least two or more column unit images are set in advance by causing the two-dimensional imaging device to capture images in synchronization with the movement of the belt-like body and obtaining the plurality of column unit images obtained by the imaging Only by arranging each of the at least two column-unit images, a control device that constitutes a frame image obtained by imaging the belt-shaped light irradiation area from each of the vertical reflection angles corresponding to each other,
A wrinkle image processing device that performs image processing on the frame image and detects wrinkles on the surface of the belt-like body;
A surface flaw inspection apparatus is provided.

The two-dimensional imaging apparatus includes a two-dimensional CMOS camera having a two-dimensional pixel array composed of a plurality of CMOS imaging elements, and exposes an arbitrary pixel row at an arbitrary exposure time according to an instruction from the control apparatus. Outputting the column unit image of the pixel column from the pixel array;
The control device includes a plurality of pixels of the two-dimensional CMOS camera for imaging the plurality of partial regions corresponding to the plurality of vertical reflection angles set in advance with an exposure time for each vertical reflection angle. You may instruct | indicate the exposure time when imaging in a row | line | column and this pixel row | line to the said two-dimensional imaging device.

  Further, the control device extracts a plurality of pixels from the frame image by a predetermined processing method for each frame image composed of the two or more column-unit images, and brightness of the plurality of pixels. An exposure time corresponding to each of the plurality of vertical reflection angles may be determined based on the value.

Moreover, in order to solve the said subject, according to another viewpoint of this invention, in the area | region where the strip | belt-shaped light is irradiated to the whole width direction area | region of the surface of the strip | belt-shaped body conveyed in a longitudinal direction, and this strip | belt-shaped light is irradiated. A surface wrinkle inspection method for imaging reflected light of the band light from a certain band light irradiation region and inspecting wrinkles on the surface of the band,
The illumination device irradiates the band light so that the vertical incident angle of the band light with respect to the band is equal to a predetermined value set in advance at all points of the band light irradiation region;
A two-dimensional imaging device that is arranged opposite to the illumination device with the band-shaped light irradiation region interposed therebetween and capable of column-by-row exposure and partial readout is positioned in the movement direction of the band-shaped body in the band-shaped light irradiation region. Imaging reflected light from a plurality of different partial areas at a plurality of different vertical reflection angles with an exposure time for each vertical reflection angle, and outputting a plurality of column unit images;
The control device causes the two-dimensional imaging device to capture images in synchronization with the movement of the belt-like body, and presets at least two or more column unit images among the plurality of column unit images obtained by the imaging. Arranging each of the number of columns arranged to form a frame image obtained by imaging the band-shaped light irradiation region from each of the vertical reflection angles corresponding to each of the at least two column unit images; and
A wrinkle image processing device performing image processing on the frame image to detect wrinkles on the surface of the strip;
A method for inspecting surface defects is provided.

Moreover, in order to solve the said subject, according to another viewpoint of this invention, in the area | region where the strip | belt-shaped light is irradiated to the whole width direction area | region of the surface of the strip | belt-shaped body conveyed in a longitudinal direction, and this strip | belt-shaped light is irradiated. In order to image the reflected light of the band light from a certain band light irradiation area and inspect the wrinkles on the surface of the band body,
An illumination device that irradiates the strip light so that the vertical incident angle of the strip light with respect to the strip is equal to a predetermined value set in advance at all points of the strip light irradiation region;
A plurality of different reflected lights from a plurality of partial regions that are disposed opposite to the illumination device across the belt-shaped light irradiation region, and that are different in position in the moving direction of the belt-shaped body in the belt-shaped light irradiation region. A two-dimensional imaging device capable of column unit exposure and partial readout, which images at an exposure time for each vertical direction reflection angle and outputs a plurality of column unit images;
A computer for controlling a surface defect inspection apparatus having
The number of columns in which at least two or more column unit images are set in advance by causing the two-dimensional imaging device to capture images in synchronization with the movement of the belt-like body and obtaining the plurality of column unit images obtained by the imaging Only by arranging each of the at least two column unit images, a function of configuring a frame image obtained by imaging the band-shaped light irradiation region from each of the vertical reflection angles corresponding to each of the column unit images,
A function of performing image processing on the frame image to detect wrinkles on the surface of the strip;
A program for realizing the above is provided.

  As described above, according to the present invention, it is possible to detect the surface defects of the belt-like belt to be transported more accurately and reliably in real time.

  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.

<Related technologies>
Prior to describing the embodiments of the present invention, the related art of the present invention will be described.
The surface defect inspection apparatus according to the related art includes an illuminating device that makes planar light incident on a two-dimensional irradiation region on the surface of a steel plate, and a partially readable two-dimensional CMOS camera that is arranged to face the illuminating device. The image signal of only the reflected light specified at two or more predetermined different light receiving angles in the reflected light from the irradiation area is obtained from the pixel row partially read out in synchronization with the steel plate conveyance speed, and the length of the steel plate Combine and arrange in the direction. As a result, the surface flaw inspection apparatus according to the related art simultaneously creates inspection images of two or more light receiving angles in real time.

  The surface defect inspection apparatus according to this related technology includes an imaging device including a two-dimensional CMOS camera having a two-dimensional image array including a plurality of partially readable CMOS imaging elements, and has a planar shape extending in the width direction of the steel plate. Inspection images can be obtained at a plurality of light receiving angles at the same time from within the planar light irradiation region irradiated on the light irradiation region. Therefore, by selecting a plurality of light receiving angles arbitrarily according to the generated wrinkles, real-time detection of wrinkles that can obtain a change in luminance only at a predetermined light receiving angle is supported.

  However, the surface flaw inspection apparatus according to this related technique also has a large difference in luminance between the regular reflection optical system image and the irregular reflection optical system image because the exposure to the CMOS image sensor is at the same time and the same time. Therefore, significant brightness information is lost due to overexposure or blackout, and there is a risk that the detection / recognition accuracy of the hips may be lowered.

  As described in Non-Patent Document 1, the CMOS image pickup device starts the next photoelectric conversion from a pixel column that outputs a photoelectrically converted signal due to the characteristics of its circuit configuration. Therefore, unlike a CCD image sensor in which the start and end of photoelectric conversion are synchronized in all pixels, in a CMOS image sensor, photoelectric conversion is performed asynchronously in units of pixel columns. This characteristic of the CMOS image sensor is a defect that the image is distorted when shooting an object that moves or rotates at high speed. A mechanical shutter is added to the optical system or a separate circuit is added. Thus, efforts have been made to ensure the synchronism of photoelectric conversion similar to that of a CCD image sensor.

  The inventor of the present invention has arrived at the present invention as a result of earnest research on the improvement point in the related technology that significant luminance information is lost. The surface flaw detection apparatus according to an embodiment of the present invention, for example, actively utilizes the asynchronousness of photoelectric conversion possessed by a CMOS image sensor, for example, and individually controls the period of photoelectric conversion for each pixel column. As a result, this surface flaw detection apparatus can inspect the surface of a strip at a plurality of light receiving angles using an exposure time suitable for each light receiving angle without using a plurality of imaging means. In addition, since a plurality of light receiving angles can be selected according to the generated wrinkles, real-time detection of wrinkles that can only change the luminance at a predetermined light receiving angle is possible, and luminance information for each light receiving angle is lost. By making an image without any problem, it is possible to improve the wrinkle detection / recognition accuracy of the wrinkle inspection apparatus, and further, the apparatus can be made smaller and less expensive, so that wrinkle inspection can be easily introduced. Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

<One Embodiment>
FIG. 1 shows an example of an embodiment of the present invention. A strip of an object to be inspected includes a plurality of steel plates in which the rear end and the front end in the longitudinal direction of each steel plate are connected by welding at welding points. It is. FIG. 1 is a schematic configuration diagram of a surface flaw inspection apparatus disposed on a steel plate production line (hereinafter referred to as “steel plate line”) in which the steel plate is processed while moving in the longitudinal direction. In addition, although a steel plate is illustrated here as a strip | belt body here, and the case where a surface flaw inspection apparatus is arrange | positioned at a steel plate line is illustrated, the strip | belt body and application destination of this invention are not specifically limited. Absent. Since the embodiment of the present invention detects surface wrinkles using the reflected light from the strip, the surface wrinkles of various strips can be used as long as the strip has a surface that can reflect the reflected light. Needless to say, it can be detected.

  As shown in FIG. 1, the surface defect inspection apparatus according to the present embodiment images an illumination device 10 that illuminates an imaging region (also referred to as “band-shaped light irradiation region”) LA on the surface of the steel plate 1 and the imaging region. A two-dimensional imaging device 20 for outputting image data, a control device 30 for controlling the two-dimensional imaging device 20 and processing the image data to form a frame image, and subjecting the frame image to image processing. A haze image processing device 31 that detects wrinkles on the surface and an operator haze display device 32 that displays the detected wrinkle information on a screen are provided. In the steel plate line to which the surface flaw inspection apparatus is applied, the roll 2 is brought into close contact with and rotated on the back surface of the steel plate 1 to convey the steel plate 1 in the longitudinal direction. In order to detect the movement of the steel plate 1, the steel plate line includes a sensor such as a rotary encoder (hereinafter referred to as “PLG 40”) that outputs a pulse signal for each set rotation angle of the roll 2. . The surface flaw inspection apparatus according to this embodiment receives a pulse signal from the PLG 40, detects the movement of the steel plate 1, and can operate in accordance with the movement. The method for detecting the movement of the steel sheet 1 is not limited to the pulse signal from the PLG 40, but a method based on a signal from a control device or the like that controls the movement of the steel sheet 1 or a movement of the steel sheet 1 is detected. Various methods such as a method based on the detection result by the movement detection device can be used. In addition, this surface flaw inspection apparatus measures a portion while the steel plate 1 is in close contact with the roll 2, and can inspect with little error due to pass line fluctuation.

(Lighting device 10)
The illumination device 10 includes, for example, a light source 11 having a fluorescent lamp, an incandescent lamp, a halogen lamp, and the like, an optical fiber bundle 12, an input end 13 that guides light emitted from the light source into the optical fiber, and an output that emits light from the optical fiber. It consists of an end 14. In the optical fiber bundle 12, a large number of optical fibers are arranged in a strip shape, the light source 11 is close to the input end 13, and the strip light L <b> 1 is emitted from the output end 14 installed in parallel with the width direction of the steel plate 1. The strip light L1 is preferably parallel light. In addition, the lighting device 10 may be irradiated with band-like light, for example, with ON / OFF and illuminance controlled by a control device 30 described later, but can also operate independently.

  FIG. 2 is a plan view of the steel plate surface on the steel plate line as viewed from above, and shows a planar arrangement of the illumination device 10, the strip-shaped light irradiation surface LA, and the two-dimensional imaging device 20. FIG. 3 is a side view of the steel plate line as viewed from the side, and shows the arrangement of the side surfaces of the illumination device 10, the strip-shaped light irradiation region LA, and the two-dimensional imaging device 20.

  As shown in FIG. 3, the band-shaped light L1 has a band-shaped angle α (vertical incident angle) formed by a rod lens or a cylindrical lens (not shown) provided at the output end 14 and the vertical direction of the steel sheet surface of the steel sheet 1. It is adjusted so as to be equal at all points in the light irradiation area LA, and the vertical light incident angle α is irradiated to the belt-shaped light irradiation area LA at a predetermined angle of about 45 to 70 °, for example. The band-shaped light L1 has a uniform illuminance, and is irradiated so that the irradiation width in the moving direction of the steel plate 1 satisfies the imaging range of the two-dimensional imaging device 20 over the entire width of the band-shaped light irradiation area LA. For example, the illuminating device 10 may use an illuminating device at the output end of a fiber bundle using one 180 W halogen lamp as a light source, and the vertical incident angle α is 45 degrees so as to irradiate the steel plate surface. The band-shaped light irradiation area LA on the steel plate 1 may be set to 1700 mm in the width direction and 70 mm in the length direction. Although the distance from the output end 14 of the optical fiber bundle 12 to the strip light irradiation area LA on the steel plate surface of the strip light L1 is determined by the apparatus configuration of the steel plate line, it is about 200 to 400 mm in the present embodiment. The lighting device 10 may use other belt-like light sources other than those described above, such as a commercially available LED laser light lighting device. In this case, when the light emitting diodes are arranged in a strip shape, at least a part of another configuration such as an optical fiber can be omitted.

(Two-dimensional imaging device 20)
The optical axis of the two-dimensional imaging device 20 that images the strip-shaped light irradiation area LA on the steel plate has an angle β (vertical reflection angle) with the vertical direction of the steel plate surface of the steel plate 1 of, for example, 45 degrees, and the vertical direction. The two-dimensional imaging device 20 is preferably installed so as to be equal to the incident angle α (see FIG. 3). However, the position is not particularly limited as long as it is a position where regular reflection light is included in the imaging surface of the two-dimensional imaging device 20.

  The two-dimensional imaging device 20 is disposed at a position facing the illumination device 10 across the belt-shaped light irradiation area LA, and may be disposed as a single unit, but usually a plurality of units are arranged in parallel in the width direction of the steel plate 1. It may be arranged. The number of two-dimensional imaging devices 20 depends on the required resolution of how small wrinkles are to be detected based on the inspection width of the steel plate 1 and the number of pixels in the lateral direction of the imaging element (the width direction of the steel plate 1). To decide. Usually, there are a plurality of types of steel plates 1 through which the steel plate line passes, and the width of each is about 1 m to 3 m. Therefore, it is desirable to arrange three or four imaging devices in parallel in accordance with this. Note that the distance from the two-dimensional imaging device 20 to the belt-shaped light irradiation area LA on the steel plate surface of the strip-shaped light L1 is the inspection width of the steel plate 1, the viewing angle of the imaging device 20, the required resolution, and the like. It is determined by the arrangement of the apparatus and is usually about 300 to 1200 mm, for example. For example, when the steel plate has a width of 1500 mm and the two-dimensional imaging device 20 uses a camera composed of an image sensor having 2352 horizontal pixels and 200 vertical pixels, the resolution is a pixel size of 0.35 mm in the plate width direction and pixels in the line direction. To obtain an image with a size of 0.35 mm, if the optical conditions are set so that the width direction field of view of one camera is 820 mm, the size of the field of view in the camera width direction is 1640 mm with two cameras. You can get an image.

  The two-dimensional imaging device 20 includes a two-dimensional pixel array in which a plurality of CMOS (Complementary Metal Oxide Semiconductor) imaging elements 22 are two-dimensionally arranged (rows and columns) on the light receiving unit 21 as schematically shown in FIG. The optical axis of the optical axis of the steel plate 1 is parallel to the scanning direction (lateral direction) of the image sensor, and the moving direction of the steel plate 1 is perpendicular to the scanning direction of the image sensor ( (Vertical direction). The two-dimensional imaging apparatus 20 arbitrarily selects a plurality of different pixel columns (hereinafter referred to as “pixel columns”) during one imaging operation, and further arbitrarily sets an exposure time in the pixel columns. Thus, a partial readout means 23 for partially reading out an image signal of the pixel row (hereinafter referred to as “column unit image”) after the exposure is completed is provided.

  Therefore, the two-dimensional imaging device 20 includes a p-row unit image obtained from the p-th pixel row of the CMOS image pickup device 22 and a q-row unit image obtained from the q-th pixel row of the CMOS image pickup device 22. Each can be photographed at an arbitrary exposure time, and they can be obtained during the same imaging operation (where p and q represent any two different column numbers of the CMOS image sensor 22). Although an example in which the CMOS image sensor 22 is used is described here, various image sensors can be used as long as the image sensor has an asynchronous property that allows the exposure time to be set for each pixel column. Is possible. The exposure time is preferably set for each vertical reflection angle corresponding to each pixel column.

  In the embodiment of the present invention, the belt-shaped light L1 irradiated from the output end 14 of the lighting device 10 with the vertical incident angle α inclined is reflected by the belt-shaped light irradiation region LA and is reflected on the light receiving unit 21 of the two-dimensional imaging device 20. An image is formed and the band-shaped light irradiation area LA is imaged. At this time, the two-dimensional imaging device 20 images a plurality of partial regions having different positions in the moving direction of the strip 1 in the strip light irradiation region LA at a plurality of different vertical reflection angles. More specifically, as shown in FIGS. 2 and 3, the reflected light having a constant vertical reflection angle βp is always two-dimensionally from the partial region Lp long in the width direction of the steel plate 1 on the strip-shaped light irradiation region LA. Light is received at the p-th row of the CMOS image sensor 22 of the light receiving unit 21 of the imaging device 20. Similarly, the reflected light having a constant vertical reflection angle βq is always received from the partial region Lq on the belt-shaped light irradiation region LA by the q-th column of the CMOS image sensor 22 of the light-receiving surface 21 of the two-dimensional imaging device 20. It will be. The correspondence between the partial regions Lp and Lq on the strip-shaped light irradiation area LA and the p-column and q-column on the CMOS image sensor 22 is determined in advance by using the optical system for the strip-shaped light irradiation area LA. A captured image is obtained by the image sensor 22 and calibrated.

  By the way, since the band-shaped light irradiation area LA forms an image on the light receiving portion 21 of the two-dimensional imaging device 20, the band-shaped light L1 always has a constant vertical direction from the partial area Lp where the band-shaped light L1 is incident at the vertical direction incident angle α. An image of one scanning line is incident at the reflection angle βp. Therefore, if the pixel row of the CMOS image sensor 22 is arbitrarily selected, the vertical reflection angle can always be selected uniquely within the range of the belt-shaped light irradiation region LA. Although the case where the partial areas Lp and Lq are imaged is described here, the number of the partial areas may be plural and may be three or more.

(Control device 30)
The control device 30 controls the two-dimensional imaging device 20 and the like. More specifically, the control device 30 is a steel plate in which the pixel size in the steel plate line direction (through plate direction) on one image captured by the two-dimensional imaging device 20 moves during the time of one imaging operation. The two-dimensional imaging device 20 outputs an imaging drive signal for imaging once in synchronization with the pulse signal output from the PLG 40 corresponding to the line speed so as to be equal to the length of. In addition, the control device 30 selects a plurality of pixel columns of the CMOS image sensor 22 from which an image signal is read out in a process that will be described in detail later, and at the same time, exposes each pixel column (for each vertical reflection angle). Set the exposure time. Then, the control device 20 sequentially synthesizes each column unit image obtained during one imaging operation from the two-dimensional imaging device 20 in the longitudinal direction of the steel plate, creates a frame image, and sequentially outputs it to the bag image processing device 31. To do.

  FIG. 5 is an explanatory diagram for explaining an example of processing when two pixel columns (p-th column and q-th column) on the CMOS image sensor 22 are selected when the steel sheet 1 is inspected. That is, in the example of FIG. 5, two column-unit images are obtained from the p-th pixel column and the q-th pixel column by one imaging, but the number of column-unit images for this one imaging. There is no particular limitation as long as there is a plurality. This will be described more specifically. FIG. 5A shows a column unit image (p column unit image or q column unit image) captured by the p-th pixel column and the q-th pixel column of the two-dimensional imaging device 20, each steel plate 1. The partial area (Lp or Lq), which is the upper imaging position, and the conveying direction of the steel sheet 1 are upward, the time axis is rightward, and each imaging time t, t + i, t + 2i (where i is required for imaging) A state of movement of the steel plate 1 in (time) is schematically shown. If the number of pixels in the lateral direction of the CMOS image sensor is W, the widths of the p-th and q-th pixel columns are both W pixels. In the embodiment of the present invention, the imaging drive signal output from the control device 30 includes the length of the steel plate that the two-dimensional imaging device 20 moves during one imaging time, and the pixel size in the steel plate line direction on the image to be captured. Since the control device 30 is synchronized so as to be equal to each other, the control device 30 obtains H column unit images (p column unit images or q column unit images) that are obtained by sequential imaging operations and are set in advance. 5), a frame image Ip in which p-row unit images are combined in the longitudinal direction of the steel plate and a q-row unit image in the longitudinal direction of the steel plate simply by arranging and synthesizing them sequentially at each imaging time. The frame image Iq synthesized with the above can be created. In addition, since the frame images Ip and Iq are arranged and combined with H pixel columns each having the number of pixels W as described above, the number of pixels in the steel plate width direction of Ip and Iq (hereinafter referred to as the height of the frame image). And the number of pixels in the longitudinal direction (hereinafter referred to as the width of the frame image) are W and H, respectively.

  The selectable pixel columns of the column-unit image that can be obtained during the same imaging operation are the two p-th image columns and the q-th column corresponding to the two different vertical reflection angles βp and βq in the above. Although the case where the image sequence is used has been described, as described above, the number of pixel columns corresponding to a plurality of different vertical reflection angles β is not limited to two, but can be selected up to the number of image sequences that the light receiving unit 21 has. Is possible.

(About update of operation and exposure time)
As described above, the control device 30 sets the exposure time for each of the plurality of p-th pixel columns and the q-th pixel column. Here, the process for determining the exposure time will be described in detail together with the frame image composition processing with reference to FIGS.

  6 to 9 are explanatory diagrams for explaining the frame image combining process shown in FIG. 5 and the process of determining an exposure time suitable for each vertical reflection angle from the combined frame image. Hereinafter, a process for updating the exposure time corresponding to the vertical reflection angle from the frame image corresponding to one vertical reflection angle will be described. However, the same applies to the case where a plurality of vertical reflection angles are used as in the present embodiment, that is, the case where a plurality of column unit images are captured, and the exposure time is calculated from the frame image corresponding to each vertical reflection angle. do it.

  As will be described later, a preset initial value is set for the exposure time. Then, the exposure time is updated according to the processing flow shown in FIG. 6 using the frame image captured and created with the initial value.

  First, the illumination device 10 starts irradiating the band-shaped light irradiation area LA with the band-shaped light L1 (an example of a step of irradiating the band-shaped light), and the two-dimensional imaging apparatus 20 sets the band-shaped light irradiation area LA to a plurality of pixels. An image is taken in a row (an example of a step of outputting a plurality of row-unit images), and then the processing after step 51 is started.

  In steps 51, 52 and 53, as shown in FIG. 5, the control device 30 acquires column unit images corresponding to the respective vertical reflection angles from the two-dimensional imaging device 20, and creates a frame image (frame image). An example of steps constituting the above. The above-described column-unit image acquisition and frame image synthesis are repeated until the frame image is completed (see step 53). In other words, a column unit image having a vertical reflection angle height of 1 is obtained in steps 51 and 52, and one frame image is obtained by repeating this H times.

  Next, in step 54, the control device 30 extracts pixels from the frame image by a predetermined processing method. As a specific example of the extraction method, there is a method of extracting only the pixels of the steel plate portion by detecting the left and right edges of the steel plate and extracting the pixels in the edges as shown in the processing flow of FIG. An example of the extraction method performed in step 54 will be described as follows.

As shown in FIG. 7, in step 61, the control device 30 adds the luminance value of each pixel of the frame image to the height direction (that is, the stacking direction of the column unit image or the sheet passing direction of the steel plate 1), A distribution (referred to as vertical projection) P (x) in the width direction x (1 ≦ x ≦ W) is calculated (see FIG. 8). That is, assuming that the luminance value of the pixel at the height direction position y and the width direction position x of the frame image is I (x, y), the vertical projection P (x) is calculated as follows.
P (x) = I (x, 1) + I (x, 2) +... I (x, H), 1 ≦ x ≦ W

  Next, in step 62, the control device 30 scans the calculated vertical projection P (x) from x = 1 to the positive direction of the axis x and from x = W to the negative direction of the axis x, respectively. The positions in the width direction exceeding the specified value (threshold value) are determined as the left edge position XL and the right edge position XR, respectively (see FIG. 9). Then, the control device 30 regards the region of the frame image XL ≦ X ≦ XR as a steel plate portion from the determined left and right edge positions, and extracts pixels of the steel plate portion from the frame image in step 63.

Referring again to FIG. 6, the description returns to the exposure time update processing.
In step 55, the control device 30 calculates an average value Vm of the luminance values of the pixels extracted in step 54. In step 56, the control device 30 further corrects the exposure time so that the calculated average value Vm falls within a preset luminance range. Assuming that the maximum value of the set luminance range is Vmax, Vmin, the exposure time change rate is ΔE (> 0), the current exposure time is E, and the updated exposure time is E ′,
When Vm> Vmax, E ′ = E (1−ΔE)
When Vm <Vmin, E ′ = E (1 + ΔE)
In other cases, E '= E
The updated exposure time E ′ is determined as follows. Here, the luminance range (Vmin, Vmax) is set to a luminance range suitable for processing in the later-described image processing apparatus 31 from the range that the luminance value of each pixel of the frame can take. For example, when a pixel is represented by 8 bits, the range of luminance values that can be taken is from 0 to 255. Therefore, it is desirable to set Vmin = 80 and Vmax = 220, but this value is merely an example.

  Then, after the exposure time is updated in step 56, the control device 30 confirms whether or not imaging has been completed in step 57, that is, whether or not all desired frame images have been captured, and further captures frame images. If it should be, imaging is performed with the updated exposure time by repeating the processing from step 51 onward.

  Note that if the exposure time is changed during the eyelid inspection, a luminance change occurs between a plurality of frame images with respect to the same vertical reflection angle. When referring to the frame image, it may be difficult to determine wrinkles. Therefore, the exposure time change rate ΔE is sufficiently small with respect to the wrinkle detection sensitivity in the wrinkle image processing device 31 so as not to cause such a situation that detection of wrinkles becomes difficult, for example, ΔE = 0.05. It is desirable to set the degree.

  Further, in the exposure time changing process shown in FIG. 6, the exposure time for each pixel row for creating one frame image is calculated using the average value of the luminance of the frame image captured before the frame image. Although changed, this change may be performed in the middle of acquiring a pixel row forming one frame image. However, in this case, for example, it is preferable that the average value Vm of the luminance values is calculated and the above processing is performed when a predetermined number or more of column-unit images for the same pixel column are acquired. For example, when a cocoon image is included in a column-unit image, the average value Vm of the luminance values includes abnormal luminance from the cocoon, so a plurality of column-unit images instead of a frame image as described above. When calculating the average value Vm from the above, it is desirable to calculate the average value Vm of luminance values for a predetermined number of column-unit images. However, in this case, when the exposure time is changed during the eyelid inspection, the boundary between the area before the exposure time change and the area after the change is reflected in one frame image, and a large luminance change occurs in this boundary portion. It can be considered that this boundary is erroneously detected as a trap. Therefore, also in this case, the change rate ΔE of the exposure time is desirably sufficiently small with respect to the wrinkle detection sensitivity in the haze image processing device 31, and is preferably set to about ΔE = 0.05, for example.

  Further, as described above, a plurality of pixel columns (for example, the q-th column and the q-th column) for capturing a column-unit image are set by the control device 30, and before the exposure time is updated by the processing of FIG. It is desirable that a predetermined initial value is set in advance as the exposure time. The selection of the pixel column and the initial value of the exposure time will be described as follows.

  That is, if the control device 30 images the steel plate at any vertical reflection angle (uniquely determined by the pixel row read by the CMOS image sensor 22) in advance corresponding to the type and surface properties of the steel plate 1, the accuracy is high. It is desirable that whether or not wrinkles can be detected well is examined in advance by experiments or the like and registered in a database (not shown) built in as a vertical reflection angle or pixel column number for each type of steel plate 1 or the like. Similarly, with respect to the exposure time, the control device 30 similarly corresponds to the type and surface properties of the steel plate 1 in advance, and what exposure time is used to image the steel plate so that the exposure time does not need to be changed frequently. It is desirable to study in advance through experiments or the like and register it in the database as the initial value of the exposure time for the product type and surface texture. Then, when actually inspecting the steel plate line, the control device 30 determines the initial values of the vertical reflection angle and the exposure time suitable for the type and surface properties of the steel plate to be inspected based on the database. It is desirable to control the imaging by outputting the readout pixel column setting signal of the CMOS image sensor 22 and the exposure time setting signal of the pixel column to the two-dimensional imaging device 20.

(疵 image processing device 31)
The image processing device 31 receives a frame image created from a plurality of column-unit images by the control device 30 and, for example, improves image quality such as shading correction, labeling processing, and geometric feature amount for the frame image. Image processing such as image analysis such as extraction, and wrinkle determination processing are performed to detect harmful wrinkles (an example of a step for detecting wrinkles). For example, as disclosed in Patent Document 2 filed by the present applicant, the hail image processing device 31 binarizes the luminance value using the threshold value and extracts a region above the threshold value as a haze candidate image, for example. The candidate image is extracted from the obtained frame image, such as extracting the candidate image by pattern fitting the luminance distribution of the frame image. Then, the haze image processing device 31 extracts an image including a pattern of a specific shape from the extracted haze candidate images, and performs predetermined image processing (for example, pattern fitting, feature amount extraction and determination) on the extracted image. The target of image processing to be performed for detecting harmful poisons may be narrowed down by performing harmful detection. By performing such processing, the wrinkle image processing device 31 can detect wrinkles at high speed and with high accuracy, can efficiently perform image processing only on the surface portion of the steel sheet, and moves on the roll 2. Even if the steel sheet is meandering so-called “walk”, the occurrence position of harmful flaws can be accurately calculated from the edge of the steel sheet.

(Operator 疵 display device 32)
The operator bag display device 32 is configured by a display device such as a computer display, for example, and superimposes a signal including a detected bag image sent from the bag image processing device 31 and a feature amount extracted from the bag image. Then, the image of the cocoon and the feature amount are displayed. Here, for example, the feature amount may include data such as the type of soot, the occurrence position on the steel plate, the size, and the degree of harmfulness.

<Examples of effects according to this embodiment>
As described above, according to the surface flaw inspection apparatus of the embodiment of the present invention, by selecting the pixel column of the light receiving unit 21 of the two-dimensional imaging device 20 and arbitrarily setting the exposure time of the pixel column, It is suitable for each of a plurality of light receiving angles in an imaging time short enough to be regarded as the same time as the normal imaging time for exposing all the pixel columns of the light receiving unit 21 of the two-dimensional imaging device 20 without increasing the number of imaging devices. A plurality of soot images captured using the exposure time can be obtained, and a regular reflection optical system and an irregular reflection optical system can be easily constructed.
At this time, the surface wrinkle inspection apparatus can also select a plurality of light receiving angles in accordance with wrinkles generated. In other words, for example, the surface flaw inspection device has little scattered / reflected light from the formation that is the normal part of the band, and the difference in irregular reflection due to minute wrinkles on the surface of the band and the angle dependence of the reflectance A suitable light receiving angle can be selected. Then, as described above, the surface wrinkle inspection apparatus sets an exposure time suitable for each of the light receiving angles, efficiently captures an image with the exposure time, and performs high-speed and high-precision image processing to detect wrinkles. it can.
Therefore, this surface wrinkle inspection apparatus can realize real-time detection of wrinkles that can obtain a change in luminance only at a predetermined light receiving angle with higher accuracy than before, and can be easily and inexpensively performed on a steel plate. Can be tested for various wrinkles. In addition, this surface defect inspection apparatus is simple, can be downsized, and can be inexpensive. Therefore, the apparatus can be easily introduced.

  More specifically, in the embodiment of the present invention, in the embodiment of the present invention, the two-dimensional imaging device 20 similarly uses a plurality of predetermined different pixel columns selected by the control device 30 during one imaging operation. Means is provided for reading out the image signal of the pixel column after exposure using the time set for each pixel column by the control device 30. Therefore, according to this surface defect inspection apparatus, if the control device 30 previously selects the p-th column of the CMOS image sensor 22 and the q-th column of the CMOS image sensor 22, an arbitrary image can be obtained by one imaging operation. A p-row unit image with a vertical reflection angle βp and a q-row unit image with a vertical reflection angle βq captured at the exposure time can be obtained simultaneously.

  Furthermore, as described above, the two-dimensional imaging device 20 repeats the imaging operation using the imaging drive signal output from the control device 30 and synchronized with the pulse signal PLG corresponding to the movement of the steel plate 1. Therefore, the surface defect inspection apparatus according to the present embodiment can easily perform a frame image (FIG. 5, Ip) captured at the vertical reflection angle βp and a frame image (FIG. 5, Iq) captured at the vertical reflection angle βq. Can be obtained continuously over the entire length of the steel sheet. Therefore, this surface wrinkle inspection apparatus can construct an arbitrary irregular reflection optical system suitable for the ground surface of the steel sheet 1, and it has a predetermined difference due to a slight difference in irregular reflection occurring on the wrinkle surface and the angle dependency of the reflectance. With respect to the eyelid that can obtain the luminance change only at the light receiving angle, the image can be taken at the optimum light receiving angle (that is, the vertical reflection angle).

  Furthermore, since this surface defect inspection apparatus can set an exposure time suitable for each vertical reflection angle, it is possible to obtain a frame image corresponding to the luminance range (dynamic range) at each vertical reflection angle. Significant loss of luminance information due to overexposure or underexposure, which is a problem in shooting with exposure time, can be reduced, and the accuracy of detection and recognition of the buttock can be improved.

  Also, as shown in FIG. 3, the optical axis of the two-dimensional imaging device 20 reflects at a vertical reflection angle β (difference from the vertical incident angle is 0 ° to 5 °) substantially equal to the vertical incident angle α. A specular reflection optical system can be constructed by selecting a pixel row with the control device 30 so as to receive light at the r-th row of the CMOS image sensor 22 of the light receiving surface 21 of the two-dimensional imaging device 20 that receives reflected light. As a result, it becomes easier to detect the irregularity wrinkles. Here, in order to construct the regular reflection optical system, in order to select the r-th column of the CMOS image sensor 22, the intensity of the reflected light may be measured and the pixel column having the highest intensity may be selected. In addition, when the steel plate formation is a mirror surface, intense reflected light from the formation is obtained by selecting a pixel row that satisfies the condition (twisting zone) shifted by about ± 1 to 2 ° from the regular reflection optical system. As a result, it is possible to obtain a result that the fine concavo-convex characteristic flaws can be realized. FIG. 5B shows, as an example, an image obtained when photographing regular reflection unevenness wrinkles (疵 H) and foreign matter adhesion (疵 G) that is colored wrinkles. In FIG. 5B, Ip is a frame image obtained by the regular reflection optical system (that is, an example in which p = r), and Iq is a frame image obtained by the irregular reflection optical system. From the image obtained with the regular reflection optical system, it is possible to reliably detect minute irregularities and the like, and from the image obtained from the irregular reflection optical system, colored wrinkles such as foreign matter adhesion.

  Then, as described above, the surface defect inspection apparatus selects one or two or more pixel columns from the plurality of pixel columns of the light receiving unit 21 of the two-dimensional imaging device 20, and selects a column unit image from each of the selected pixel columns. To get. Then, the surface defect inspection apparatus superimposes the column unit images acquired from the respective pixel columns for each pixel column, and forms a frame image corresponding to each pixel column (that is, the vertical reflection angle). Therefore, for example, as in the technique described in Patent Document 1, a plurality of frame images including all pixel columns are formed, and an inspection image for each light receiving angle is formed from the plurality of frame images. Thus, the surface defect inspection apparatus according to the present embodiment does not need to expose all the pixel columns, and only needs to expose the selected pixel columns, so that the real-time property can be greatly improved. Such a surface defect inspection apparatus can also set an appropriate exposure time for each pixel column by selecting a pixel column from which a column-unit image is obtained.

  As mentioned above, although preferred embodiment of this invention was described in detail, referring an accompanying drawing, it cannot be overemphasized that this invention is not limited to this example. 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.

(Example of pixel extraction process change)
For example, in the above-described embodiment, the case where the control device 30 obtains the edge from the vertical projection as illustrated in FIG. 7 when extracting a plurality of pixels in step 54 has been described. However, this processing method is not limited to this example, as long as it is a method that can extract a plurality of pixels in the region within the edge, that is, the region in which the steel plate 1 to be inspected is imaged, from the frame image. Various methods can be used. For example, the control device 30 can operate for each pixel column (that is, a column-unit image) of the frame image using a threshold value to detect the left and right edge positions and extract pixels between the left and right edge positions. is there. In this case, the control device 30 can detect the edge position for each column unit image, for example, when the steel plate 1 is wavy or the width is not constant, and the pixel representing the steel plate 1 is more accurately detected. It becomes possible to extract. Note that the edge detection method described here is merely an example, and it is needless to say that various edge detection methods can be used during pixel extraction processing.

(Example of changing exposure time adjustment)
In the above embodiment, the control device 30 uses the average value Vm calculated in step 55 shown in FIG. 6 so that the average value Vm falls within the preset luminance range Vmax to Vmin in step 56. In addition, the exposure time was changed by a change rate ΔE. However, the exposure time adjustment method is not limited to this example, and various exposure time adjustments are possible as long as the brightness value of a frame image to be captured thereafter is a value that does not cause overexposure or blackout. The method can be used.

For example, the control device 30 can adjust the exposure time so that the average value Vm approaches the predetermined target value Vt. More specifically, in this case, assuming that the change rate of the exposure time is ΔE (> 0), the current exposure time is E, and the updated exposure time is E ′,
When Vt> Vm, E ′ = E (1−ΔE)
When Vt <Vm, E ′ = E (1 + ΔE)
In other cases, E '= E
It is also possible to determine the updated exposure time E ′ as shown in FIG. Also in this case, the exposure time E ′ is adjusted within the range where the average value Vm is close to Vt, and appropriate exposure time adjustment is possible.

Further, as another example of changing the exposure time adjustment, a method using a correlation (for example, a linear relationship) existing between the luminance value and the exposure time can be used. More specifically, when there is a linear relationship between the brightness value and the exposure time, it is expected that the brightness value will be increased by n times if the exposure time is increased by n times. Therefore, when the target value of the average value Vm is Vt, the exposure time magnification is n, the current exposure time is E, and the updated exposure time is E ′, the control device 30 has the magnification n:
n = Vt / Vm
Further, the exposure time E ′ after the change is
E ′ = n × E
It is also possible to determine as follows. At this time, for example, when the exposure time is changed during the formation of one frame image, the change magnification n is multiplied by the limiter value nl in order to prevent over-detection due to a sudden change in brightness. To n ′ n ′ = min (n, nl)
To change the exposure time E ′ after the change,
E ′ = n ′ × E
It is desirable to decide as follows.

  In the case of the above-described embodiment shown in FIG. 6 and the example of adjusting the exposure time described here, an area excluding the edge portion is extracted from the frame image, and an average value of luminance values for the area is used. However, the control device 30 can use the median value or the like instead of the average value of the luminance values.

(Program etc.)
In addition, for example, the control device 30 and the eyelid image processing device 31 may be configured by individual dedicated devices as shown in FIG. 1, but the predetermined function may be realized by a computer by software. When the above functions are performed by software or the like, the functions of the control device 30 and the eyelid image processing device 31 can be realized by causing a general-purpose or dedicated computer to execute the program. In other words, the control device 30 and the eyelid image processing device 31 and the like each include an input device such as a keyboard and a mouse for inputting the set values for the above-described eyelid detection by an operator who realizes a predetermined function, and an operator eyelid display. An input / output interface for the device 32, an internal memory including an image memory, an external recording device such as a DVD-RAM or an HDD, and a computer having a computer display or the like can also be used. Further, the control device 30 and the eyelid image processing device 31 may be configured by separate computers, respectively, or may be configured by a single unit in order to reduce the installation location and production cost. In addition, a process for controlling a production line for steel plates, etc. = an I / O board connected to a computer or a LAN or a dedicated cable, etc., so that information on the type of steel plate to be inspected can be obtained from the process computer Alternatively, the wrinkle detection result may be output to the process computer. The above-described control and each data processing (information processing) performed by the control device 30 and the eyelid image processing device 31 are devices configured by a computer or the like as described above. It is also possible to load and execute it in the internal memory.

It is a figure which shows the outline of the surface flaw inspection apparatus of one embodiment of this invention. It is a top view which shows arrangement | positioning of the illuminating device and two-dimensional imaging device which were seen from the upper direction of the surface flaw inspection apparatus of embodiment shown in FIG. It is a side surface arrangement | positioning figure of the surface flaw inspection apparatus of embodiment shown in FIG. It is a figure which shows typically the CMOS image sensor of the light-receiving part of a two-dimensional imaging device. It is explanatory drawing of the outline of the method of deriving | requiring a frame image from two column unit images. It is explanatory drawing of the outline of the method of updating exposure time using the derived | led-out frame image. It is explanatory drawing of the outline of the method of updating exposure time using the derived | led-out frame image. It is explanatory drawing of the outline of the method of updating exposure time using the derived | led-out frame image. It is explanatory drawing of the outline of the method of updating exposure time using the derived | led-out frame image.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Steel plate 2 Roll 10 Illuminating device 11 Light source 12 Optical fiber bundle 13 Optical fiber bundle input end 14 Light is a fiber bundle output end 20 Two-dimensional imaging device 21 Light-receiving part 22 CMOS imaging element 23 Partial reading means 30 Control device 31 疵 Image processing device 32 Operator疵 Display device 40 PLG
L1 Band-shaped light LA Band-shaped light irradiation region Lp, Lq Image pickup position (partial region) in the band-shaped light irradiation region

Claims (5)

The band-shaped light is irradiated to the entire width direction of the surface of the band-shaped body conveyed in the longitudinal direction, and the reflected light of the band-shaped light from the band-shaped light irradiation area that is the area irradiated with the band-shaped light is imaged. A surface wrinkle inspection device for inspecting surface wrinkles of
An illumination device that irradiates the strip light so that the vertical incident angle of the strip light with respect to the strip is equal to a predetermined value set in advance at all points of the strip light irradiation region;
A plurality of different reflected lights from a plurality of partial regions that are arranged opposite to the illumination device across the belt-shaped light irradiation region, and that are different in position in the moving direction of the belt-shaped body in the belt-shaped light irradiation region. A two-dimensional imaging device capable of column unit exposure and partial readout, which images at an exposure time for each vertical direction reflection angle and outputs a plurality of column unit images;
The number of columns in which at least two column unit images are set in advance among the plurality of column unit images obtained by the two-dimensional imaging device in synchronization with the movement of the strip. Only by arranging each of the at least two column unit images, a control device that constitutes a frame image obtained by imaging the band-shaped light irradiation region from each of the vertical reflection angles corresponding to each of the column unit images,
A wrinkle image processing device that performs image processing on the frame image to detect wrinkles on the surface of the strip;
A surface defect inspection apparatus comprising: The two-dimensional imaging device includes a two-dimensional CMOS camera having a two-dimensional pixel array composed of a plurality of CMOS imaging elements, and exposes an arbitrary pixel row at an arbitrary exposure time according to an instruction from the control device. Output the column-unit image of the pixel array,
The control device includes a plurality of pixels of the two-dimensional CMOS camera for imaging the plurality of partial regions corresponding to the plurality of vertical reflection angles set in advance with an exposure time for each vertical reflection angle. Instructing the two-dimensional imaging device the exposure time when imaging with a column and the pixel column,
The surface flaw inspection apparatus according to claim 1. The control device extracts a plurality of pixels from the frame image by a predetermined processing method for each of the frame images configured from the two or more column-unit images, and sets the luminance values of the plurality of pixels. And determining an exposure time corresponding to each of the plurality of vertical reflection angles.
The surface flaw inspection apparatus according to claim 1 or 2, wherein The band-shaped light is irradiated to the entire width direction of the surface of the band-shaped body conveyed in the longitudinal direction, and the reflected light of the band-shaped light from the band-shaped light irradiation area that is the area irradiated with the band-shaped light is imaged. A surface wrinkle inspection method for inspecting surface wrinkles of
The illumination device irradiates the band light so that the vertical incident angle of the band light with respect to the band is equal to a predetermined value set in advance at all points of the band light irradiation region;
A two-dimensional imaging device that is arranged opposite to the illumination device with the band-shaped light irradiation region interposed therebetween and that can perform column-by-column exposure and that can be partially read out has a position in the band-shaped light irradiation region in the moving direction. Imaging reflected light from a plurality of different partial areas at a plurality of different vertical reflection angles with an exposure time for each vertical reflection angle, and outputting a plurality of column unit images;
The control device causes the two-dimensional imaging device to capture images in synchronization with the movement of the belt-like body, and presets at least two or more column unit images among the plurality of column unit images obtained by the imaging. Forming a frame image obtained by imaging the belt-shaped light irradiation area from each of the vertical reflection angles corresponding to each of the at least two or more column-unit images by arranging the number of columns arranged; and
A wrinkle image processing device image-processing the frame image to detect wrinkles on the surface of the strip;
A method for inspecting surface wrinkles, comprising: The band-shaped light is irradiated to the entire width direction of the surface of the band-shaped body conveyed in the longitudinal direction, and the reflected light of the band-shaped light from the band-shaped light irradiation area that is the area irradiated with the band-shaped light is imaged. In order to inspect the surface wrinkles
An illumination device that irradiates the strip light so that the vertical incident angle of the strip light with respect to the strip is equal to a predetermined value set in advance at all points of the strip light irradiation region;
A plurality of different reflected lights from a plurality of partial regions that are arranged opposite to the illumination device across the belt-shaped light irradiation region, and that are different in position in the moving direction of the belt-shaped body in the belt-shaped light irradiation region. A two-dimensional imaging device capable of column unit exposure and partial readout, which images at an exposure time for each vertical direction reflection angle and outputs a plurality of column unit images;
A computer for controlling a surface defect inspection apparatus having
The number of columns in which at least two column unit images are set in advance among the plurality of column unit images obtained by the two-dimensional imaging device in synchronization with the movement of the strip. Only by arranging each of the at least two column unit images, a function of configuring a frame image obtained by imaging the band-shaped light irradiation region from each of the vertical reflection angles corresponding to each of the column unit images,
A function of performing image processing on the frame image to detect wrinkles on the surface of the strip;
A program to realize