JP2017129427A - Defect inspection device, defect inspection method, and defect inspection program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a defect inspection device and method for detecting surface defects of a sheet material with high accuracy.SOLUTION: The defect inspection device includes a line sensor camera for capturing the image of the surface of a sheet material A carried by a roller 2 in order to detect surface defects of the sheet material A. It also includes a dark-side reference luminosity detection member 3 and a light-side reference luminosity detection member 4 having appropriate average luminosity at an end of the roller 2. It further includes a defect detection processing unit for detecting the presence and type of defect from the output data of the line sensor camera. The defect detection processing unit finds the average luminosity of the dark-side reference luminosity detection member 3, the average luminosity of the light-side reference luminosity detection member 4, and a luminosity change of the sheet material A from the image data of the line sensor camera, and expands the range of luminosity changes of the sheet material A on the basis of the found average luminosity of the dark-side reference luminosity detection member 3 and average luminosity of the light-side reference luminosity detection member 4.SELECTED DRAWING: Figure 2

Description

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

  Conventionally, a color line sensor camera as a defect inspection device that inspects the surface condition of a sheet material, the printing condition of a pattern or characters printed on the surface of the sheet material, the state of baking or painting applied to the surface of the sheet material, etc. There is known a technique for optically detecting defects by imaging the surface of a sheet material using, and processing the obtained image (see, for example, Patent Document 1).

  The defect inspection apparatus described in Patent Document 1 captures the surface of a sheet material and the white area and black area of a reference plate with a three-plate CCD line sensor camera, and obtains each color obtained from image data on the surface of the sheet material to be inspected. The component signal is subjected to white balance correction based on white balance parameters calculated from the image data of the white area and black area of the reference plate. Thereby, even if there is a change in the inspection environment, the defect inspection apparatus described in Patent Document 1 can correct the optimum white balance at that time in real time.

  Note that the defect inspection apparatus described in Patent Document 1 converts each color component signal that has been subjected to white balance correction processing into H (hue), S (saturation), and I (lightness) signals, and performs the subsequent processing. This is done using the HSI signal. Thereby, the defect of the color-printed sheet material can be detected with high accuracy.

  However, even if the prior art is effective for defect inspection of color-printed sheet material, it is possible to accurately detect defects on the surface of a plain sheet material with a small difference in brightness between a defective portion and a non-defective portion. There is a problem that it cannot be detected.

  In other words, the surface defects of plain sheet materials include scratches, dents, foreign matter contamination, stains, uneven material, etc., but the difference in brightness between this type of defect and non-defect parts is common. Therefore, it is difficult to detect a defect with high accuracy only by optimally correcting the white balance. Therefore, in order to detect the surface defect of the plain sheet material with high accuracy, it is indispensable to expand the brightness difference between the defective portion and the non-defective portion.

  The present invention has been made based on such knowledge, and an object thereof is to detect a surface defect of a sheet material with high accuracy.

  In order to solve the above-described problem, an aspect of the present invention includes a roller that conveys a sheet material that is an inspection object, a dark-side reference brightness detection member and a light-side reference brightness detection member that are attached to the roller, A line sensor camera that images the sheet material conveyed by the roller, the dark-side reference brightness detection member, and the light-side reference brightness detection member for each scan; and image processing that outputs image data output from the line sensor camera A defect detection processing unit that detects the presence or absence of a defect on the surface of the sheet material, and the defect detection processing unit detects an average brightness of the dark side reference brightness detection member from the image data, and the light side reference brightness. The average brightness of the detection member and the brightness change of the sheet material in the scanning direction of the line sensor camera are obtained, and the obtained dark Based on the average brightness of the reference brightness detection member and the average brightness of the light side reference brightness detection member, the range of brightness change of the sheet material is expanded, and based on the brightness change of the sheet material after the range is expanded A surface defect of the sheet material is detected.

  According to the present invention, a surface defect of a sheet material can be detected with high accuracy.

It is a figure showing the whole defect inspection device composition concerning an embodiment. It is a figure which shows the structure of the roller with which the defect inspection apparatus which concerns on embodiment, a dark side reference | standard brightness detection member, and a light side reference | standard brightness detection member, and the imaging width of a line sensor camera. It is a figure which shows the hardware constitutions of the defect inspection apparatus which concerns on embodiment. It is a figure which shows the picked-up image of the line sensor camera with which the defect inspection apparatus which concerns on embodiment was equipped. It is a flowchart which shows the process sequence of the defect inspection process which the defect inspection apparatus which concerns on embodiment performs. It is a graph which shows the 1st example of the brightness range expansion process which the defect inspection apparatus which concerns on embodiment performs. It is a graph which shows the 2nd example of the brightness range expansion process which the defect inspection apparatus which concerns on embodiment performs. It is a figure which shows the captured image of the line sensor camera after the defect inspection apparatus which concerns on embodiment performs the brightness range expansion process. It is a figure which shows the other structural example of the roller with which the defect inspection apparatus which concerns on embodiment is equipped, the dark side reference | standard brightness detection member, and the light side reference | standard brightness detection member.

  Hereinafter, embodiments of a defect inspection apparatus, a defect inspection method, and a defect inspection program according to the present invention will be described with reference to the drawings.

  As shown in FIG. 1 and FIG. 2, the defect inspection apparatus 1 according to the present embodiment includes a plurality of rollers 2 that convey a sheet material A that is an inspection object, and one selected from the plurality of rollers 2. A dark side reference brightness detecting member 3 attached near one end of the book roller 2, a light side reference brightness detecting member 4 attached near the other end of the roller 2, a sheet material A conveyed by the roller 2, dark The side reference lightness detection member 3 and the light side reference lightness detection member 4 are output from the illumination device 5, the line sensor camera 6 that images the illumination portion of the illumination device 5 for each scan, and the line sensor camera 6. The image processing is mainly composed of a defect detection processing unit 7 that performs image processing on image data to detect the presence or absence of defects on the surface of the sheet material A.

  The defect inspection apparatus 1 according to the present embodiment can be incorporated as an in-line inspection apparatus in the manufacturing process of the sheet material A or the manufacturing process of the member using the sheet material A as a raw material, or offline inspection that inspects the sheet material A alone. It can also be used as a device.

  The sheet material A may be wound in a roll shape or may be cut into a predetermined dimension. The sheet material A to be inspected includes a plastic film and a thermal sheet, but is not particularly limited here. In short, any material that can be stably conveyed using the roller 2 may be used.

  The dark-side reference brightness detection member 3 and the light-side reference brightness detection member 4 are used for performing calibration on the brightness of the image data of the sheet material A output from the line sensor camera 6.

  That is, when the sheet material A is conveyed using the plurality of rollers 2, the sheet material A is conveyed in the conveying direction in a undulating state, and the surface of the sheet material A fluctuates in the vertical direction. On the other hand, the brightness of the image data of the sheet material A output from the line sensor camera 6 varies depending on the distance from the line sensor camera 6 to the sheet material A at the imaging timing. Therefore, when calibration is not performed, the brightness of the image data of the sheet material A output from the line sensor camera 6 varies every scan, and the detection of defects becomes extremely difficult.

  On the other hand, the defect inspection apparatus 1 according to the present embodiment uses the brightness of the image data of the sheet material A as the brightness of the image data of the dark-side reference brightness detection member 3 output from the line sensor camera 6 during the same scan. Since the calibration is performed based on the brightness of the image data of the light side reference brightness detection member 4, the change in brightness of the image data of the sheet material A caused by the vertical movement of the surface of the sheet material A can be canceled, and the defect Detection can be made easy.

  It is possible to store the lightness data for calibration in the defect detection processing unit 7 without the dark side reference brightness detection member 3 and the light side reference brightness detection member 4, but in this case, Accurate calibration cannot be performed when the inspection environment changes such that the illumination device 5 deteriorates with time. On the other hand, since the defect inspection apparatus 1 according to the present embodiment includes the dark side reference brightness detection member 3 and the light side reference brightness detection member 4 on the roller 2, an inspection such that the illumination device 5 deteriorates with time is provided. Accurate calibration can be performed even when the environment changes.

  As the dark-side reference brightness detection member 3, one having an average brightness lower by a preset margin value than the darkest of the various sheet materials A that are inspection objects is used. Moreover, the light side reference brightness detection member 4 uses a member whose average brightness is higher by a preset margin value than the brightest of the various sheet materials A that are inspection objects. Thus, since the defect inspection apparatus 1 according to the present embodiment considers the margin values for the average brightness of the dark-side reference brightness detection member 3 and the average brightness of the light-side reference brightness detection member 4, changes in the inspection environment Therefore, a stable defect inspection result can be obtained.

  The outer diameter of the roller 2, the outer diameter of the dark side reference brightness detecting member 3, and the outer diameter of the light side reference brightness detecting member 4 are the brightness of the image data of the sheet material A output from the line sensor camera 6 for each scan, In order to make the brightness of the image data of the dark side reference brightness detection member 3 and the brightness of the image data of the light side reference brightness detection member 4 uniform, it is desirable to have the same diameter.

  The illumination device 5 is disposed above the roller 2 with the sheet material A interposed therebetween as shown in FIG. Further, as shown in FIG. 2, the illumination device 5 is arranged in a direction orthogonal to the conveyance direction of the sheet material A, and the sheet material A conveyed by the roller 2, the dark side reference brightness detection member 3, and the light side. The reference lightness detection member 4 has a length that can be illuminated in a series of lines. As a light source of the illuminating device 5, LED (light emitting diode), organic EL (electroluminescence), a fluorescent lamp, etc. can be mentioned.

  The line sensor camera 6 images the surface of the sheet material A conveyed by the roller 2, the surface of the dark side reference brightness detection member 3, and the surface of the light side reference brightness detection member 4 in one scan, and an image corresponding to the captured image Output data. The line sensor camera 6 applied to the defect inspection apparatus according to the present embodiment may be a color line sensor camera or a monochrome line sensor camera. As the line sensor camera 6 applied to the defect inspection apparatus according to this embodiment, a CCD line sensor camera is preferably used because it is relatively inexpensive and has high reliability.

  As shown in FIG. 1, the illumination position of the illumination device 5 and the imaging position of the line sensor camera 6 have less vertical movement of the sheet material A accompanying the rotational driving of the roller 2 and can reduce the change in brightness of image data for each scan. In addition, it is desirable to use the top side portion of the roller 2 provided with the dark side reference brightness detection member 3 and the light side reference brightness detection member 4.

  The defect detection processing unit 7 is realized by a computer having a hardware configuration as shown in FIG. As is apparent from this figure, the defect detection processing unit 7 according to the present embodiment includes a CPU (central processing unit) 11, a memory 12, a hard disk 13, a general-purpose external device interface 14, a camera interface 15, and a graphic board 16. It is configured. A keyboard 21, a mouse 22, and a LAN (local area network) 23 are connected to the general-purpose external device interface 14. The line sensor camera 6 is connected to the camera interface 15. Further, a display device 24 such as a liquid crystal display device is connected to the graphic board 16.

  The CPU 11 is a calculation means and controls the operation of the entire defect inspection apparatus 1.

  The memory 12 includes a RAM and a ROM (not shown). The RAM is a volatile storage medium capable of reading and writing information at high speed, and is used as a work area when the CPU 11 processes information. The ROM is a read-only nonvolatile storage medium, and temporarily holds programs such as firmware and various data necessary for the defect inspection apparatus 1 to execute the defect inspection processing. The hard disk 13 is a non-volatile storage medium capable of reading and writing information, and stores an OS (basic software), various control programs, application programs, and the like.

  The data required when the defect inspection apparatus 1 executes the defect inspection process includes area data of the sheet material A, the dark side reference brightness detection member 3 and the light side reference brightness detection member 4, and a reference when executing the brightness extension. The minimum value and the maximum value of the dynamic range of the signal handled by the defect inspection apparatus 1 can be given. Other data includes defect width and area for specifying the type of defect, data on the side of a circumscribed square, the coordinates of the center of gravity point, etc., the defect criteria for the sheet material A required by the defect inspection apparatus 1 Can be mentioned. The area data can be specified by the coordinates of both ends of the sheet material A, the coordinates of both ends of the dark side reference brightness detection member 3, and the coordinates of both ends of the light side reference brightness detection member 4.

  The CPU 11 implements control and functions of the entire defect inspection apparatus 1 by reading necessary programs, data, setting information, and the like from the ROM and the hard disk 13 on the RAM and executing the processing.

  FIG. 4 shows an example of an image generated based on the image data output from the line sensor camera 6. This image is obtained by driving the illuminating device 5 and the line sensor camera 6 while rotating the roller 2 to convey the sheet material A at a constant speed, and from the outer end portion of the dark side reference brightness detection member 3 to the light side reference. The lightness detection member 4 is repeatedly scanned up to the outer end portion at a predetermined shutter timing, and a linear image obtained for each scan is continuously arranged in the conveying direction of the sheet material A to form a planar image.

  In the planar image of FIG. 4, the lowest (dark) portion of the brightness is an image of the dark side reference brightness detection member 3, and the brightness corresponds to the average brightness of the dark side reference brightness detection member 3. The portion with the highest (bright) brightness is an image of the light-side reference brightness detection member 4, and the brightness corresponds to the average brightness of the light-side reference brightness detection member 4. Further, the portion having the intermediate brightness is an image of the sheet material A, and the brightness corresponds to a change in brightness in the scanning direction of the line sensor camera 6.

  The brightness of the image of the sheet material A is uniform over the entire surface of the image when there is no defect on the surface of the sheet material A, and the line sensor camera 6 scans when there is a defect on the surface of the sheet material A. A non-uniform part is produced regarding the direction and the conveyance direction of the sheet material A. Therefore, the defect inspection apparatus 1 can detect whether or not a defect has occurred on the surface of the sheet material A by monitoring the change in brightness of the image of the sheet material A. In the example of FIG. 4, it can be seen from the change in brightness of the image of the sheet material A that there is one linear defect D1, three point defects D2, and one spot-like defect D3 on the surface of the sheet material A. .

  However, as apparent from FIG. 4, the planar image generated based on the image data output from the line sensor camera 6 has a small difference in brightness between the portion with the lowest brightness and the portion with the highest brightness. Is difficult to detect with high accuracy. Such inconvenience becomes particularly noticeable when a defect occurring on the surface of the plain sheet material A is detected.

  Therefore, the defect inspection apparatus 1 according to the present embodiment expands the range of the brightness change of the sheet material A based on the average brightness of the dark side reference brightness detection member 3 and the average brightness of the light side reference brightness detection member 4. A surface defect of the sheet material A is detected based on a change in brightness of the sheet material A after extending the range.

  FIG. 5 shows a processing procedure of defect inspection performed by the defect inspection apparatus 1 according to the present embodiment.

  The CPU 11 starts driving the illumination device 5 and the line sensor camera 6 in a state where the defect inspection device 1 and the roller 2 are activated, and the light side reference brightness detection member 4 starts from the outer end portion of the dark side reference brightness detection member 3. Images up to the outer end are repeatedly taken at a predetermined shutter timing (step S1).

  Next, the CPU 11 stores the image data output from the line sensor camera in the memory 12 via the camera interface 15 and the CPU 11 (step S2).

  Next, the CPU 11 determines the area of the image data output from the line sensor camera based on the area data stored in advance in the memory 12 (step S3), and determines the area of the image data output from the line sensor camera as the sheet portion. Divide into dark reference part and light reference part. The area determination of the image data is performed by determining the coordinates of both end portions of the dark side reference brightness detecting member 3 stored in advance in the memory 12, the coordinates of both end portions of the light side reference brightness detecting member 4, and both end portions of the sheet material A. Based on the coordinates of

  For the area determined to be the dark reference portion in step S3, the process proceeds to step S4, the average brightness of the dark side reference brightness detection member 3 is obtained, and the obtained average brightness of the dark side reference brightness detection member 3 is obtained. Is set as the dark side reference value (step S5). The set dark side reference value is stored in the memory 12.

  On the other hand, for the region determined to be the light reference portion in step S3, the process proceeds to step S6, the average lightness of the light side reference lightness detection member 4 is obtained, and the obtained light side reference lightness detection member 4 is obtained. The average brightness is set as the light side reference value (step S7). The set write side reference value is stored in the memory 12.

  Next, the CPU 11 proceeds to step S8, compares the brightness of the sheet portion, dark reference portion, and light reference portion determined in step S3, and is set in the dark side reference value set in step S5 and step S7. Using the light side reference value, a brightness expansion process is performed on the image of the sheet material A to perform brightness comparison with higher accuracy. A specific method of brightness enhancement processing will be described later with reference to FIGS.

  Next, the CPU 11 proceeds to step S <b> 9, performs binarization processing on the image of the sheet material A whose brightness is expanded, and binarizes the difference in brightness. Thereby, the part in which the brightness has changed compared with the ground part can be detected as a defect. Depending on the type of defect, it is desirable to perform ternary processing instead of binarization processing when the brightness is higher or lower than the ground portion.

  Next, the CPU 11 proceeds to step S10 and performs a labeling process on the binarized or ternarized image of the sheet material A. The labeling process is a process of detecting adjacent connections of individual pixels determined to be defective parts by binarization or ternarization, finding isolated defect masses, assigning numbers (labels), and defining them as defective parts It is. Next, the CPU 11 calculates the length, width, area, side length of the circumscribed square, barycentric coordinates and the like for each labeled defect, and lists them as the characteristics of each defect.

  Next, the CPU 11 proceeds to step S11, compares the various characteristics of each defect label with a standard according to the request, and performs a defect determination process for determining whether the defect should be eliminated or overlooked.

  Next, the CPU 11 proceeds to step S12 and compares the various characteristics of each defect label with the defect criteria of the sheet material A stored in the memory 12 to determine whether the defect should be eliminated or missed. The defect determination process is performed. The result of this defect determination processing is stored in the memory 12 and used for final processing (disposition, excision, repair, etc.) in a later process.

  Note that there are no particular limitations on the specific methods of the binarization (ternarization) processing, labeling processing, and defect type determination processing, and any method that is publicly known can be used.

  Next, a specific method of the brightness expansion process performed in step S8 will be described.

  FIG. 6 shows a first example of brightness extension processing. FIG. 6A shows the brightness change of image data for one scan from the outer end of the dark side reference brightness detecting member 3 to the outer end of the light side reference brightness detecting member 4 captured by the line sensor camera 6. Is shown. As is apparent from this figure, the brightness change of the image data for one scan has the lowest brightness of the dark side reference brightness detection member 3 and the highest brightness of the light side reference brightness detection member 4. Further, the lightness of the sheet material A is in the middle, and the high lightness portion and the low lightness portion appear continuously in the scanning direction. There is a possibility that these high-lightness portions and low-lightness portions are detected defects.

  In FIG. 6A, the entire area of the brightness change is the dynamic range of the brightness of the line sensor camera 6. For example, black and white reflected in the human eye are the dynamic range of the entire area. As can be seen from FIG. 6A, in the state where the brightness expansion process is not performed, the distribution of brightness including defects of the sheet material A is smaller than that in the entire dynamic range. It is difficult to do accurately.

  FIG. 6B shows a change in brightness of image data for one scan after the brightness extension process is performed. After performing the brightness expansion process, the vertical axis is not the dynamic range of the line sensor camera 6 but the dynamic range of the data range handled by the defect detection processing unit 7 as shown in FIG. For example, when the data input range of the defect detection processing unit 7 is 8 bits, the minimum value of the dynamic range is “0” and the maximum value is “255”.

  Therefore, the brightness of the dark side reference brightness detection member 3 shown in FIG. 6A is set to “0”, which is the minimum value of the dynamic range of the defect detection processing unit 7, and the brightness of the write side reference brightness detection member 4 is set. Is set to “255” which is the maximum value of the dynamic range of the defect detection processing unit 7. Then, the brightness of the sheet material A is expanded and converted linearly or nonlinearly in a form corresponding to the brightness. Thereby, the range of the brightness of the sheet material A is expanded.

  FIG. 7 shows a second example of brightness extension processing. The second example of the brightness extension process is characterized in that the brightness extension process is performed by a technique using gamma conversion. Gamma conversion here is gamma conversion in a broad sense and refers to so-called one-to-one conversion of an output value for an input value linearly or nonlinearly.

  The horizontal axis of FIG. 7 is an input value (pixel value of pixel data), and the vertical axis is an output value corresponding to the input value. In the example of FIG. 7, the lightness of the dark side reference lightness detection member 3 and the lightness of the light side reference lightness detection member 4 are connected by a straight line, and the entire dynamic range of the output is linearly mapped therebetween.

  In other words, in this example, the output from “0” to the lightness of the dark side reference lightness detection member 3 is converted to “0”. Further, regarding the input from the lightness of the light side reference lightness detection member 4 to “255”, the output is converted to “255”. The input value in the meantime is linearly associated with the range from “0” to “255” from the lightness of the light-side reference lightness detection member 4 to the lightness of the dark-side reference lightness detection member 3. Thereby, the range of the brightness of the sheet material A is expanded.

  When the brightness range of the sheet material A is expanded as described above, the contrast of the image of the sheet material A is enhanced as shown in FIG. 8, so that it is easy to determine the presence or absence of a defect, the size of the defect, and the like. Become. In addition, smaller defects can be detected.

  The dark side reference brightness detection member 3 and the light side reference brightness detection member 4 may be configured to be detachable from the roller 2 as shown in FIG.

  That is, the minimum value and the maximum value of the brightness distribution of the sheet material A to be inspected may change depending on the type of the sheet material A or the change of the lot. Therefore, when the dark side reference brightness detection member 3 and the light side reference brightness detection member 4 are configured to be attachable to and detachable from the roller 2, the dark side reference brightness detection member 3 and the light side reference brightness detection member 4 are replaced with the sheet material A. Therefore, it is possible to appropriately perform the process of extending the brightness of the sheet material A.

DESCRIPTION OF SYMBOLS 1 Defect inspection apparatus 2 Roller 3 Dark side standard brightness detection member 4 Light side standard brightness detection member 5 Illumination device 6 Line sensor camera 7 Defect detection process part 11 CPU
12 memory 13 hard disk 14 general-purpose external device interface 15 camera interface 16 graphic board 21 keyboard 22 mouse 23 LAN
24 Display device A Sheet material (inspection object)
D1-D3 defect

Japanese Patent No. 4028654

Claims (10)

A roller for conveying a sheet material to be inspected;
A dark side reference brightness detection member and a light side reference brightness detection member attached to the roller;
A line sensor camera that images the sheet material conveyed by the roller, the dark side reference brightness detection member, and the light side reference brightness detection member for each scan;
A defect detection processing unit that performs image processing on image data output from the line sensor camera and detects the presence or absence of defects on the surface of the sheet material;
With
The defect detection processing unit obtains an average brightness of the dark side reference brightness detection member, an average brightness of the light side reference brightness detection member, and a change in brightness of the sheet material in the scanning direction of the line sensor camera from the image data, Based on the obtained average brightness of the dark side reference brightness detection member and the average brightness of the light side reference brightness detection member, the range of the brightness change of the sheet material is expanded, and the sheet material after the range is expanded A defect inspection apparatus for detecting a surface defect of the sheet material based on a change in brightness.   The defect detection processing unit replaces the average brightness of the dark-side reference brightness detection member with a minimum value of a dynamic range of a signal handled by the defect detection processing unit, and detects the average brightness of the light-side reference brightness detection member as the defect detection The defect inspection apparatus according to claim 1, wherein the range of brightness change of the sheet material is expanded by replacing with a maximum value of a dynamic range of a signal handled by the processing unit.   The defect detection processing unit creates a gamma conversion table by linearly connecting the average brightness of the dark side reference brightness detection member and the average brightness of the light side reference brightness detection member, and an input value is the dark side reference brightness detection If it is less than the average brightness of the member, the output value is the minimum value of the dynamic range of the signal handled by the defect detection processing unit, and if the input value is greater than the average brightness of the light side reference brightness detection member, The defect inspection apparatus according to claim 1, wherein an output value is set to a maximum value of a dynamic range of a signal handled by the defect detection processing unit, thereby extending a range of brightness change of the sheet material.   The dark-side reference brightness detection member uses an average brightness lower by a preset margin value than the darkest of the various sheet materials that are inspection objects, and the light-side reference brightness detection member The defect inspection apparatus according to claim 1, wherein an average brightness is set higher by a preset margin value than a brightest one of the various sheet materials that are inspection objects.   The defect inspection apparatus according to claim 1, wherein the dark side reference brightness detection member and the light side reference brightness detection member are detachably attached to the roller.   6. The defect inspection according to claim 1, wherein a diameter of the dark-side reference brightness detection member, a diameter of the light-side reference brightness detection member, and a diameter of the roller are the same. apparatus.   The defect inspection apparatus according to claim 1, wherein the line sensor camera images a portion of the surface of the sheet material that contacts the roller. A first line sensor camera that images the sheet material conveyed by the roller and the dark side reference brightness detection member, and a second line imager that images the sheet material and light side reference brightness detection member conveyed by the roller. A plurality of line sensor cameras including at least a line sensor camera,
The defect detection processing unit obtains an average brightness of the dark-side reference brightness detection member and a change in brightness of the sheet material in the scanning direction of the dark-side line sensor camera from image data output from the first line sensor camera. In addition, the average brightness of the light-side reference brightness detection member and the brightness change of the sheet material in the scanning direction of the second line sensor camera are obtained from the image data output from the second line sensor camera. Based on the average brightness of the dark side reference brightness detection member and the average brightness of the light side reference brightness detection member, the range of the brightness change of the sheet material is expanded, and the brightness change of the sheet material after the range is expanded The surface defect of the sheet material is detected based on any one of claims 1 to 7. Defect inspection apparatus according. A roller for conveying a sheet material as an inspection object, a dark side reference brightness detection member and a light side reference brightness detection member attached to the roller, the sheet material conveyed by the roller, and the dark side reference brightness detection A line sensor camera that images the member and the light side reference brightness detection member every scan, and a defect that detects the presence or absence of a defect on the surface of the sheet material by performing image processing on image data output from the line sensor camera A defect inspection method using a defect inspection apparatus comprising a detection processing unit,
The defect detection processing unit obtains an average brightness of the dark side reference brightness detection member, an average brightness of the light side reference brightness detection member, and a change in brightness of the sheet material in the scanning direction of the line sensor camera from the image data. And expanding the range of the change in brightness of the sheet material based on the average brightness of the dark side reference brightness detection member and the average brightness of the light side reference brightness detection member obtained in advance. And a step of detecting a surface defect of the sheet material on the basis of a change in brightness of the sheet material later. A roller for conveying a sheet material as an inspection object, a dark side reference brightness detection member and a light side reference brightness detection member attached to the roller, the sheet material conveyed by the roller, and the dark side reference brightness detection A line sensor camera that images the member and the light side reference brightness detection member every scan, and a defect that detects the presence or absence of a defect on the surface of the sheet material by performing image processing on image data output from the line sensor camera A defect inspection apparatus including a detection processing unit,
Obtaining the average brightness of the dark side reference brightness detection member, the average brightness of the light side reference brightness detection member, and the change in brightness of the sheet material in the scanning direction of the line sensor camera from the image data, Extending the range of brightness change of the sheet material based on the average brightness of the dark side reference brightness detection member and the average brightness of the light side reference brightness detection member, and the brightness of the sheet material after extending the range And a step of detecting a surface defect of the sheet material based on a change.