JP2013200157A - Method and apparatus for inspecting end face - Google Patents

Abstract

PROBLEM TO BE SOLVED: To highly accurately find a defective site generated in a core material of a laminate in a conveyance process, from an end face.SOLUTION: The method for inspecting end faces includes: photographing an end face of an electrode plate as a laminate; obtaining a region of the electrode plate from the acquired image; obtaining a region of a core material from the region of the electrode plate; expanding/contracting the image along the longitudinal direction of the core material by a prescribed pixel unit to linearly complement the image of the core material; thinning the obtained linear core material image to obtain a skeleton of the core material; obtaining a difference of the skeleton image from the image in the region of the core material before the complement; comparing a measured distance in the longitudinal direction of the image with a reference distance for identifying a predetermined adhesion to find a defective site of the core material.

Description

  The present invention relates to an end face inspection method and an end face inspection apparatus for inspecting an end face of a laminate formed by applying a coating substance to a core material such as a metal foil or a metal plate.

  Conventionally, an electrode plate of a secondary battery is known as a laminate in which an active material is applied to a core material. In the process of applying an active material to a metal foil or metal plate and manufacturing an electrode plate cut into a predetermined sheet, defects such as chips and burrs generated on the end face of the electrode plate are inspected. Specifically, a single electrode plate is placed on a background color tone plate that emits light, and the surface of the electrode plate is photographed with a camera while irradiating light toward the surface of the electrode plate with a lighting device. To do.

  The obtained image data is subjected to a differential filter to differentiate the change in the degree of reflection of light from the background color tone plate, the core material and the active material, and the contour of the electrode plate is emphasized by the change in the gray level in the image data. Thereafter, noise processing is performed to compare the contour extraction data with a predetermined value. The chipping and burrs of the electrode plate are detected depending on whether or not the contour extraction data is within the predetermined value range (Patent Document 1).

Japanese Patent Laid-Open No. 10-40383

  Until the electrode plate is cut into a sheet of a predetermined shape, the long electrode plate is slit while being conveyed from the roll to the roll, or the electrode plate after the slit is conveyed from the roll to the roll at a high speed. .

  Therefore, the conventional inspection method functions effectively to detect burrs protruding from the width direction when a planar electrode plate is viewed in plan rather than a long electrode plate. However, there is a problem that defective products are mixed even if the electrode plate is inspected for defects by this method.

  Therefore, as a result of intensive studies to solve the problem, the following new findings were obtained. That is, when a long and wide laminate is slit to a predetermined width, the active material dust applied to the core material adheres to the cut end face, or a part of the active material near the cut end face is deformed. It was newly found that there is a defective part that can be discriminated only from the cut end face side.

  If the dust adhesion part does not remove the dust, the positive and negative electrode plates alternately stacked via separators inside the battery when the product is completed may be short-circuited by the dust or contaminated inside the battery. I found out that However, the dust can be used as a non-defective product if it is removed unlike the defective part. Therefore, it was found that it is important to discriminate between the dust adhering to the electrode end face and the defective part and to detect only the defective part.

  Further, it has been found that some deformation of the active material in the vicinity of the cut end face is caused by deterioration such as wear of the blade for slitting or poor conditions such as the position of the blade. If the slit is continued in a state where the defect occurs, not only the quality is deteriorated due to the frequent occurrence of the defect, but also the wear of the blade is advanced, and there is a possibility that the blade is damaged. Therefore, it was found that it is important to detect this defect in order to perform slitting stably with high quality.

  In defect inspection using conventional image processing, an image of a normal sheet-shaped laminate is previously learned, and a comparison process is performed by pattern matching with the normal image for each captured image to detect a defective portion. The method is also often used. However, in the process of slitting the laminated body to be inspected, the end face image of the laminated body obtained by the material and thickness of the core material and coating material of the laminated body, or the change in tension applied to the laminated body at the time of slitting is Contains irregular and minute bends and distortions. These small bends include those that do not reach defects. Therefore, in the comparison process with the normal image, there is a problem that a non-defective part is erroneously determined as a defective part.

  In the process of slitting the laminated body, a slit blade is installed at a position where the laminated body is cut to a specified width during conveyance of the laminated body fed from the roll, and the laminated body cut by the blade is wound up. I'm taking it. In this case, the long laminate is more susceptible to changes in tension during the transport process than handling a single-wafer laminate. Therefore, in the case of a long laminated body, the problem of erroneously determining a non-defective part as a defective part as in the case of a sheet-like laminated body occurs remarkably.

  Furthermore, the conveyance speed at the time of slitting the laminate is, for example, 50 m / min. Therefore, in order to perform defect inspection by image processing following such high-speed conveyance, defect detection processing can catch up with conventional complicated filtering processing such as differential filter processing and noise processing and high calculation load. There is no problem.

  The present invention has been made in view of such circumstances, and the deposits adhering to the end surface of the core material or the laminate in which the coating substance is applied to the core material and the defect site occurring on the end surface are obtained. The main object is to provide an end face inspection method and an end face inspection apparatus that can accurately discriminate in a short time.

In order to achieve such an object, the present invention has the following configuration.
That is, an end face inspection method for inspecting an end face of a long laminate formed by applying a coating substance to a core material,
A shooting process of shooting the end face of the laminate;
A first region extraction step of obtaining a region of the laminate from the image data acquired in the photographing step;
A second region extraction process for obtaining a core region from the laminate region;
A complementing process for complementing the core material image in a linear form by expanding and contracting the image in predetermined pixel units along the longitudinal direction of the core material;
A skeleton extraction process for obtaining a skeleton of the core material from the linear core material image obtained by the complementary processing;
A subtraction process for obtaining a difference between the skeleton image obtained in the skeleton extraction process from the image obtained in the first area extraction process;
A defect calculation process for obtaining a defect portion of the core by comparing a measured distance in the longitudinal direction of the image obtained in the subtraction process with a reference distance for specifying a predetermined deposit,
It is provided with.

  (Function / Effect) According to this method, the image of the core material obtained from the captured image is expanded and contracted only in the longitudinal direction, so that the divided core image is combined and complemented linearly. Is done. By thinning the complemented linear core material image, an ideal shape of the core material that should originally exist is obtained. By taking the difference between the skeleton images extracted from the image including the fragmented portion indicating the core region obtained in the first region extraction process, the candidate for the defective portion is extracted. By comparing the actual measured distance in the longitudinal direction of the candidate with the reference distance, only the defective portion excluding the deposit can be obtained.

  Therefore, unlike the conventional method, it is possible to accurately detect only the defective part occurring on the end face of the core material in a short time with processing with low calculation load without performing complicated calculation processing such as differential filter processing and noise removal processing. can do.

  In this embodiment, in the complementing process, the expansion and contraction process may be performed once using a plurality of pixel units that can be complemented based on the occurrence history of the defective part, and the predetermined number of times may be set. You may repeat. What is necessary is just to change a setting suitably according to the incidence rate and calculation load of the defective part which changes with materials, such as a core material to be used and a coating substance.

  In the skeleton extraction process, it is preferable that the skeleton of the core material obtains a center pixel from a plurality of pixels in the thickness direction of the core material image. Moreover, when the thickness of the core material image of the center extraction region is composed of even-numbered pixels, it is preferable to select one of the upper and lower pixels according to the inclination of the core material.

  According to this embodiment, depending on the degree of inclination of the core material, even if it is difficult to reproduce the extraction of the ideal core material skeleton in which images are continuously connected by the image expansion and contraction processing, It is possible to accurately reproduce an ideal skeleton in which pixels are connected.

The present invention has the following configuration in order to achieve such an object.
That is, an end surface inspection device that inspects an end surface of a long laminate formed by applying a coating substance to a core material,
An irradiation unit for irradiating light toward an end face of the laminate;
A photographing unit for photographing the end surface of the laminate irradiated with light;
An image processing unit that inspects a defective portion of the core material using image data acquired by the imaging unit,
The image processing unit includes a storage unit that stores image data acquired by the photographing unit;
A region extraction unit for obtaining a region of the laminate from the image data acquired in the photographing process;
A core material extraction unit for obtaining an area of the core material from the area of the laminate;
A complementary processing unit that expands and contracts the image in predetermined pixel units along the longitudinal direction of the core material, and complements the core material image in a linear shape,
A skeleton extraction unit for obtaining a skeleton from a linear core material image by the complementary processing;
The difference between the skeleton images obtained by the skeleton extraction unit is taken from the image of the laminate obtained by the region extraction unit, the measured distance in the longitudinal direction of the difference image is obtained, and the measured distance is compared with a predetermined reference distance. A defect calculation unit for obtaining a defect part of the core material;
It is provided with.

  According to this configuration, the above method can be suitably implemented.

  According to the end face inspection method and the end face inspection apparatus of the present invention, it is possible to accurately detect only a defective portion in a short time except for the attached matter attached to the end face of the core material.

It is a schematic block diagram of the end surface inspection apparatus which concerns on a present Example. It is AA arrow sectional drawing of FIG. 1 which shows the end surface test | inspection of an electrode plate. It is a block diagram which shows schematic structure of an Example apparatus. It is a flowchart which shows the test | inspection process by an Example apparatus. It is the image which imaged the end surface of the electrode plate. It is an image of the electrode plate after a binarization process. It is the image which emphasized the core material after a complementation process. It is the skeleton image of the core material obtained by thinning. It is the image which extracted the candidate of the defect site | part. It is a flowchart which shows the test | inspection process by a modification apparatus. It is a flowchart which shows the test | inspection process by a modification apparatus.

  An embodiment of the present invention will be described below with reference to the drawings. In this embodiment, as shown in FIG. 2, a positive electrode in which a coating material 5 (for example, an active material) is applied on both surfaces of a core material 4 used for a secondary battery (for example, a lithium ion battery) or the like. A negative electrode plate 3 is used. An apparatus for inspecting a defect portion such as a burr generated on the end face obtained by cutting the long electrode plate 3 wound around a bobbin into a predetermined width along the longitudinal direction will be described as an example.

  FIG. 1 is a diagram showing a schematic configuration of an end surface inspection apparatus according to an embodiment of the present invention.

  This end face inspection apparatus is composed of an optical unit 1, an image processing apparatus 2, and the like.

  The optical unit 1 includes an optical camera 7 disposed toward an end face of the electrode plate 3 to which a predetermined tension is applied by a guide roller 6, an irradiation unit 8 that irradiates light toward an imaging region of the electrode plate 3, and an electrode plate 3 is composed of a background plate 9 and the like disposed opposite to the optical camera 7.

  As the background plate 9, a white plate for reflection or a black plate that absorbs light is appropriately changed according to the object to be photographed. In this embodiment, a black plate is used. Note that, depending on the imaging object, the inspection may be performed without the background plate 9 being installed.

  As shown in FIG. 3, the image processing apparatus 2 includes an operation unit 10 for setting and inputting imaging conditions and the like, a display unit 11 for displaying a setting input image and a captured image, and a defective portion generated on the end face of the electrode plate 3. The image processing unit 12 for obtaining

  The image processing unit 12 includes a storage unit 13, a region extraction unit 14, a core material extraction unit 15, a complement processing unit 16, a skeleton extraction unit 17, and a defect site calculation unit 18. Note that the function of each part constituting the image processing unit 12 will be described in detail in the following description of the operation of the embodiment apparatus using the flowchart shown in FIG.

<Step S1> Condition setting The total length and width of the electrode plate 3 to be inspected, the thicknesses of the core material 4 and the coating substance 5 are selected from the operation unit 10 or information displayed on the display unit 11 is selected from a touch panel or the like. The In addition, a reference value and a set value used for determining a defective portion of the core material 4 and the coating material 5 are set and inputted in advance. These setting conditions are stored in the storage unit 13.

<Step S2> Image acquisition The condition setting is completed, and the optical camera 7 starts photographing the end face of the electrode plate 3 simultaneously with the conveyance of the electrode plate 3. The image data of the captured image shown in FIG. 5 is stored in the storage unit 13 in real time. Here, in FIG. 5, the horizontal direction is the longitudinal direction, and the vertical direction is the width direction. In the subsequent FIGS. 6 to 9, the direction is set under the same conditions. Step S2 corresponds to the imaging process of the present invention.

<Step S <b>3> First Region Extraction Process The region extraction unit 14 obtains a region of the coating substance 5. For example, the photographing objects displayed in the image are the background plate 9 and the electrode plate 3. The electrode plate 3 includes a core material 4 and a coating material 5. These three materials have different reflectivities. That is, the brightness of the core material 4 that is a metal foil is the highest, and the brightness decreases in the order of the coating substance 5 and the background plate 9. Therefore, in the binarized image data, a reference value is obtained in advance by an experiment or the like corresponding to the density level of each substance in the range of 0 to 255 gradations, and the obtained actual image data and the reference value are obtained. In comparison, as shown in FIG. 6, the region of the electrode plate 3 is obtained. Here, the portions of the electrode plate 3 and the core material 4 are highlighted by the binarization process. Note that step S3 corresponds to the first region extraction process of the present invention.

<Step S4> Core Material Extraction Processing The core material extraction unit 15 extracts the region of the core material 4 that is highlighted with the highest luminance from the image after the binarization processing. Step S4 corresponds to the second region extraction process of the present invention.

<Step S5> Complement Processing The complement processing section 15 performs the closing processing of expanding and contracting only the longitudinal direction of the core material 4 on the extracted region pixels of the core material 4 only once. Here, the pixel unit to be subjected to the closing process is a value set in advance from the size of the adhering material including dust of the coating substance 5 adhering to the end face of the core material 4 obtained from the history of the experiment and the manufacturing process, for example, The maximum length and average value of the deposits are set as appropriate. Therefore, by the closing process, the pixel of the defective part candidate 19 having a lower luminance than the core material 4 in the area of the core material 4 is changed to the same luminance as the core material 4 as shown in FIG. An image of the linear core material 4 having a width corresponding to a predetermined pixel connected to is obtained. Step S5 corresponds to the complementing process of the present invention.

<Step S6> Core Material Skeleton Extraction Processing As shown in FIG. 8, the skeleton extraction unit 17 obtains a center pixel along the longitudinal direction of the complemented core material 4 image. That is, the image of the skeleton 4A of the thinned core material 4 connected by one pixel is obtained from the image of the core material 4 connected linearly. Here, when the width of the core material 4 is composed of even-numbered pixels, the inclination of the core material 4 before and after the longitudinal direction of the pixel is obtained, and either one of the upper and lower pixels is selected according to the inclination. Step S6 corresponds to the skeleton extraction process of the present invention.

<Step S7> Subtraction Processing The difference between the image of the core material 4 area including the defective part candidate 19 obtained in step S3 and the thinned core material 4 image obtained in step S6 is obtained. By the subtraction process, only the image of the skeleton 4B of the defect candidate 19 is extracted as shown in the region surrounded by the wavy line in FIG. Step S7 corresponds to the skeleton extraction process of the present invention.

<Step S8> Defect site determination process 1
The defective part calculation unit 18 calculates the distance in the longitudinal direction of the defective part candidate 19 obtained by the subtraction process. The measured distance is compared with a predetermined reference distance. As a result of comparison, when the actually measured distance exceeds the reference distance, the part is determined as a defective part. If the measured distance is less than or equal to the reference distance, it is determined that it is not a defective part. In other words, when the measured distance is equal to or less than the reference distance, it can be estimated that the deposit is attached to the part.

<Step S9> Inspection end determination processing An encoder is mounted on a rotating shaft or the like on which a bobbin around which the reel-shaped electrode plate 3 is wound is mounted, and the length of the electrode plate 3 is calculated from the number of rotations detected by the encoder. Determine whether the set length has been reached. When the feeding length of the electrode plate 3 does not reach the set length, the processing from step S2 to step S8 is repeatedly executed. When the set length is reached, the series of processing ends.

  As described above, an image of the continuously connected linear core material 4 can be obtained by subjecting only the portion of the core material 4 to the longitudinal direction. By thinning the image of the core material 4 after the closing process, an ideal skeleton 4A of the core material 4 without defects can be obtained, and the skeleton 4A can be obtained from the image of the core material 4 obtained by the binarization process. Only the defect part candidate 19 can be extracted by subtracting these images. Furthermore, only the defective part can be detected by comparing the measured distance of the candidate 19 with the reference distance. Moreover, it can be estimated that the deposit | attachment, such as dust, has adhered to the site | part determined not to be a defect site | part.

  Therefore, as in the conventional method, it is possible to detect the defective portion of the end face that cannot be detected when the defective portion is detected by viewing the electrode plate 3 in plan view, and it is also possible to determine the presence or absence of the deposit. .

  Moreover, since the defective part is calculated | required by the low load calculation process by the subtraction of image data instead of the complicated calculation process which performs a differential filter process and a noise process like the conventional method, the strip | belt-shaped electrode plate conveyed at high speed 3 can immediately determine the defective part.

  Furthermore, the electrode plate 3 is not easily applied with an equal tension depending on the material of the core material 4 and the coating material 5 and tends to bend. Further, the core material itself may have minute distortions and irregularities. Reflected light is scattered by these minute deflections and irregularities, and a low-luminance part is generated in the core material 4 region. That is, it tends to be induced as a defective part in a pseudo manner. However, even in such a case, by reproducing the skeleton 4A of the core material 4, it is possible to avoid the influence of the induction of the pseudo defect site and detect the defect site.

  The present invention is not limited to the embodiment described above, and can be modified as follows.

  (1) In the above embodiment, the closing process is performed only once, but the same process may be repeated a plurality of times. In this case, the pixel corresponding to the minimum length of the deposit including the dust of the coating substance 5 adhered to the end face of the core material 4 obtained from the history of the experiment and the manufacturing process is set as a unit. After that, as shown in step S5 'of FIG. 10, every time the closing process is performed, it is determined whether or not the complement process has been completed. If the image of the core material 4 is not connected as a result of the determination, the process returns to step S5 and the closing process is repeated again. If the image of the core material 4 is connected, it will progress to step S6 and will perform the process of step S6 to step S8 similarly to the above-mentioned implementation.

  (2) In the above-described embodiment device, the defect portion that is within the thickness of the core material 4 is obtained. However, in addition to the defect portion, a defect portion protruding in the vertical direction from the coating substance 5 may be obtained together. Good.

  For example, the processing shown in the flowchart of FIG. 11 may be performed. In this case, steps S1 to S8 perform the same processing as in the above embodiment. This will be described in detail below from step S9, which is a feature of the method.

<Step S <b>9> Region Enhancement Processing The region extraction unit 14 performs at least one closing process for expanding and contracting the obtained region pixel of the coating substance 5 vertically and horizontally. At this time, pixels that are deficient due to unevenness of brightness caused by light diffusion due to unevenness or dust adhering at the time of cutting on the end surface of the coating substance 5 are complemented. Therefore, the area | region of the coating substance 5 connected continuously is emphasized. The region enhancement process may use a filtering process in addition to the closing process.

<Step S10> Rectangle Approximation Processing The region of the coating material 5 is approximated to a rectangle by the least square method using the image data of the region of the laminate obtained in step S2.

<Step S11> First Fitting Process The image of the rectangle and the area of the coating substance 5 are combined.

<Step S12> Extraction Processing of Defect Site of Coating Material The difference between the coating material 5 region and the rectangular region after the fitting process is taken to determine the defective site of the coating material 5.

<Step S <b>13> Second Fitting Process Using the image data of the core material 4 region, the region is approximated to a rectangle by the least square method, and is aligned with the core material 4 region.

<Step S <b>14> Extraction Process of Defect Site of Core Material The difference between the core material 4 region and the rectangular region after fitting is taken, and the defect site protruding in the vertical direction of the core material 4 is obtained.

<Step S15> Defect site determination process 2
A determination value for determining the shape of the defective part is calculated for the defective part obtained in step 14. The determination value is compared with a predetermined reference value, and when the determination value exceeds the reference value as a result of the comparison, the portion is determined as a defective portion. If the determination value is less than or equal to the reference value, it is determined that the defect portion is not present.

<Step S16> Inspection End Determination The same process as in the above embodiment is performed. That is, an encoder is mounted on a rotating shaft or the like on which a bobbin around which a reel-shaped electrode plate 3 is wound is mounted, the length of the electrode plate 3 is calculated from the number of rotations detected by the encoder, and the set length is reached. Determine whether or not. When the feeding length of the electrode plate 3 has not reached the set length, the processing from step S1 to step S14 is repeatedly executed. When the set length is reached, the series of processing ends.

  In the modification, the rectangle approximation process divides the image into a plurality of parts according to the degree of distortion of the electrode plate 3 displayed on one screen, performs the approximation process for each divided image, and starts from step S10 for each divided image. You may make it perform the process of step S13.

  (3) In the process for obtaining the defective part in the above modification, the process for obtaining the defective part of the core material 4 and the process for obtaining the defective part of the coating substance 5 are serially processed, but parallel processing may be performed. Specifically, it may be configured to branch from step S1 and perform each process in parallel.

  (4) Although the above-described embodiment device shows a configuration in which the optical camera 7 is provided only on one side of the electrode plate 3, it is preferably provided on both sides. In the case of this configuration, the image processing unit 12 may perform parallel processing or serial processing on the processing shown in FIG.

  (5) The above-described embodiment apparatus can be used for end face inspection of a laminated body cut into a single wafer in addition to a long laminated body.

  (6) In the above embodiment, the position information of the part of the defective part candidate 19 from which the attached substance is determined out of the defective part is stored in the storage unit 13, and the information is used in the subsequent process. You may perform the process which removes a kimono.

DESCRIPTION OF SYMBOLS 1 ... Optical unit 2 ... Image processing apparatus 3 ... Electrode plate 4 ... Core material 4A ... Skeleton 5 ... Application | coating substance 12 ... Image processing unit 13 ... Memory | storage part 14 ... Area | region extraction part 15 ... Core material extraction part 16 ... Complementary processing part 17 ... Skeletal extraction unit 18 ... Defective part calculation part 19 ... Defective part candidate

Claims (7)

An end face inspection method for inspecting an end face of a long laminate formed by applying a coating substance to a core material,
A shooting process of shooting the end face of the laminate;
A first region extraction step of obtaining a region of the laminate from the image data acquired in the photographing step;
A second region extraction process for obtaining a core region from the laminate region;
A complementing process for complementing the core material image in a linear form by expanding and contracting the image in predetermined pixel units along the longitudinal direction of the core material;
A skeleton extraction process for obtaining a skeleton of a core material by thinning the linear core material image obtained by the complementary processing;
A subtraction process for obtaining a difference between the skeleton image obtained in the skeleton extraction process from the image obtained in the first area extraction process;
A defect calculation process for obtaining a defect portion of the core by comparing a measured distance in the longitudinal direction of the image obtained in the subtraction process with a reference distance for specifying a predetermined deposit,
An end face inspection method characterized by comprising: The end face inspection method according to claim 1,
The end face inspection method is characterized in that the complementary processing step performs one expansion and contraction processing using a plurality of pixel units which can be complemented in advance based on a defect site occurrence history. The end face inspection method according to claim 1,
In the end face inspection method, the complementary processing step repeatedly performs expansion and contraction processing of a predetermined rank in units of a plurality of predetermined pixels. In the end surface inspection method according to claim 3,
In the complementing process, the expansion and contraction processing is performed until the core material images are continuous in a linear shape. In the end face inspection method according to any one of claims 1 to 4,
The end face inspection method, wherein in the skeleton extraction process, a center pixel is obtained from a plurality of pixels in a thickness direction of a core material image. The end face inspection method according to claim 5,
In the skeleton extraction process, when the thickness of the core material image of the center extraction part is composed of even-numbered pixels, one of the upper and lower pixels is selected according to the inclination of the core material. An end face inspection device for inspecting an end face of a long laminate formed by applying a coating substance to a core material,
An irradiation unit for irradiating light toward an end face of the laminate;
A photographing unit for photographing the end surface of the laminate irradiated with light;
An image processing unit that inspects a defective portion of the core material using image data acquired by the imaging unit,
The image processing unit includes a storage unit that stores image data acquired by the photographing unit;
A region extraction unit for obtaining a region of the laminate from the image data acquired in the photographing process;
A core material extraction unit for obtaining an area of the core material from the area of the laminate;
A complementary processing unit that expands and contracts the image in predetermined pixel units along the longitudinal direction of the core material, and complements the core material image in a linear shape,
A skeleton extraction unit for obtaining a skeleton from a linear core material image by the complementary processing;
The difference between the skeleton images obtained by the skeleton extraction unit is taken from the image of the laminate obtained by the region extraction unit, the measured distance in the longitudinal direction of the difference image is obtained, and the measured distance is compared with a predetermined reference distance. A defect calculation unit for obtaining a defect part of the core material;
An end surface inspection apparatus comprising:

Images