Yarn inspection device and selection method

Abstract

To eliminate or suppress fuzz during weaving processing of filament yarn.SOLUTION: An inspection device (100) of a filament yarn (X) is provided. The inspection device (100) includes irradiating means (10) to irradiate an inspection light to the filament yarn (X), measurement means (20) to measure shade and background brightness level of the filament yarn (X) with an area sensor, image processing means (30) to obtain image information based on information obtained from the area sensor, and determination means (40) to determine type and frequency of an error from the image information and determine an inspection result of the filament yarn (X) on the basis of a threshold corresponding to the error.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an apparatus and method for inspecting filament yarn, an apparatus and method for selecting raw yarn, and a method for weaving and processing filament yarn using the selected raw yarn. [Background technology]

[0002] Filament yarns are obtained by bundling a plurality of fiber monofilaments obtained by spinning raw materials such as glass, synthetic resin, natural resin, carbon, metal, etc. Filament yarns are used in the manufacture of various components such as laminates, printed wiring boards, stent grafts, composite materials, reinforcing materials, and concrete crack suppression materials. Among them, glass filament yarns are known as a raw material for printed wiring boards.

[0003] For example, tens to thousands of glass fiber monofilaments of several μm to several tens of μm obtained by spinning molten glass are bundled to form glass filament yarn, which is then temporarily wound on a drum to form a state called a cake. The glass filament yarn is then unwound from the cake and shipped in the form of glass yarn (glass thread) obtained by twisting the glass filament yarn. The glass thread is woven and crossed, opened and / or glued as necessary, composited and laminated with a matrix resin such as an epoxy resin, and processed into a printed wiring board.

[0004] Also, glass filaments may be shipped in the form of a bobbin unit in which the glass yarn is wound around a bobbin, a glass roving in which several dozen glass filament yarns are bundled and spun, or a chopped strand in which the glass filament yarn is cut into several mm to several tens of mm. The raw yarn from the bobbin unit can be used as a warp yarn and / or a weft yarn in a weaving process. Glass roving is used as a raw material for glass fiber reinforced plastics (FRP, FRTP), and chopped strands are used as a reinforcing material for thermoplastic resins.

[0005] In the process of bundling a plurality of fiber monofilaments into a filament thread, the process of unwinding the filament thread from the cake, the weaving process, or the processing process, the fiber monofilaments come into contact with each other and rub against each other, which may cause fluff defects in which the glass fiber monofilaments are partially cut or woven in the cut state. The fluff defects not only deteriorate the workability but also affect the quality of the final product. For example, when a glass cloth woven from glass filament threads is used for a printed wiring board, the presence of fluff on the surface of the glass filament threads may deteriorate the flatness of the printed wiring board and cause problems in the formation of copper foil. Conventionally, an inspection device and method for fluff that may occur in ultra-fine glass fiber monofilaments, which are the raw material for high-density printed wiring boards, have been considered (see, for example, Patent Document 1).

[0006] Patent Document 1 describes an inspection device for glass filament yarns, which are formed by bundling a plurality of ultrafine glass fiber monofilaments having a diameter of more than 3 μm and not more than 7 μm, comprising an irradiation means for irradiating an inspection light onto the glass filament yarn in progress, a measurement means for measuring the luminance level of the surface of the glass filament yarn in a line shape, an extraction means for extracting monofilaments extruding from the glass filament yarn based on the measured luminance level, an image processing means for converting the extraction result into a luminance level image of 256 gradations and binarizing it, and a judgment means for judging the monofilament to be defective when the amount of monofilament extruding in the binarized image is found to be equal to or greater than a predetermined threshold value, and also describes a method for inspecting glass filament yarns using the above-mentioned inspection device. [Prior art documents] [Patent documents]

[0007] [Patent Document 1] JP 2012-7248 A Summary of the Invention [Problem to be solved by the invention]

[0008] However, there is still room for improvement in suppressing or eliminating fuzz defects in the inspection device and method described in Patent Document 1. In addition, there is still room for improvement in the yield and stable mass production of woven products, as well as detailed determination of fuzz defects and investigation of their causes in the inspection device and method using a line sensor described in Patent Document 1.

[0009] In view of the above problems, an object of the present invention is to eliminate or suppress fuzz defects during weaving of filament yarns by utilizing a specific inspection device or method. [Means for solving the problem]

[0010] As a result of intensive research, the present inventors have found that the above-mentioned problems can be solved by performing specific inspections and selections at the stage of raw yarn before the weaving process, and have completed the present invention. (1) An inspection device for filament yarn, comprising: An irradiation means for irradiating the filament yarn with an inspection light; a measuring means for measuring the luminance level of the shadow of the filament yarn and the background obtained by irradiating the inspection light by the irradiating means with an area sensor; an image processing means for acquiring image information based on information obtained from the area sensor; and a judgment means for determining errors and a frequency of the errors from the image information, and judging an inspection result of the filament yarn based on a threshold value corresponding to the errors. (2) The inspection device according to item (1), wherein, when there are multiple types of errors, the determination means determines the inspection result of the filament yarn based on a threshold series corresponding to each error type. (3) The inspection device according to item (1) or (2), wherein when there are a plurality of thresholds, the judgment means outputs a judgment result of the image by dividing the thresholds into a first threshold or less, a first to second thresholds, and a second threshold or more. (4) The inspection device according to item (3), wherein, when there are three or more thresholds, the judgment means outputs a judgment result of the image by dividing the thresholds into those below the first threshold, those between the first and second thresholds, those between the second and third thresholds, and those above the third threshold. (5) The inspection device according to any one of (1) to (4), wherein the determining means determines the threshold value by combining the image information with a measurement result of a diameter of the filament yarn. (6) The inspection device according to any one of (1) to (5), wherein the image processing means corrects the image. (7) The inspection device according to any one of items (1) to (6), wherein the filament yarn is made of at least one selected from the group consisting of glass fiber, acrylic resin fiber, fiber containing acrylic resin and other synthetic resin, and carbon fiber. (8) The inspection device according to item (7), wherein the filament yarn is made of glass fiber. (9) The inspection device according to any one of items (1) to (8), further comprising a control unit that controls the irradiation unit, the measurement unit, the image processing unit, and the determination unit. (10) The inspection device according to any one of (1) to (9), wherein the judging means further judges whether the raw filament yarn is pass or fail based on the inspection result of the filament yarn. (11) The inspection device according to any one of items (1) to (10), wherein the filament yarn is made up of 10 to 500 single yarns having a single yarn diameter of 1 to 10 μm. (12) The inspection device according to any one of items (1) to (11), which is used offline. (13) The inspection device according to any one of items (1) to (11), which is used in-line. (14) A method for inspecting a filament yarn, comprising the steps of: an irradiation step of irradiating the filament yarn with an inspection light; a measuring step of measuring the luminance levels of the shadow of the filament yarn and the background obtained by irradiating the inspection light in the irradiating step with an area sensor; an image processing step for acquiring image information based on information obtained from the area sensor; a determining step of determining an error and a frequency of the error from the image information and determining an inspection result of the filament yarn based on a threshold value corresponding to the error; The inspection method includes: Effect of the Invention

[0011] According to the present invention, by utilizing a specific inspection device or method, it is possible to eliminate or suppress fuzz defects during weaving processing of filament yarn, thereby achieving improved yields, stable mass production, efficiency, and high quality of woven products. [Brief description of the drawings]

[0012] [Figure 1] 1 is a perspective view showing a schematic diagram of an offline inspection apparatus according to an embodiment of the present invention; [Diagram 2] 1 is a perspective view showing a schematic diagram of an in-line inspection device according to an embodiment of the present invention; [Diagram 3] 1 is a schematic configuration diagram of an inspection device according to an embodiment of the present invention; [Figure 4] 1 is a flowchart of an inspection method according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] Below, an embodiment of the present invention (hereinafter referred to as "the present embodiment") will be described in detail with reference to the drawings. However, the present invention is not limited to the embodiment or drawings described below, and various modifications are possible without departing from the gist of the present invention.

[0014] [Filament yarn inspection device] (Inspection equipment use) The filament yarn inspection device according to this embodiment can be used for, for example, filament yarn inspection, raw yarn production, or raw yarn processing.

[0015] When the inspection device according to this embodiment is used in yarn production, it is possible to prevent or suppress the shipment or continued production of yarn with poor quality, and / or to achieve process management or process improvement by quickly identifying the cause of the poor quality yarn.

[0016] When the inspection device according to this embodiment is used in raw yarn processing, it is possible to prevent the use of raw yarn of unsuitable quality, thereby achieving quality control or quality improvement in the processing step.

[0017] The filament yarn inspection device according to this embodiment can be used, for example, offline or inline.

[0018] Specifically, the inline inspection device can be incorporated into the raw yarn production process (e.g., spinning bushing, sizing, winding, twisting process, etc.) or the raw yarn processing process (e.g., warping, weaving, post-processing, etc.), thereby achieving the effects of improving the quality of fuzz, etc., and increasing productivity in raw yarn production.

[0019] Specifically, offline inspection devices are used to inspect the filament yarn itself, or before the raw yarn is processed, and by selecting raw yarns of poor quality in advance, it is possible to recover the raw yarns of poor quality without using them, or to manage or improve the yarn quality.

[0020] (Inspection equipment configuration) 1 is a perspective view showing a schematic diagram of offline use of a filament yarn inspection device according to the present embodiment (hereinafter simply referred to as the "inspection device") 100. The inspection device 100 can store one or more bobbins of the filament yarn X to be inspected, and can receive the filament yarn X unwound from the bobbin into an inspection section 101 within the device via a stable delivery jig such as a roller, reel, or guide.

[0021] The filament yarn X to be inspected is drawn out from a bobbin installed in the inspection device 100 and introduced into the inspection section 101. The filament X inspected by the inspection device 100 can be wound onto a jig such as another bobbin, or further delivered to an external line via a stable delivery jig. In addition, the filament yarn X after the measurement can be provided for weaving processing, collected, reused, diverted to another use, or discarded.

[0022] As shown in Fig. 1, the inspection device 100 includes an inspection unit 101, which is an optical system that irradiates light onto the filament yarn X and measures the brightness level of the shadow of the filament yarn X and the background in a two-dimensional area, and a calculation unit equipped with a computer, an arithmetic processing device, a monitor, a memory unit, etc. for analyzing the measurement results. More specifically, the optical system includes an irradiation means 10 that irradiates the filament yarn X with inspection light, and a measurement means 20 that measures the brightness level of the shadow of the filament yarn X and the background obtained by irradiating the inspection light by the irradiation means 10 with an area sensor. The calculation unit includes an image processing means 30 that acquires image information based on information obtained from the area sensor, and a judgment means 40 that determines errors and error frequency from the image information and judges the inspection result of the filament X based on a threshold value according to the errors, and may further include a control means 50 that controls the irradiation means 10, the measurement means 20, the image processing means 30, and the judgment means 40 as desired.

[0023] 2 is a perspective view showing a schematic diagram of an in-line use of the inspection device according to the present embodiment. The in-line type inspection device includes an image processing means 30, a determination means 40, and a calculation unit including a control means 50 as required (all of which are omitted in FIG. 2).

[0024] In Fig. 2, the arrows indicate the flow direction of the filament yarn X to be inspected, and a yarn unwinding section (not shown) may be disposed upstream of the line (lower side of Fig. 2), and a yarn winding section (not shown) may be disposed downstream of the line (upper side of Fig. 2). The filament yarn X indicated by the solid arrow is inspected by a set of irradiation means 10 and measurement means 20 indicated by a solid line, the filament yarn X indicated by the dotted arrow is inspected by a set (10,20) indicated by a dotted line, the filament yarn X indicated by the dashed-dotted arrow is inspected by a set (10,20) indicated by a dashed-dotted line, and the filament yarn X indicated by the two-dot dash ... line arrow is inspected by a set (10,20) indicated by a dashed-dotted line, and each of these may be judged by a calculation section (not shown).

[0025] In the inspection device 100, the number of sets of the irradiation means 10 and the measurement means 20 and the number of filament yarns to be inspected can be arbitrarily set. For example, the number of sets of the irradiation means 10 and the measurement means 20 can be set to 1 to 100, 1 to 80, 1 to 60, 1 to 40, 1 to 20, 1 to 16, 1 to 12, 1 to 8, 1 to 7, 1 to 5, 1 to 4, 1 to 3, 1 to 2, or 1. and the number of filament yarns may be 1 to 500, 10 to 500, 1 to 400, 1 to 300, 1 to 250, 1 to 230, 1 to 200, 1 to 160, 1 to 130, 1 to 100, 1 to 80, 1 to 60, 1 to 40, 1 to 20, 1 to 18, 1 to 16, 1 to 14, 1 to 12, 1 to 10, 1 to 8, 1 to 6, 1 to 4, 1 to 3, 1 to 2, or 1. For example, in FIG. 1, two filament yarns X are inspected by two sets of irradiation means 10 and measurement means 20. In FIG. 2, four filament yarns X are inspected by one set of irradiation means 10 and measurement means 20, and a total of 16 filament yarns X are inspected by a total of four sets (10, 20).

[0026] The filament yarn X may be made of at least one selected from the group consisting of glass fiber, acrylic resin fiber, fiber containing acrylic resin and other synthetic resin, and carbon fiber. Among them, from the viewpoint of identifying the cause of fluff defects in woven products and reducing fluff defects, it is preferable that the filament yarn to be inspected is made of glass fiber.

[0027] The filament yarn may be made up of a plurality of single yarns, and from the viewpoint of improving the accuracy of the judgment by the calculation unit, the single yarn diameter is preferably 1 μm to 10 μm, and more preferably 3 μm to 7 μm. From the same viewpoint, the filament yarn is preferably made up of 10 to 500 single yarns having the above single yarn diameter.

[0028] 3 is a schematic diagram showing the configuration of the inspection device 100. The inspection device 100 includes an irradiation means 10 that irradiates the filament yarn X with inspection light; a measurement means 20 that measures the brightness levels of the shadow of the filament yarn X and the background with an area sensor; an image processing means 30 that acquires image information based on information obtained from the area sensor; and a judgment means 40 that determines errors and their frequency from the acquired image information and judges the inspection result of the filament yarn X based on a threshold value according to the errors. If desired, the inspection device 100 may further include a control means 50 that controls the irradiation means 10, the measurement means 20, the image processing means 30, and the judgment means 40.

[0029] The inspection unit 101 includes as its components an irradiation means 10 that irradiates light onto the advancing filament yarn X, and a measurement means 20 that uses an area sensor to measure the brightness levels of the shadow of the filament yarn X and the background. If desired, the inspection unit 101 may also include rollers, guides, reels, and the like as stable delivery jigs for stably advancing the filament yarn X.

[0030] (irradiation means) As shown in Figs. 1 to 3, the irradiation means 10 is configured as a single or multiple light sources that irradiate light onto the filament yarn X in progress. Although two light sources are shown in Fig. 1, four in Fig. 2, and one in Fig. 3, three or five or more light sources may be used as long as the measurement means 20 can measure a two-dimensional image of the brightness level. For example, an LED, a halogen lamp, or the like may be used as the light source, and among them, an LED is preferable. The irradiation means 10 is preferably configured to irradiate planar light so as to cooperate with the measurement means 20, and is preferably equipped with a strobe illumination that can be synchronized with the measurement means 20 in order to capture an image of an inspection object moving at high speed without blurring, and / or is preferably of a highly directional type that can grasp even thin fibers. The illumination used preferably has an output voltage of 24 V or more, and more preferably has an output voltage of 40 V or more.

[0031] (Method of measurement) As shown in Figs. 1 to 3, the measuring means 20 is configured as a camera including an area sensor that measures the luminance level of the shadow of the filament yarn X and the background in a two-dimensional area according to the irradiated light. The number of cameras can be determined according to the number of light sources. The area sensor included in the camera is configured to measure the luminance level of the shadow and the background that occurs when the filament yarn X blocks the irradiated light, so in this embodiment, the area where the irradiation range of the light source and the image receiving range of the area sensor overlap two-dimensionally is the inspection target area of ​​the filament yarn X. In this embodiment, direct area measurement by the area sensor makes it easier to set defect judgment criteria, improves defect inspection accuracy, and reduces the influence of positional deviation of the filament yarn X, compared to inspection by a conventional line sensor or two-dimensionalization by stacking conventional one-dimensional data.

[0032] The at least one light source and the at least one camera may be arranged with an optical axis offset as long as all measurement objects are measured by the area sensor under the same positional conditions, but from the viewpoint of accurate measurement of the filament yarn X within the inspection target area, it is preferable that the two are arranged to face each other across the filament yarn X. From the viewpoint of measurement accuracy in the inspection target area, it is preferable that the filament yarn X is delivered to the inspection target area at a speed of 80 to 1500 m / min and is imaged continuously and without gaps for each imaging range in the overall length direction.

[0033] The area sensor may comprise a CMOS image sensor, which may have a resolution of 5 μm or less per pixel, and may perform area measurements with dimensions of 15 mm length by 20 mm width per measurement.

[0034] As shown in Figs. 1 and 3, the irradiation means 10, the measurement means 20, the stable delivery jig, and the like are preferably installed within the inspection section 101 from the viewpoint of the accuracy of irradiation and measurement.

[0035] Data on the brightness levels of the shadow of the filament yarn X and the background measured by the area sensor is sent to the calculation unit. In the calculation unit, image processing, analysis, judgment, storage, machine learning, etc. of the data on the brightness level are automatically performed. The calculation unit can be realized by, for example, a general-purpose personal computer, a large computer, a calculation processing device, a GPU, a monitor, a memory, image processing software, an image processing application, artificial intelligence, etc. The calculation unit includes an image processing means 30 that acquires an image from the area sensor, and a judgment means 40 that determines errors and their frequency from the acquired image, stores a threshold value in memory according to the error, and judges the inspection result of the filament yarn X based on the stored threshold value. If desired, the inspection device 100 or its calculation unit may further include a control means 50 that controls the irradiation means 10, the measurement means 20, the image processing means 30, and the judgment means 40.

[0036] (Image processing means) The image processing means 30 acquires data on the luminance level of the shadow of the filament yarn X and the background measured by the area sensor as a two-dimensional image. The acquired image can be processed at a rate of, for example, 12 million pixels and 68 images per second. The size of the acquired image can be preset within the range of the area to be inspected, and can be enlarged or reduced. It is preferable that the image processing means 30 corrects the acquired image from the viewpoints of improving the efficiency and speed of processing, making it easier to set criteria for defect judgment, and improving the inspection accuracy of defects. A specific example of image correction will be described below.

[0037] Image processing involving correction is performed, for example, in the following order: (1) inspection area setting and image reduction, (2) shading correction, (3) filter processing, and (4) noise removal.

[0038] In (1), the processing load is reduced by limiting the inspection area to the position of (the shadow of) the filament yarn X and its surrounding area, thereby improving the efficiency of image processing, and by reducing the size of the captured image as necessary, speeding up processing and suppressing noise.

[0039] Due to the lens characteristics of camera 2, the peripheral image tends to be dark, making it impossible to detect the difference from defective parts of the yarn, and it may be impossible to distinguish between the two images by simple binarization processing. In contrast, in (2), in order to detect bare wires, the background of the object to be measured is corrected by image processing, and the brightness of the fiber (yarn) image and the peripheral image are made uniform, reducing or eliminating the effect of the difference in light quantity between the two images.

[0040] In (3), when detecting defects in the measurement target, the acquired image is smoothed by, for example, Gaussian filter processing to reduce noise while maintaining the contrast of the image of the defective part of the yarn. Note that the load increased by the filter processing in (3) can be reduced or distributed by strengthening the hardware.

[0041] In (4), the image is binarized, and the binarization results are combined using at least two thresholds, a relatively high threshold and a relatively low threshold, to remove noise from the image and obtain an optimized image. In the obtained image, some of the defect detection results may be interrupted. In response to this, thinning processing is used to extract short dots in the area and combine them with nearby short dots to connect the interruptions.

[0042] (judgment means) The determination means 40 automatically determines errors and their frequency from the acquired image, saves and accumulates output examples, stores thresholds according to the errors, and determines the inspection result of the filament yarn X based on the stored thresholds. More specifically, the determination means 40 is configured to determine errors and their frequency from the acquired image, store thresholds according to the errors in memory, and automatically determine the inspection result of the filament yarn X based on the stored thresholds.

[0043] The errors are causes of fuzz defects, and vary depending on the use, raw material, dimensions, required quality, etc. of the filament yarn to be inspected, or the required quality of the final product, etc. The threshold value can be determined by machine learning of the determination means 40 based on the error frequency, or can be appropriately set by the operator of the inspection device 100 or input to the memory in advance.

[0044] When there are multiple types of errors, it is preferable that the judgment means 40 stores a threshold series for each error type in the memory, calls up the stored threshold series according to each error type, and automatically judges the inspection result of the filament yarn X based on the called threshold series. The error type can be determined according to the raw material, appearance, or dimensions of the filament yarn or raw yarn to be measured, and affects the degree of fluff defects or the accuracy and stability of the weaving process. The error types include, but are not limited to, appearance errors, such as cut, loose, split, etc.; or yarn diameter errors, such as maximum diameter, minimum diameter, difference (R) between maximum diameter and minimum diameter, yarn diameter distribution (e.g., skewness, kurtosis, etc.), yarn diameter standard deviation, average diameter, etc. Among them, from the viewpoint of judgment accuracy, judgment based on appearance errors or judgment based on a combination of appearance errors and yarn diameter errors is preferable.

[0045] When there are multiple error frequency thresholds determined from the acquired image, or when there are multiple inflection points in the error frequency distribution graph, the judgment means 40 preferably divides the acquired image into a first threshold or less, first and second thresholds, and second threshold or more, and automatically outputs the judgment result of the acquired image. For example, when there are three or more thresholds, the judgment means 40 can divide the acquired image into a first threshold or less, first and second thresholds, second and third thresholds, and third threshold or more, and automatically outputs the judgment result of the acquired image.

[0046] In this embodiment, the number of thresholds and the thresholds that become the pass / fail criteria may be determined depending on the raw material to be measured, the distinction between warp and weft, or the type of error. For example, in the case of an error type of "cut" of a glass filament yarn as a weft, one threshold tends to be the pass / fail criteria value as it is. In the case of an error type of "loose" of a glass filament yarn as a weft, three thresholds tend to be determined, and an automatic output of the first threshold or less (no looseness), the first to second thresholds (looseness level "small (S)"), and the second to third thresholds (looseness level "medium (M)") is pass, and an automatic output of the third threshold or more (looseness level "large (L)") may fail. In addition, the threshold of the error type of "split" of a glass filament yarn tends to be set in the inspection of the warp yarn rather than the weft yarn.

[0047] From the viewpoint of judgment accuracy, it is preferable that the judgment means 40 determines the threshold value by combining the acquired image information and the yarn diameter measurement result of the filament yarn X. Information regarding the yarn diameter of the filament yarn X, such as the maximum diameter, the minimum diameter, the difference between the maximum diameter and the minimum diameter (R), the yarn diameter distribution (e.g., skewness, kurtosis, etc.), the yarn diameter standard deviation, the average diameter, etc., can be determined from a two-dimensional acquired image including data regarding the brightness levels of the shadow of the filament yarn X and the background.

[0048] The filament yarn X that has been automatically determined as passing the inspection result based on the threshold value by the determination means 40 can be supplied to another line or to a subsequent process, such as a weaving process. Also, the filament yarn X that has been automatically determined as failing for a particular use can be collected, returned, or diverted to another use.

[0049] The judgment means 40 preferably further judges the pass / fail of the raw yarn of the inspected filament yarn, or judges the pass / fail of the raw yarn on a bobbin basis, based on the inspection result of the filament yarn X judged as described above. For example, a raw yarn or a bobbin unit that is automatically judged as unacceptable for a specific use can be collected, returned, or diverted to another use. In the case of glass yarn, even if it is unacceptable as a weft yarn, it can be used as a warp yarn in the weaving process, and vice versa.

[0050] (Control Means) The control means 50 automatically controls the irradiation means 10, the measurement means 20, the image processing means 30, and the determination means 40, and can be executed by, for example, an automatic calculation processing device, a general-purpose personal computer, a large computer, a memory, an artificial intelligence, etc. The control means 50 can be included in the computer that controls the calculation unit, or can be arranged as a separate computer, or can be installed as an external server or cloud, and can be in data communication with the calculation unit by wire or wirelessly.

[0051] [Inspection method for filament yarn] (Overview of the inspection method) The filament yarn inspection method of the present invention is carried out using an inspection device including the above-mentioned irradiation means, measurement means, image processing means, and judgment means.

[0052] More specifically, a method for inspecting filament yarns, comprising the steps of: an irradiation step of irradiating the filament yarn with an inspection light; a measuring step of measuring the luminance level of the shadow of the filament yarn and the background obtained by irradiating the inspection light in the irradiation step with an area sensor; an image processing step for acquiring image information based on information obtained from the area sensor; a determining step of determining an error and a frequency of the error from the image information and determining an inspection result of the filament yarn based on a threshold value corresponding to the error; Execute.

[0053] (Flow chart of inspection method) Fig. 4 shows a flow chart of a filament yarn inspection method according to one embodiment of the present invention. Each step of the inspection method can be performed according to Fig. 4. Note that the steps in parentheses in Fig. 4 represent optional steps, which can be performed as desired.

[0054] According to the inspection method of the present invention, by inspecting and selecting filament yarn or its raw yarn at a stage before weaving processing, it is possible to eliminate or suppress fuzz defects during weaving processing of the filament yarn, thereby achieving improved yields of woven products and stable mass production. EXAMPLES

[0055] The following describes an actual inspection of a glass filament yarn using the inspection device 100 according to the present invention.

[0056] <Example 1> <Test subject> A total of 12 bobbins were prepared, each of which was wound with four different types of glass filament yarn (manufactured by AGY: LCD1020, LCD510, ECBC3750, and ECBC3000), three of each type. ·LCD1020 Filament yarn (5μm x 100 strands) LCD510 filament yarn (5μm x 200 strands) ·ECBC3750 filament yarn (4μm x 40 strands) ·ECBC3000 filament yarn (4μm x 50 strands)

[0057] <Device configuration> (A) An inspection device 100 having an external appearance as shown in FIG. 1 was prepared. (a) As shown in FIG. 1, two highly directional surface lights “TH2-43X35-PW” manufactured by CCS were used as the light source of the irradiation means 10. (c) As shown in FIG. 1, two cameras of Basler "boost boA4112-68cm" equipped with SONY CMOS sensors were used as the measuring means 20. (d) The calculation unit (30, 40, 50) includes a personal computer and a monitor, and is connected to the measurement means 20 by wire or wirelessly.

[0058] <Testing> The bobbin on which the above-prepared glass filament yarn was wound was set in the inspection device 100, and then pulled out at 1200 m / min. The shape of the filament yarn was imaged and the yarn diameter was measured by the area sensor of the measuring means 20.

[0059] The image from the area sensor was taken into the calculation unit, and (1) inspection area setting and image reduction, (2) shading correction, (3) filter processing, and (4) noise removal were performed in this order, and the corrected image data was sent to the judgment process. Based on the judgment criteria for the weft yarn for glass cloth, the judgment means automatically output (i) two thresholds for "cut" as the error type of the appearance morphology result, three thresholds for "loose", and one threshold for split, and (ii) "Average", "max", "min", "exceeding threshold 1", "exceeding threshold 2", and "distribution (standard deviation, kurtosis, skewness)" as the error type of the yarn diameter result, and stored them in memory.

[0060] <Result> Tables 1 to 4 show the judgment results and defect judgments for a total of 12 bobbins wound with four types of glass filament yarn (AGY, LCD1020, LCD510, ECBC3750, ECBC3000), three of each type, as well as the results of cloth production using the inspected raw yarn and bobbin residual yarn. Inspection length: 400m (20k shots) (*Average value for 50 bobbins)

[0061] [Table 1]

[0062] [Table 2]

[0063] [Table 3]

[0064] [Table 4]

[0065] <Example 2> The accuracy of the fuzz hit rate (those that predicted fuzz generation and actually became fuzzed) for 216 glass filament yarns (AGY LCD1020) was compared among the following: A) prediction from visual inspection by a human; B) yarn diameter judgment only; C) visual appearance judgment according to the present invention only; and D) visual appearance judgment according to the present invention + yarn diameter judgment. The results are shown in Table 5. The following accuracy comparison methods were also adopted and are summarized in Table 6.

[0066] [Table 5]

[0067] [Table 6] <Types of hairiness accuracy> A. Precision rate = TP / (TP+FP) B. Recall rate = TP / (TP+FN) C. Correct answer rate = (TP+TN) / (TP+FP+FN+TN) D. F value = 2 x A x B / (A + B)

[0068] The image data imported into the personal computer and the data after image processing were all saved and used for machine learning, for identifying the cause of fluff defects in the glass cloth, and for accurately determining the defects.

[0069] Tables 1 to 6 show that by using the filament yarn inspection device and inspection method according to the present invention, it is possible to predict with high accuracy whether or not fuzz will occur in the filament yarn. [Industrial Applicability]

[0070] The inspection device and inspection method according to the present invention can be used to inspect filament yarns or their raw yarns used in the manufacture of various components such as laminates, printed wiring boards, stent grafts, composite materials, reinforcing materials, and concrete crack suppression materials. The inspection device and inspection method according to the present invention can also be used to inspect defects not only in glass filament yarns, but also in fibers other than glass fibers (e.g., synthetic fibers, natural fibers, and metal fibers). By selecting and not using poor quality filament yarns in advance, it becomes possible to provide customers with high-quality products, and the industrial applications are extremely high. [Explanation of symbols]

[0071] 100 Inspection Equipment 101 Inspection Department 10 Irradiation means 20 Measurement methods 30 Image processing means 40 Judgment means 50 Control Means X-filament yarn

Claims

1. An inspection device for filament yarn, comprising: An irradiation means for irradiating the filament yarn with an inspection light; a measuring means for measuring the luminance level of the shadow of the filament yarn and the background obtained by irradiating the inspection light by the irradiating means with an area sensor; an image processing means for acquiring image information based on information obtained from the area sensor; and a judgment means for determining errors and a frequency of the errors from the image information, and judging an inspection result of the filament yarn based on a threshold value corresponding to the errors.

2. 2. The inspection device according to claim 1, wherein, when there are a plurality of types of errors, the determination means determines the inspection result of the filament yarn based on a threshold series corresponding to each type of error.

3. The inspection device described in claim 1 or 2, wherein when there are multiple thresholds, the judgment means divides the thresholds into those below a first threshold, those between the first and second thresholds, and those above the second threshold, and outputs the judgment result of the image acquired by the image processing means.

4. The inspection device described in claim 3, wherein when there are three or more thresholds, the judgment means divides the thresholds into a first threshold or less, a first to second thresholds, a second to third thresholds, and a third threshold or more, and outputs a judgment result of the image acquired by the image processing means.

5. The inspection device according to any one of claims 1 to 4, wherein the determination means determines the threshold value by combining the image information with a measurement result of a diameter of the filament yarn.

6. 6. The inspection device according to claim 1, wherein the image processing means corrects the image acquired by the image processing means.

7. The inspection device according to any one of claims 1 to 6, wherein the filament yarn is formed of at least one selected from the group consisting of glass fiber, acrylic resin fiber, fiber containing acrylic resin and other synthetic resin, and carbon fiber.

8. The inspection device of claim 7 , wherein the filament yarn is formed of fiberglass.

9. 9. The inspection device according to claim 1, further comprising a control means for controlling the irradiation means, the measurement means, the image processing means, and the determination means.

10. The inspection device according to any one of claims 1 to 9, wherein the judging means further judges whether the raw filament yarn is pass or fail based on the inspection result of the filament yarn.

11. The inspection device according to any one of claims 1 to 10, wherein the filament yarn is made up of 10 to 500 single yarns each having a single yarn diameter of 1 to 10 µm.

12. The inspection apparatus according to any one of claims 1 to 11, which is used offline.

13. The inspection device according to any one of claims 1 to 11, which is used in-line.

14. 1. A method for inspecting a filament yarn, comprising: an irradiation step of irradiating the filament yarn with an inspection light; a measuring step of measuring the luminance levels of the shadow of the filament yarn and the background obtained by irradiating the inspection light in the irradiating step with an area sensor; an image processing step for acquiring image information based on information obtained from the area sensor; a determination step of determining an error and a frequency of the error from the image information, and determining an inspection result of the filament yarn based on a threshold value corresponding to the error; The inspection method includes:

Images