Abnormality detection system

Abstract

To provide an abnormality detection system which can detect abnormalities such as a damaged yarn or the like in a fiber.SOLUTION: An abnormality detection system includes: an illumination section which irradiates a fiber belt which has a plurality of fibers expanded in a belt shape and travels in a constant direction with light; a plurality of imaging sections which are arranged side by side in a width direction of the fiber belt and captures an image by receiving reflection light in the fiber belt; a binarization processing section which performs each binarization processing for each image captured by each imaging section by defining a luminance threshold having a prescribed relation with an average value of luminance of the fiber belt in each image as a reference and generates each binarization image constituted of a first color and a second color; an image connection section which connects each binarization image and forms a connected image; and an abnormality determination section which determines abnormalities in the plurality of fibers on the basis of the connected image.SELECTED DRAWING: Figure 5

Description

[Technical field]

[0001] The present invention relates to an anomaly detection system for detecting anomalies in textiles. [Background technology]

[0002] 2. Description of the Related Art Synthetic fibers have been used conventionally as fibers for artificial hair that imitates human hair (human hair) and is attached to the head (for example, Patent Document 1). For example, in Patent Document 1, a flame-retardant polyester resin is used as a fiber for artificial hair, and the fiber for artificial hair in Patent Document 1 is said to have excellent flame retardancy and a luster close to that of human hair. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2018-090913 A Summary of the Invention [Problem to be solved by the invention]

[0004] During the manufacturing process, synthetic fibers may suffer from defects such as damaged yarns caused by threads breaking due to tension or shrinking due to heat from drying, lumps caused by poor yarn formation, fused yarns formed by the merging of multiple threads, and discolored yarns caused by heat. If defects such as damaged threads occur, there are problems in that the appearance becomes poor and the quality deteriorates when the hair is processed into artificial hair.

[0005] Therefore, an object of the present invention is to provide an abnormality detection system capable of detecting abnormalities such as damaged threads in textiles. [Means for solving the problem]

[0006] One aspect for solving the above-mentioned problems is an anomaly detection system including an illumination unit that irradiates light onto a fiber band in which a plurality of fibers are spread out in a band shape and which runs in a fixed direction; a plurality of imaging units that are arranged in parallel in the width direction of the fiber band and receive and capture light reflected from the fiber band; a binarization processing unit that performs binarization processing on each image captured by each imaging unit based on a luminance threshold that has a predetermined relationship with an average value of the luminance of the fiber band in each image, thereby generating each binary image composed of a first color and a second color; an image combining unit that combines each binary image to create a combined image; and an anomaly determination unit that determines abnormalities in the plurality of fibers based on the combined image.

[0007] According to this aspect, since fiber abnormalities are detected in a state in which a plurality of fibers are spread out in a band shape, fiber abnormalities are easily detected. According to this aspect, since the binarization process is individually performed on each image captured separately by each imaging unit, a combined image can be created without being affected by the brightness difference between the images captured by the imaging units. According to this aspect, the abnormality determination unit can determine abnormalities in the fibers based on the combined image, so that abnormalities in the fibers can be detected based on the determination result.

[0008] Incidentally, since the defect thread is formed by one or more intertwined threads during manufacturing, it has a certain size and is formed in a protruding state in most cases relative to other parts. Therefore, when light is irradiated from the illumination unit, the brightness of the light received by the imaging unit is likely to be lower at the location where the defect thread occurs than at other parts. In other words, the brightness of the image captured by the imaging unit is often lower at the location where the defect thread occurs, and when the brightness value is expressed in a histogram, it is often on the lower brightness side than the average brightness and appears at the foot of the distribution peak.

[0009] Thus, in a preferred aspect, the brightness threshold is smaller than the average brightness of the fibre bands.

[0010] According to this aspect, the brightness threshold is set lower than the average brightness of the fiber band, so that abnormal locations caused by damaged yarns can be easily extracted and abnormalities in damaged yarns can be easily detected.

[0011] In a preferred aspect, the brightness threshold is calculated from the standard deviation of the brightness of the fibre bands.

[0012] According to this aspect, binarization processing is performed on each image captured by the imaging unit in accordance with the image, so that abnormal areas can be more accurately separated from normal areas by the binarization processing.

[0013] In a preferred aspect, the combined image is expanded or contracted in the width direction of the fiber band.

[0014] According to this aspect, noise can be removed from the combined image without dividing an anomaly that straddles a boundary between images captured by each imaging unit.

[0015] In a preferred aspect, the plurality of imaging units are arranged in a straight line or a staggered manner.

[0016] This aspect makes it easy to generate a continuous combined image.

[0017] In a preferred aspect, the illumination unit and the plurality of imaging units form an imaging device, and the imaging device is provided such that an imaging axis of each imaging unit is inclined at a predetermined angle with respect to the fiber band.

[0018] According to this aspect, irregularities in the fiber belt are easily reflected, and abnormal locations can be detected more accurately.

[0019] In a preferred aspect, the abnormality determination unit determines that an abnormality has occurred when an area of ​​the portion displayed in the first color is equal to or larger than a predetermined area.

[0020] According to this aspect, abnormalities can be detected more accurately.

[0021] In a preferred aspect, the combined image is divided into a plurality of areas corresponding to each binary image, and the abnormality determination unit determines an abnormality to exist when the portion represented in the first color spans three or more areas.

[0022] According to this aspect, anomalies can be more easily detected. Effect of the Invention

[0023] According to the anomaly detection system of the present invention, anomalies such as damaged threads in textiles can be detected. [Brief description of the drawings]

[0024] [Figure 1] FIG. 1 is an explanatory diagram of an anomaly detection system according to a first embodiment of the present invention, where (a) is a side view showing a schematic diagram of the anomaly detection system, and (b) is a perspective view of a main part of (a). [Diagram 2] FIG. 2 is a block diagram of the anomaly detection system of FIG. [Diagram 3] 3 is an explanatory diagram of a luminance threshold used in the binarization process by the binarization processing unit in FIG. 2, showing a luminance value histogram of a captured image (solid line) and a normalized luminance value histogram of the captured image (two-dot chain line). [Figure 4] 2 is an explanatory diagram of the anomaly detection operation by the anomaly detection system of FIG. 1, where (a) shows an example of each captured image obtained in the imaging process, and (b) shows a combined image generated in a combining process based on each captured image in (a). [Diagram 5] 4A and 4B are explanatory diagrams of the anomaly detection operation by the anomaly detection system of FIG. 1, in which (a) shows a combined image in which noise has been removed by a noise removal process based on the combined image of (b) in FIG. 4, and (b) shows a combined image in which contours have been extracted by a contour extraction process based on the combined image of (a). [Figure 6] FIG. 11 is a side view showing a schematic diagram of an anomaly detection system according to another embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0025] Hereinafter, an embodiment of the present invention will be described in detail.

[0026] The abnormality detection system 1 of the first embodiment of the present invention detects abnormalities such as damaged yarns in a fiber belt 201 in which a plurality of fibers are spread out in a belt shape and run in a fixed direction (transport direction). The abnormality detection system 1 includes a fiber supplying section 2, a flattening section 3, a conveying section 5, a fiber collecting section 6, and a detection device 7, as shown in FIG.

[0027] (Fiber supply section 2) The fiber supplying section 2 is a section for supplying a bundle of multiple fibers (hereinafter, also referred to as a fiber bundle 200).

[0028] (Flat processing part 3) The flattening section 3 is a section for spreading the fiber bundle 200 supplied from the fiber supply section 2 in the width direction substantially without any gaps, and processing it into a fiber belt 201 extending in a belt shape in the conveyance direction. The flattening section 3 of this embodiment is equipped with at least two clamping rollers 30a, 30b, as shown in FIG. 1(a), and is capable of processing the fiber bundle 200 supplied from the fiber supply section 2 into a fiber belt 201 by clamping it between the clamping rollers 30a, 30b.

[0029] (Transportation section 5) The conveying section 5 is a section for conveying the fiber belt 201 at a predetermined speed, and is composed of a plurality of conveying rollers 50a to 50e as shown in FIG. 1(a). At least one of the transport rollers 50a to 50e is a drive roller that has a drive source and can be rotated by the drive source, and the remaining transport rollers are driven rollers that do not have a drive source and can rotate freely. All of the transport rollers 50a to 50e may be drive rollers. The transport roller 50d is a glossy metal roller made of metal and has light reflectivity. The conveying roller 50d is a conveying direction changing roller that changes the conveying direction to another direction, and also a tension roller that generates tension on the fiber belt 201.

[0030] (Fiber recovery section 6) The fiber recovery section 6 is a section for recovering the fiber belt 201 transported by the transport section 5.

[0031] (Detection device 7) The detector 7, as shown in FIG. 1, is disposed opposite the transport roller 50d and detects any abnormality in the fiber belt 201 traveling on the transport roller 50d. The detection device 7 is mainly composed of a management device 102 and an imaging device 103, as shown in FIG.

[0032] (Management device 102) The management device 102 is a computer having, in terms of its hardware configuration, a central processing unit consisting of a control device that controls each device and a calculation device that performs calculations on data, a storage device that stores data, an input device that inputs data from the outside, and an output device that outputs data to the outside. As shown in FIG. 2, the management device 102 includes a data storage unit 110, an image extraction unit 111, a binarization processing unit 112, an image combination unit 113, a noise removal unit 115, a contour extraction unit 116, an abnormality determination unit 117, an alarm unit 118, and an input / output unit 120.

[0033] The data storage unit 110 is a section that stores various data such as past and present captured image data, binarized image data, combined image data, and data related to abnormality detection.

[0034] The image extraction unit 111 is a component that extracts a portion in which only the fiber band 201 is shown from each captured image captured by each imaging unit 135 of the imaging group 132 of the imaging device 103 . Specifically, the image extraction unit 111 is capable of generating an image in which only the fiber belt 201 is shown by deleting the portions of the conveying roller 50d and the gaps between the fibers from the images captured by each imaging unit 135.

[0035] The binarization processing unit 112 is a part that converts the brightness of each captured image from which the fiber band 201 portion has been extracted by the image extraction unit 111 into a normal distribution as shown in FIG. 3, performs binarization processing based on the brightness threshold value Lt, and generates a binary image represented only in the first and second colors. The binarization processor 112 of this embodiment sets a range in which the luminance is equal to or less than the luminance threshold Lt as a first color, and sets a range in which the luminance exceeds the luminance threshold Lt as a second color.

[0036] The image combining unit 113 is a component that combines the binarized images, which have been binarized by the binarization processing unit 112 and expressed in the first and second colors, into one combined image. In other words, the image combining unit 113 combines each binary image in the width direction of the fiber band 201 side by side, and deletes overlapping portions, thereby making it possible to generate one continuous combined image.

[0037] The noise removal unit 115 is a part that removes noise by performing expansion or contraction processing in the juxtaposition direction of each binary image, without separating abnormalities that straddle areas 140 (140A to 140L) corresponding to each binary image in the combined image.

[0038] The contour extraction unit 116 is a unit that extracts the contour of the portion expressed in the first color in the combined image that has been subjected to the expansion process or the contraction process by the noise removal unit 115.

[0039] Anomaly determination section 117 is a section that determines an abnormality based on the area of ​​the portion expressed in the first color from the outline of the portion expressed in the first color in the combined image.

[0040] The notification unit 118 is a part that notifies an operator when the abnormality determination unit 117 determines that an abnormality has occurred. The notification method by the notification unit 118 is not particularly limited, and for example, the notification can be made by sound, color, video, image, language, or the like.

[0041] The input / output unit 120 is a part capable of two-way communication with an input / output unit 133 of the imaging device 103 via wired or wireless means, and capable of inputting and outputting data.

[0042] (Imaging device 103) The imaging device 103 is a line sensor that captures an image of a fiber band 201, which is an imaging target, in a line shape, and is specifically a contact image sensor (CIS). The imaging device 103 includes illumination units 130 and 131, an imaging group 132, and an input / output unit 133, as shown in FIGS.

[0043] As shown in FIG. 1(b), the illumination units 130 and 131 are units that irradiate light onto an imaging area P of a fiber band 201, which is an object to be imaged. The illumination units 130 and 131 have one or more light emitting elements arranged in parallel in the width direction of the fiber strip 201 . The light-emitting element is not particularly limited as long as it can emit light. For example, an LED, an organic EL, etc. can be used.

[0044] As shown in FIG. 1(b), the imaging group 132 is made up of a plurality of imaging sections 135 (135A to 135L). In the imaging device 103 of this embodiment, the imaging group 132 is made up of 12 imaging sections 135A to 135L, and 12 captured images can be obtained by each of the imaging sections 135A to 135L. That is, the imaging device 103 of this embodiment is capable of acquiring captured images by dividing the fiber belt 201 into 12 areas 140A to 140L in the width direction by the imaging sections 135A to 135L.

[0045] The imaging section 135 is a section that receives light that is irradiated from the illumination sections 130 and 131 and reflected from an imaging region P of the fiber band 201, which is an imaging target, and captures an image. Of the adjacent imaging sections 135, the distance between the closest imaging sections 135 is preferably 0.1 cm or more and 5 cm or less, and more preferably 1 cm or more and 2 cm or less.

[0046] (fiber) The fiber is not particularly limited in terms of its use, and can be used, for example, as an artificial hair fiber for use in wigs, braids, and the like. The material of the fibers is not particularly limited, but is preferably a chemical fiber such as a regenerated fiber, semi-synthetic fiber, or synthetic fiber, and more preferably a synthetic fiber that can be produced by a wet spinning apparatus. The cross-sectional shape of the fibers is not particularly limited, and may be, for example, a circle, an ellipse, or an irregular shape (for example, a horseshoe shape, a C-shape, or a Y-shape).

[0047] Here, the positional relationship between the imaging device 103 and the imaging area P of the fiber web 201 when the fiber web 201 is transported by the transport section 5 will be described.

[0048] 1B, the imaging device 103 has an imaging group 132 disposed between illumination units 130 and 131, and the imaging units 135 of the imaging group 132 are arranged in a staggered manner in the width direction. That is, they are arranged side by side in the width direction while being shifted alternately in the conveying direction. When the imaging device 103 is viewed in a plane, the first illumination section 130 and the second illumination section 131 are positioned on either side of the imaging area P in the transport direction, and the imaging group 132 overlaps with the imaging area P or is located upstream of the imaging area P in the transport direction. As shown in Figure 1(a), when viewed from the rotation axis direction of the conveying roller 50d, the optical axis of the first illumination section 130 is inclined toward the upstream side in the conveying direction with respect to the imaging area P, and the optical axis of the second illumination section 131 is inclined toward the downstream side in the conveying direction with respect to the imaging area P. The imaging device 103 is provided such that the imaging axis of each imaging section 135 of the imaging group 132 is inclined at a predetermined inclination angle θ1 with respect to the imaging area P. That is, the imaging device 103 is provided such that the imaging axis of each imaging section 135 of the imaging group 132 is inclined with respect to the direction connecting the rotation axis of the conveying roller 50d and the imaging area P (the normal direction of the rotation axis of the conveying roller 50d). 1(a), the inclination angle θ1 between the imaging axis of each imaging unit 135, the rotation axis of the conveying roller 50d, and the direction connecting the imaging area P is preferably greater than 0 degrees and equal to or greater than 30 degrees. The inclination angle θ1 is preferably less than 90 degrees and equal to or less than 60 degrees.

[0049] Next, an abnormality detection operation in the abnormality detection system 1 of this embodiment will be described.

[0050] The abnormality detection operation of the abnormality detection system 1 of this embodiment involves supplying a fiber bundle 200 from the fiber supply unit 2, forming the supplied fiber bundle 200 into a fiber strip 201 by the flattening unit 3, passing the fiber bundle 200 through the imaging area P at a predetermined speed by the conveying unit 5, and recovering the fiber strip 201 by the fiber recovery unit 6. The abnormality detection operation is an operation for detecting an abnormality such as a damaged yarn in the fiber belt 201 traveling in the conveying section 5 during this series of operation.

[0051] Specifically, while the fiber belt 201 is traveling at a predetermined speed by the conveying section 5, each imaging section 135 of the imaging group 132 images an imaging area P of the fiber belt 201, and the imaging data is input from the input / output section 133 of the imaging device 103 to the input / output section 120 of the management device 102 (imaging process).

[0052] The conveying speed of the conveying section 5 at this time is not particularly limited as long as it is a speed that can be detected by each imaging section 135. The conveying speed of the conveying section 5 is preferably 0.5 m / s or more and 3.0 m / s or less, and more preferably 1.0 m / s or more and 2.5 m / s or less.

[0053] At this time, among the imaging units 135A to 135L, there is imaging unit 135A that straddles the boundary portion between one end of the fiber band 201 and the conveying roller 50d in the width direction, and imaging unit 135L that straddles the boundary portion between the other end of the fiber band 201 and the conveying roller 50d, and the images captured by imaging unit 135A and imaging unit 135L capture the exposed portion of the fiber band 201 on the conveying roller 50d, as shown in Figure 4(a).

[0054] Next, the image extraction unit 111 uses the input imaging data to perform shading processing on each image captured by each imaging unit 135, removes data for parts that are the same color as the underlying conveying roller 50d, and extracts data for only the fiber band 201 portion (band extraction process).

[0055] Specifically, since the conveying roller 50d is light reflective and the reflected light is considered to have a high brightness, the captured images from each imaging unit 135 are converted to grayscale, and the data for the highly luminous white parts are deleted from the captured images as if they were the background color of the conveying roller 50d, and only the parts of the fiber belt 201 that are represented in black or gray are extracted.

[0056] 3, the binarization processing unit 112 normalizes the luminance value histogram (solid line) in the captured image from which the fiber band 201 has been extracted by the image extraction unit 111 to produce a normal distribution (two-dot chain line), and sets a luminance threshold Lt to a value that is a predetermined luminance L1 smaller than the average luminance value La of the portion of the fiber band 201 in the normal distribution. The binarization processing unit 112 then performs binarization processing based on the luminance threshold Lt, and generates a binarized image represented in the first and second colors by making the low luminance side a first color (black in this embodiment) and the high luminance side a second color (white in this embodiment) (binarization step).

[0057] In this case, the luminance threshold value Lt is preferably calculated from the average luminance value La of the fiber band 201 and the standard deviation σ of the luminance value of the fiber band 201 . The luminance threshold Lt is preferably equal to or greater than (La-1.5σ), and more preferably equal to or less than (La-2.5σ). That is, the predetermined luminance L1 is preferably equal to or greater than 1.5σ and equal to or less than 2.5σ In this embodiment, the predetermined luminance L1 is set to 2σ.

[0058] Next, the image combining unit 113 combines the binarized images binarized by the binarization processing unit 112 to generate one combined image as shown in FIG. 4(b) (combining step).

[0059] At this time, the image combining unit 113 combines the binary images in the width direction based on the positional relationship of the imaging units 135, and prioritizes one side of the adjacent imaging units 135 and deletes the other side in the overlapping portion. At this time, as shown in FIG. 4(b), linear portions expressed in the first color due to noise are formed in the areas 140A to 140L corresponding to each of the binarized images. The areas 140A to 140L correspond to the images captured by the imaging units 135A to 135L, and are defined in the width direction of the fiber band 201.

[0060] Next, the noise removal unit 115 performs an expansion process or a contraction process in the width direction on the combined image combined by the image combination unit 113 (noise removal step). That is, the noise removal unit 115 shrinks the first color portion in the width direction or expands the second color portion in the width direction to an extent that the noise (linear portion) between adjacent areas 140, 140 in the combined image is eliminated. In this embodiment, the noise removal unit 115 performs an expansion process, expanding the portion expressed in the second color in the juxtaposition direction and contracting the portion expressed in the first color.

[0061] Next, the contour extraction unit 116 extracts the contour of the part of the combined image shown in FIG. 5(a) that has had noise removed by the noise removal unit 115, which is represented by the first color, as shown in FIG. 5(b) (contour extraction process).

[0062] At this time, the contour extraction unit 116 removes, as necessary, contours whose area is equal to or smaller than a certain removal threshold from the combined image, as shown in FIG. 5(b). The removal threshold is set appropriately depending on the material and thickness of the fiber, for example, 0 mm. 2 Excess 499mm 2 It can be done as follows.

[0063] Next, when the area of ​​the contour (see FIG. 5(b)) extracted by the contour extraction unit 116 is equal to or greater than a judgment threshold (predetermined area), the abnormality judgment unit 117 judges that an abnormality has occurred, terminates the abnormality detection operation, and notifies the operator via the notification unit 118. On the other hand, when the area of ​​the contour extracted by the contour extraction unit 116 is less than the determination threshold (predetermined area), the abnormality determination unit 117 determines that the system is normal, and continues the abnormality detection operation (determination step).

[0064] At this time, the judgment threshold is appropriately set depending on the material and thickness of the fiber. For example, 2 More than 1000mm 2 It can be done as follows.

[0065] According to the abnormality detection system 1 of the present embodiment, fiber abnormalities are detected in the fiber band 201 in which a plurality of fibers are spread out in a band shape, and therefore, fiber abnormalities can be easily detected. Here, in the abnormality detection system 1 of this embodiment, the imaging units 135 are arranged in a staggered manner and are shifted in the conveying direction, which causes a problem that the color of the captured image varies for each imaging unit 135. Therefore, according to the anomaly detection system 1 of this embodiment, the binarization processing unit 112 individually performs binarization processing on each image captured by the multiple imaging units 135, so that the image combination unit 113 can create a combined image without being affected by the brightness difference or color of the images captured by the imaging units 135. According to the abnormality detection system 1 of the present embodiment, the abnormality determination unit 117 can determine an abnormality in the fibers based on the combined image, so that an abnormality in the fibers can be detected based on the determination result.

[0066] According to the anomaly detection system 1 of the present embodiment, since the luminance threshold value Lt is smaller than the average luminance value La of the fiber band 201, it is particularly easy to detect an anomaly such as a damaged yarn.

[0067] According to the abnormality detection system 1 of this embodiment, the brightness threshold Lt is calculated from the standard deviation σ of the brightness of the fiber band 201, making it possible to perform binarization processing suited to each captured image, and the binarization processing makes it easier to isolate abnormal areas more accurately.

[0068] According to the anomaly detection system 1 of the present embodiment, an expansion process or a contraction process is performed on the combined image in the width direction of the fiber band 201, so that noise can be removed from the combined image without separating anomalies that cross the boundaries between the images captured by each imaging unit.

[0069] According to the abnormality detection system 1 of the present embodiment, the multiple imaging units 135 are arranged in a staggered manner, which makes it easy to generate a continuous combined image.

[0070] According to the anomaly detection system 1 of this embodiment, the imaging device 103 is arranged so that the imaging axis of each imaging section 135 is inclined at a predetermined angle with respect to the fiber band 201, so that irregularities in the fiber band 201 are easily reflected, enabling more accurate detection of abnormalities.

[0071] According to the abnormality detection system 1 of the present embodiment, the abnormality determination unit 117 determines that an abnormality has occurred when the area of ​​the portion expressed in the first color is equal to or greater than the determination threshold, so that abnormalities can be detected more accurately.

[0072] According to the abnormality detection system 1 of the present embodiment, the composite image is expressed only in the first and second colors, so that when the abnormality determination unit 117 determines that an abnormality has occurred, the location of the abnormality can be identified by the composite image.

[0073] In the above embodiment, the abnormality determination unit 117 determines that an abnormality exists when the area of ​​the portion represented by the first color is equal to or greater than the determination threshold value, but the present invention is not limited to this. An abnormality may be determined by other methods using the combined image. For example, the abnormality determination unit 117 may determine that an abnormality exists when the portion represented by the first color of the combined image spans three or more areas 140, 140, 140. In this way, calculation of the area is not required, and abnormalities can be detected more easily. Furthermore, the abnormality determination unit 117 may determine that an abnormality has occurred when the portion of the combined image that is expressed in the first color crosses one of the areas 140. In this way, it is not necessary to calculate the area, and abnormalities can be detected more easily.

[0074] In the above embodiment, the imaging area P is set on the portion of the fiber belt 201 located on the conveying roller 50d to detect an abnormality, but the present invention is not limited to this. For example, as shown in FIG. 6, the imaging area P may be set on the portion of the fiber belt 201 located between the conveying rollers 50a and 50b to detect an abnormality. In this case, it is preferable to provide a light-reflective metallic or white base member 150 as a base for the fiber belt 201 so as to belong to the imaging area P. In this way, only the portion of the fiber belt 201 can be extracted in the belt extraction step based on the color of the base member 150, which is the background color in the captured image.

[0075] In the embodiment described above, the captured images from each image capturing unit 135 are converted to a grayscale image with black and white shading in the band extraction step, and the fiber band 201 portion is extracted from the captured images, but the present invention is not limited to this. The fiber band 201 portion may be extracted from the captured images using a gradation of two different colors.

[0076] In the above embodiment, the first color is black and the second color is white, but the present invention is not limited to this. The first color and the second color may be other colors. For example, the first color and the second color may be primary colors such as red, green, and blue, respectively.

[0077] In the embodiment described above, the imaging group 132 is composed of 12 imaging units 135A to 135L, but the present invention is not limited to this. The imaging group 132 only needs to be composed of at least two imaging units 135, and may be composed of 11 or less imaging units 135, or may be composed of 13 or more imaging units 135.

[0078] In the embodiment described above, the imaging units 135 of the imaging group 132 are arranged in a staggered manner in the width direction, but the present invention is not limited to this. The imaging units 135 of the imaging group 132 may be arranged in a straight line in the width direction.

[0079] In the above embodiment, the fiber bundle 200 supplied from the fiber supply unit 2 is processed into the fiber belt 201 by the flattening unit 3, but the present invention is not limited to this. When the fiber belt 201 is supplied from the fiber supply unit 2, the flattening unit 3 does not need to be provided.

[0080] In the above embodiment, the imaging device 103 is provided so that the imaging axis of the imaging section 135 is inclined to the upstream side in the conveying direction with respect to the imaging area P, but the present invention is not limited to this. The imaging device 103 may be provided so that the imaging axis of the imaging section 135 is inclined to the downstream side in the conveying direction with respect to the imaging area P.

[0081] In the above embodiment, the management device 102 is configured by one computer, but the present invention is not limited to this. It may be configured by multiple computers. In this case, the computers that configure the management device 102 may be connected wirelessly or by wire, or may be connected via a network such as the Internet or an intranet.

[0082] In the above-described embodiments, each component may be freely substituted or added between the respective embodiments as long as it falls within the technical scope of the present invention. [Explanation of symbols]

[0083] 1 Anomaly Detection System 103 Imaging device 112 Binarization processing unit 113 Image Combination Unit 117 Abnormality determination section 130 First Lighting Section (Lighting Section) 131 Second Lighting Section (Lighting Section) 135, 135A~135L Imaging section 140, 140A~140L area 201 Fiber belt

Claims

1. a lighting unit that irradiates light onto a fiber band in which a plurality of fibers are spread out in a band shape and which runs in a fixed direction; A plurality of imaging units arranged in parallel in a width direction of the fiber belt, the imaging units receiving light reflected by the fiber belt and imaging the fiber belt; a binarization processing unit that performs binarization processing on each image captured by each imaging unit based on a brightness threshold value that has a predetermined relationship with an average brightness value of the fiber band in each image, and generates a binarized image composed of a first color and a second color, an image combining unit that combines the binarized images to create a combined image; An abnormality detection system comprising an abnormality determination unit that determines abnormalities in the plurality of fibers based on the combined image.

2. The anomaly detection system of claim 1 , wherein the brightness threshold is less than an average brightness value of the fiber band.

3. The abnormality detection system according to claim 1 or 2, wherein the brightness threshold is calculated from a standard deviation of the brightness of the fiber band.

4. The anomaly detection system according to claim 1 or 2, further comprising a process for expanding or contracting the combined image in a width direction of the fiber band.

5. The anomaly detection system according to claim 1 , wherein the plurality of imaging units are arranged in a straight line or a staggered pattern.

6. The illumination unit and the plurality of imaging units constitute an imaging device, The abnormality detection system according to claim 1 , wherein the imaging device is provided such that an imaging axis of each imaging unit is inclined at a predetermined angle with respect to the fiber band.

7. The anomaly detection system according to claim 1 , wherein the anomaly determination unit determines that an anomaly has occurred when an area of ​​the portion represented in the first color is equal to or larger than a predetermined area.

8. the combined image is partitioned into a plurality of areas corresponding to each of the binarized images; The anomaly detection system according to claim 1 , wherein the anomaly determination unit determines that an anomaly has occurred when the portion represented by the first color spans three or more areas.

Images