Method for detecting defects in yarns in the production of elastane textile machines

Abstract

The present application belongs to the technical field of machine vision detection, and particularly relates to a spandex textile machine production yarn defect detection method, four light sources are arranged on each spandex textile machine, an AGV trolley is used as a cooperative robot inspection carrier, three cameras with different focal lengths are installed on the mechanical arm of the cooperative robot, and the first camera, the second camera and the third camera are in order from small to large according to the focal length; the first camera shoots images of two groups of upper and lower compression rollers, and the system performs reverse winding defect detection on the images; the first camera shoots swing images of the yarn on the first winding shaft and the second winding shaft, the second camera shoots swing images of the yarn on the third winding shaft and the fourth winding shaft, and the system performs swing anomaly detection on the swing images; and the third camera shoots images of the yarn between the first yarn guide roller and the second yarn guide roller, and the system performs yarn jump defect detection on the images. The present application realizes full-automatic detection of yarn defects in the spandex production process, and improves the detection efficiency and accuracy.

Description

Technical Field

[0001] This invention belongs to the field of machine vision inspection technology, specifically relating to a method for detecting defects in yarns during spandex textile production. Background Technology

[0002] Currently, the main method for inspecting spandex defects is manual inspection. The manual inspection method involves using a handheld flashlight to illuminate the spandex yarn from the side of the spandex spinning machine, and judging whether the spandex yarn is in good condition by observing the presence or absence of yarn and any abnormalities in its swing; then proceeding to the next spandex spinning machine.

[0003] Each workshop has 12 production lines, with 26 textile machines per line, for a total of 312 textile machines. Each production line is equipped with a person who continuously patrols and inspects the machines, which is a huge challenge to the workers' physical fitness. The long-term use of flashlights to observe the machines has a significant impact on the workers' eyesight.

[0004] Common defects in spandex spinning machine production include broken yarn, bundled yarn, skipped yarn, and reverse winding. Broken yarn: Broken yarn will cause the yarn to wrap around the winding shaft, resulting in one less yarn in the swing amplitude. Bundled yarn: Bundled yarn occurs when one yarn is missing, with two yarns bundled together, resulting in one less yarn in the swing amplitude. Skipped yarn: The overall number of yarns is the same, but in two adjacent bundles of yarn, one or more yarns skip from bundle A to bundle B, resulting in bundle A having fewer yarns than bundle B. Reverse winding: The yarn is not wound on the winding shaft but is incorrectly wound on the pressure roller. Summary of the Invention

[0005] The purpose of this invention is to address the issues of low efficiency in manual inspection, bias due to subjective judgment, and missed detection due to visual fatigue. It proposes a method for detecting yarn defects in spandex textile production, which achieves fully automated detection of yarn defects in the spandex production process, significantly improving detection efficiency and accuracy.

[0006] To achieve the above objectives, the technical solution adopted is:

[0007] A method for detecting yarn defects in spandex textile production, wherein each spandex textile machine includes a first winding shaft, a second winding shaft, a third winding shaft, a fourth winding shaft, a first guide roller, a second guide roller, and a third guide roller. Two pressure rollers are installed between the first and second winding shafts, and between the third and fourth winding shafts. Multiple yarns fed by the first guide roller are evenly wound onto the first, second, third, and fourth winding shafts via the second and third guide rollers. The method includes:

[0008] Four light sources are arranged on each spandex spinning machine. AGV carts are used as inspection carriers for collaborative robots. Three cameras with different focal lengths are installed on the robotic arms of the collaborative robots. According to the focal length from smallest to largest, they are the first camera, the second camera, and the third camera.

[0009] The first camera captures images of two sets of upper and lower pressure rollers, and the system uses these images to detect reverse winding defects.

[0010] The first camera captures images of the sway of the yarn on the first and second winding shafts, and the second camera captures images of the sway of the yarn on the third and fourth winding shafts. The system then performs sway abnormality detection on these images.

[0011] The third camera captures an image of the yarn between the first and second guide rollers, and the system uses this image to detect yarn skipping defects.

[0012] According to the method for detecting defects in yarn during spandex textile production of the present invention, four light sources are further installed on the side of the textile machine, and each light source is located above a winding shaft to illuminate the entire row of yarns fed to the winding shaft. The light sources are emitted from the emitting end to the fixed end face of the winding shaft. The focal length of the first camera is 8mm, the focal length of the second camera is 12mm, and the focal length of the third camera is 25mm.

[0013] According to the method for detecting yarn defects in spandex textile production of the present invention, the emitting end of the light source is a circular emitting port with a diameter of 55mm-100mm, and the light is focused by a plano-concave mirror to form a bright band.

[0014] According to the method for detecting yarn defects in spandex spinning machine production of the present invention, further, a first camera captures images of each group of upper and lower pressure rollers, and the system performs reverse winding defect detection on the images, including:

[0015] The AGV carrying the collaborative robot moves to the aisle in front of the winding shaft. At this time, the first camera is 1700mm-1800mm above the ground and takes an image of the pressure roller using the first camera.

[0016] For dark pressure rollers in the image, reverse winding defects are detected, and the reverse winding defects are extracted using the background difference method.

[0017] In detecting bright reverse-winding defects of pressure rollers, the background difference method is first used to obtain the area of ​​the reverse-winding defect to be determined. Since the pressure roller is bright and there is interference from shaft dirt, further processing is required to distinguish between true and false defects: First, the pressure roller is extracted in the image using threshold segmentation. When the background around the undetermined reverse-winding defect is black, it is determined to be a true reverse-winding defect. When the undetermined reverse-winding defect has a certain gray value around it, it is determined to be a false reverse-winding defect.

[0018] Ultimately, the reverse winding defect occurs on the upper pressure roller and / or the lower pressure roller.

[0019] According to the method for detecting yarn defects in spandex spinning machine production of the present invention, further, a first camera captures an image of the yarn's sway on a first winding shaft and a second winding shaft, and the system performs sway abnormality detection on the sway image, comprising:

[0020] The AGV carrying the collaborative robot moves to the aisle in front of the winding shaft. At this time, the first camera is 1500mm-1700mm above the ground and is located to the right of the third guide roller. The first camera is used to capture the swing amplitude image.

[0021] Gaussian filtering is applied to the acquired swing amplitude image;

[0022] A swing amplitude detection model is constructed using a neural network. The model is trained, and the image to be detected after Gaussian processing is input into the trained swing amplitude detection model to detect the swing amplitude and obtain the detection results, namely the position, width and height of the silk thread swing amplitude in the image.

[0023] To determine whether the amplitude of the silk thread swing is abnormal, the average amplitude of the silk thread swing is calculated based on the number of silk threads. Then, the amplitude of each silk thread swing is compared with the average. Silk threads swinging beyond the average are considered abnormal.

[0024] According to the method for detecting yarn defects in spandex textile production of the present invention, the second camera further captures images of the yarn swing amplitude on the third and fourth winding shafts. When the system performs swing amplitude abnormality detection on the swing amplitude images, the second camera is 1700mm-1800mm above the ground and located above and to the right of the third guide roller.

[0025] According to the method for detecting yarn defects in spandex spinning machine production of the present invention, further, a third camera captures an image of the yarn between the first guide roller and the second guide roller, and the system performs yarn skipping defect detection on the image, including:

[0026] The AGV carrying the collaborative robot moves to the aisle in front of the winding shaft. At this time, the third camera is 1500mm-1700mm above the ground and is located between the first guide roller and the second guide roller. The third camera is used to capture images of wire skipping.

[0027] The acquired wire skipping image is rotated using a rotation matrix to make the wire vertical in the image;

[0028] After rotating the image, threshold segmentation is performed within the ROI region to separate the fixed end region of the second guide roller from the background.

[0029] Find the feature value of row2, the row index of the lower right corner of the fixed end area. Based on the feature value of row2, further locate the detection area of ​​the thread. Then analyze the area and find the corresponding feature value according to the shooting position. Generate the final detection area based on the above feature values.

[0030] A wire skipping defect detection model is constructed using a neural network. The model is trained, and the processed image to be detected is input into the trained wire skipping defect detection model to detect wire skipping defects and obtain the detection results.

[0031] The beneficial effects achieved by adopting the above technical solution are:

[0032] This invention relates to a mobile vision inspection platform (AGV vehicle + collaborative robot + vision acquisition system) that dynamically monitors each textile machine without requiring real-time manual observation. The system autonomously makes judgments and issues alarms, representing a novel inspection system that integrates vision inspection technology and robotics. This system is adaptable to large-scale work areas while also handling individual machine inspections in spaces with limited aisles. In summary, this invention can promptly detect problems during textile machine production, such as reverse winding, abnormal swing amplitude (broken yarn defects), and yarn skipping, improving inspection efficiency and accuracy, saving labor costs, and enabling companies to quickly locate the source of defects through defect information statistics, allowing for timely quality improvements and preventing the production of large quantities of defective products that would otherwise lead to cost waste.

[0033] This invention uses a light source to illuminate the silk thread, enabling the camera to accurately capture the swaying amplitude, rewinding, and skipping of the silk thread. For this white, semi-transparent target object, the side lighting solves the problem of not being able to capture the target object. Attached Figure Description

[0034] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings of the embodiments of the present invention will be briefly described below. The drawings are merely illustrative of some embodiments of the present invention and are not intended to limit the scope of the present invention to all embodiments.

[0035] Figure 1 This is a flowchart illustrating the method for detecting yarn defects in spandex textile machine production according to an embodiment of the present invention.

[0036] Figure 2 This is a front view of the arrangement of the guide roller, light source, and camera according to an embodiment of the present invention;

[0037] Figure 3 This is a top view of the arrangement of the guide roller, light source, and camera according to an embodiment of the present invention;

[0038] Figure 4 This is a schematic diagram of the first camera capturing images of the upper and lower pressure rollers on the left side according to an embodiment of the present invention;

[0039] Figure 5 yes Figure 4 A schematic diagram of the reverse-winding defect taken by the first camera in the middle;

[0040] Figure 6 This is a schematic diagram of the first camera capturing images of the upper and lower right pressure rollers according to an embodiment of the present invention;

[0041] Figure 7 This is a schematic diagram of the first camera capturing the swing amplitude of the yarn on the first and second winding shafts according to an embodiment of the present invention;

[0042] Figure 8 yes Figure 7 A schematic diagram of the abnormal swing amplitude captured by the first camera in the middle;

[0043] Figure 9 This is a schematic diagram of the second camera capturing the swing amplitude of the filament on the third and fourth winding axes according to an embodiment of the present invention;

[0044] Figure 10 This is a front view of the yarn between the first guide roller and the second guide roller taken by the third camera in an embodiment of the present invention;

[0045] Figure 11 This is a top view of the yarn between the first guide roller and the second guide roller taken by the third camera in an embodiment of the present invention;

[0046] Figure 12 yes Figure 10 and Figure 11 A schematic diagram of wire skipping defects taken by the third camera in the middle. Detailed Implementation

[0047] The exemplary solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art.

[0048] This embodiment discloses a method for detecting yarn defects in spandex textile machine production. Each spandex textile machine includes a first winding shaft, a second winding shaft, a third winding shaft, a fourth winding shaft, a first guide roller, a second guide roller, and a third guide roller. The first and second winding shafts are located below the second guide roller, and the third and fourth winding shafts are located below the third guide roller. Two pressure rollers are installed between the first and second winding shafts, and between the third and fourth winding shafts. These pressure rollers ensure the formation of the yarn cake. During normal operation, the pressure rollers do not wind yarn. When a reverse winding defect occurs, the yarn is not wound on the winding shaft but on the pressure roller. Of the 80 yarns fed by the first guide roller, 20 are wound onto the first, second, third, and fourth winding shafts respectively via the second and third guide rollers. Figure 1 As shown, the method includes the following steps:

[0049] Step S101, as follows Figure 2 and Figure 3As shown, four light sources are arranged on each spandex spinning machine. AGVs are used as the inspection carriers for collaborative robots. Three cameras with different focal lengths are installed on the robotic arm of the collaborative robot. The cameras are numbered first, second and third according to their focal length from smallest to largest. The robotic arm carries the three cameras to take pictures of five points to detect different defects.

[0050] AGVs, serving as the carrier mechanism for inspection, can be either track-guided or trackless. Collaborative robots refer to industrial collaborative robots.

[0051] Step S102: The first camera captures images of the two sets of upper and lower pressure rollers, and the system performs reverse winding defect detection on the images.

[0052] Step S103: The first camera captures images of the sway of the yarn on the first and second winding shafts, and the second camera captures images of the sway of the yarn on the third and fourth winding shafts. The system then performs sway abnormality detection on the images.

[0053] Step S104: The third camera captures an image of the wire between the first guide roller and the second guide roller, and the system performs wire skipping defect detection on the image.

[0054] Four light sources are mounted on the side of the textile machine, each positioned above a winding shaft, to illuminate the entire row of yarn fed onto the shaft. The light source extends from its emitting end to the fixed end face of the winding shaft. The emitting end of the light source is a circular aperture with a diameter of 55mm-100mm, which is focused by a plano-concave mirror to form a bright band with a width of 70mm-120mm and a length of 1500mm-1800mm. The light source uses LED beads with a full load output of 350-550 lumens. The plano-concave mirror has a diameter of 60mm-100mm; the focal length of the first camera is 8mm, the second camera's focal length is 12mm, and the third camera's focal length is 25mm.

[0055] The camera is a 500W-2500W monochrome high-definition global camera with a frame rate of 4.5 frames per second, a dynamic range of 63dB, and a signal-to-noise ratio of 36dB. The lens is a 500W-2500W resolution, 8mm-25mm lens with a field of view of D: 20.32°, H: 16.77°, V: 11.24°; flange focal distance: 17.526mm.

[0056] like Figure 4 and Figure 5 As shown, point one: the image of the upper and lower pressure rollers on the left side taken by the first camera.

[0057] Step 1: The AGV carrying the collaborative robot moves to the aisle in front of the winding shaft. At this time, the first camera is 1700mm-1800mm above the ground and is located between the second and third guide rollers. The first camera is used to take pictures of the two pressure rollers on the left.

[0058] Step 2: Based on the position of the pressure roller in the image, predefine two Regions of Interest (ROIs) and crop the image according to the ROIs. Apply Gaussian filtering to the cropped image to eliminate some noise interference.

[0059] Step 3: Due to differences in workshop lighting and camera angle, the two pressure rollers appear differently in the images. The upper pressure roller is very bright and reflects light, while the lower pressure roller reflects almost no light. Since the lower pressure roller is unaffected, a reverse winding defect detection is performed on the lower pressure roller in the image. The reverse winding defect is extracted using the background subtraction method: mean filtering is applied to the pressure roller to obtain the current image. Then, the current image is subtracted from the background image to obtain the possible defect areas. After feature filtering (using the average gray level of the area within a certain range), the reverse winding defect is identified.

[0060] Step 4: When detecting the reverse winding defect of the upper pressure roller in the image, the background subtraction method is first used to obtain the area (rectangle) of the defect to be determined. Due to the shooting angle and workshop lighting, if there is dirt on the pressure roller, the image on the upper pressure roller will form a blurry area, affecting the detection of the real reverse winding defect. Further processing is required to distinguish between real and false defects: First, extract the upper pressure roller. Because the pressure roller is reflective, it can be easily extracted in the image using threshold segmentation. When the background around the defect to be determined is black, it is determined to be a real reverse winding defect; when there is a certain gray value around the defect, it is determined to be a false reverse winding defect. The threshold segmentation formula is:

[0061]

[0062] Where T represents the preset segmentation threshold, g(x) represents the binarized image, and f(x,y) represents the grayscale value of the pixel.

[0063] Step 5: By extracting the reverse winding defect in steps 3 and 4, it is finally determined that the reverse winding defect occurs on the upper pressure roller and / or the lower pressure roller.

[0064] like Figure 6 As shown, at point two: the first camera captures images of the upper and lower pressure rollers on the right side. The detection method is the same as at point one, except that the first camera is 1700mm-1800mm above the ground and located above and to the right of the third guide roller. The first camera captures images of the two pressure rollers on the right side.

[0065] like Figure 7 and Figure 8 As shown, point three: The first camera captures an image of the sway of the silk thread on the first and second winding shafts.

[0066] Step 1: The AGV carrying the collaborative robot moves to the aisle in front of the winding shaft. At this time, the first camera is 1500mm-1700mm away and located to the right of the third guide roller. The first camera is used to capture images.

[0067] Step 2: Apply Gaussian filtering to the acquired swing amplitude image to eliminate noise interference. The image is smoothed using a discrete Gaussian function; the discrete approximation formula for the Gaussian function is as follows:

[0068]

[0069] In the formula, x represents a random variable, and σ represents the variance.

[0070] Step 3: Predefine the ROI for detection. Crops the image based on the ROI to obtain the processed swing image and annotates it. Construct a swing detection model, which is an object detection model using a neural network. The model detects the swing in the image and compares the prediction results with the annotations to train the swing detection model, ultimately obtaining a trained swing detection model. After Gaussian processing and image cropping, a detection image is generated and input into the trained model for swing detection, obtaining the detection results, including the position, width, and height of the silk thread swing in the image.

[0071] Step 4: Determine if the wavy amplitude of the silk thread is abnormal. Specifically, calculate the average width of the wavy amplitude of the silk thread based on the number of silk threads, and then compare the wavy amplitude of each silk thread with the average value. Silk threads that exceed the average value are considered abnormal wavy amplitudes.

[0072] like Figure 9 As shown, at point four: the second camera captures images of the sway of the yarn on the third and fourth winding shafts. The detection method is the same as at point three, except that the second camera is 1700mm-1800mm above the ground and located to the upper right of the third guide roller. The second camera is used to capture the sway of the yarn on the third and fourth winding shafts.

[0073] like Figures 10-12 As shown, point five: The third camera captures an image of the yarn between the first guide roller and the second guide roller.

[0074] Step 1: The AGV carrying the collaborative robot moves to the aisle in front of the winding shaft. At this time, the third camera is 1500mm-1700mm above the ground and is located between the first guide roller and the second guide roller. The third camera is used to capture images of wire skipping.

[0075] Step 2: Rotate the acquired wire skipping image using a rotation matrix to ensure the wire is vertical in the image; the expression for the rotation matrix is ​​as follows:

[0076]

[0077] In the formula, HomMat2D is the generated empty matrix, Phi in the R matrix is ​​the angle to be rotated, and HomMat2DRotate is the final rotation matrix.

[0078] After rotating the image, a ROI is predefined, and threshold segmentation is used within the ROI region to separate the fixed end region of the second guide roller from the background region.

[0079] Step 3: Calculate the feature value of row2, the bottom right row index of the fixed-end area. Based on the feature value of row2, further locate the detection area of ​​the silk thread, and then analyze this area, calculating the corresponding feature value according to the shooting position. Due to the limited depth of field of the camera, the third camera has four shooting positions because it cannot capture all 80 silk threads in one shot. The first shooting position captures 1-20 silk threads clearly, the second shooting position captures 15-35 silk threads clearly, the third shooting position captures 30-65 silk threads clearly, and the fourth shooting position captures 55-80 silk threads clearly. When shooting at the first position, calculate the feature value of column1, the top left corner of the silk thread area. When shooting at the second, third, and fourth positions, calculate the feature value of column2, the bottom right corner of the silk thread area.

[0080] Step 4: Generate the final detection region based on the above feature values.

[0081] Step 5: Obtain the processed images of sponge yarn skipping and classify them, construct a skipping defect detection model, perform defect detection on the spandex yarn images, compare the predicted defect classification with the actual classification labels, train the skipping defect detection model, and finally obtain the trained skipping defect detection model. The spandex yarn images processed in steps 2-4 are then input into the trained skipping defect detection model to perform skipping defect detection and obtain the detection results.

[0082] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0083] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0084] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0085] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0086] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for detecting yarn defects in spandex textile production, wherein each spandex textile machine comprises a first winding shaft, a second winding shaft, a third winding shaft, a fourth winding shaft, a first guide roller, a second guide roller, and a third guide roller; upper and lower pressure rollers are respectively arranged between the first winding shaft and the second winding shaft, and between the third winding shaft and the fourth winding shaft; multiple yarns fed by the first guide roller are evenly wound onto the first winding shaft, the second winding shaft, the third winding shaft, and the fourth winding shaft via the second guide roller and the third guide roller; characterized in that, The method includes: Four light sources are arranged on each spandex spinning machine. AGV carts are used as inspection carriers for collaborative robots. Three cameras with different focal lengths are installed on the robotic arms of the collaborative robots. According to the focal length from smallest to largest, they are the first camera, the second camera, and the third camera. The first camera captures images of two sets of upper and lower pressure rollers, and the system performs reverse winding defect detection on the images. The first camera captures images of the sway of the yarn on the first and second winding shafts, and the second camera captures images of the sway of the yarn on the third and fourth winding shafts. The system then performs sway abnormality detection on these images. The third camera captures an image of the yarn between the first and second guide rollers. The system then performs yarn skipping defect detection on this image. Specifically, this involves: an AGV carrying a collaborative robot moving to the aisle in front of the winding shaft, at which point the third camera is 1500mm-1700mm above the ground and positioned between the first and second guide rollers, capturing an image of the yarn skipping; rotating the captured image using a rotation matrix to ensure the yarn is vertical in the image; performing threshold segmentation within the ROI region after rotation to separate the fixed end region of the second guide roller from the background; calculating the feature value of row2 at the bottom right corner of this fixed end region; further locating the detection area of ​​the yarn based on the feature value of row2; and then analyzing this area to calculate the corresponding feature value based on the image capture position. Based on the above feature values, the final detection region is generated; a wire skipping defect detection model is constructed using a neural network, the model is trained, and the processed image to be detected is input into the trained wire skipping defect detection model to perform wire skipping defect detection and obtain the detection result.

2. The method for detecting yarn defects in spandex textile production according to claim 1, characterized in that, Four light sources are installed on the side of the textile machine, with each light source located above a winding shaft to illuminate the entire row of yarns fed to the winding shaft. The light sources are emitted from the emitting end to the fixed end face of the winding shaft. The focal length of the first camera is 8mm, the focal length of the second camera is 12mm, and the focal length of the third camera is 25mm.

3. The method for detecting yarn defects in spandex textile production according to claim 2, characterized in that, The light source has a circular emission port with a diameter of 55mm-100mm at the emitting end. The light is focused by a plano-convex mirror to form a bright band.

4. The method for detecting yarn defects in spandex textile production according to claim 1, characterized in that, The first camera captures images of each set of upper and lower pressure rollers. The system performs reverse winding defect detection on these images, including: The AGV carrying the collaborative robot moves to the aisle in front of the winding shaft. At this time, the first camera is 1700mm-1800mm above the ground and takes an image of the pressure roller using the first camera. For dark pressure rollers in the image, reverse winding defects are detected, and the reverse winding defects are extracted using the background difference method. In detecting bright reverse-winding defects of pressure rollers, the background difference method is first used to obtain the area of ​​the reverse-winding defect to be determined. Since the pressure roller is bright and there is interference from shaft dirt, further processing is required to distinguish between true and false defects: First, the pressure roller is extracted in the image using threshold segmentation. When the background around the undetermined reverse-winding defect is black, it is determined to be a true reverse-winding defect. When the undetermined reverse-winding defect has a certain gray value around it, it is determined to be a false reverse-winding defect. Ultimately, the reverse winding defect occurs on the upper pressure roller and / or the lower pressure roller.

5. The method for detecting yarn defects in spandex textile production according to claim 1, characterized in that, The first camera captures images of the yarn's sway on the first and second winding shafts. The system performs anomaly detection on these images, including: The AGV carrying the collaborative robot moves to the aisle in front of the winding shaft. At this time, the first camera is 1500mm-1700mm above the ground and is located to the right of the third guide roller. The first camera is used to capture the swing amplitude image. Gaussian filtering is applied to the acquired swing amplitude image; A swing amplitude detection model is constructed using a neural network. The model is trained, and the image to be detected after Gaussian processing is input into the trained swing amplitude detection model to detect the swing amplitude and obtain the detection results, namely the position, width and height of the silk thread swing amplitude in the image. To determine whether the amplitude of the silk thread swing is abnormal, the average amplitude of the silk thread swing is calculated based on the number of silk threads. Then, the amplitude of each silk thread swing is compared with the average. Silk threads swinging beyond the average are considered abnormal.

6. The method for detecting yarn defects in spandex textile production according to claim 5, characterized in that, The second camera captures images of the sway of the yarn on the third and fourth winding shafts. When the system detects abnormal sway in these images, the second camera is 1700mm-1800mm above the ground and positioned to the upper right of the third guide roller.

Images