A segmented circulating warp knitting machine yarn breakage detection system and method

Abstract

This invention relates to a segmented, iterative warp knitting machine yarn breakage detection system and method. The system includes a display, a host computer, and several detection devices. Both the detection devices and the display are electrically connected to the host computer. Each detection device includes a sliding table mechanism, a camera assembly, and a supplementary lighting assembly. The method includes acquiring yarn images and creating a reference image set; for a given yarn image, determining whether the camera assemblies on both sides are aligned when the image is acquired; if so, shadow removal is performed; if not, yarn breakage detection is performed; and displaying the coordinates of the yarn breakage point when the detection model detects a yarn breakage. The advantages of this invention are: by having the camera assemblies reciprocate to scan and acquire yarn images at the knitting mechanism on both sides of the warp knitting machine, a smaller number of camera assemblies can cover a larger area of ​​the warp knitting machine's working area, reducing deployment costs and the computational power requirements of the host computer. A shadow removal algorithm is included to effectively eliminate shadow interference generated by the supplementary lighting assemblies on opposite sides when the camera assemblies are aligned.

Description

Technical Field

[0001] This invention relates to the field of detection technology, and in particular to a segmented circulating warp knitting machine yarn breakage detection system and method. Background Technology

[0002] Warp knitting machine is short for warp knitting machine. It refers to a device that feeds one or more sets of parallel warp yarns longitudinally into a needle bed, where they are simultaneously looped and interlocked by knitting needles to form warp-knitted fabric. To ensure product quality, warp knitting machines need to be inspected for yarn breakage during operation.

[0003] Existing yarn breakage detection solutions, such as the utility model patent with application number 202620394592.8, disclose a fixed warp knitting machine yarn breakage detection device. This device uses several fixed-position cameras on both sides of the warp knitting machine to detect yarn breakage using a detection model. This technical solution is only suitable for small warp knitting machines, which are shorter in overall length and can be fully covered by fixed-position cameras. However, large warp knitting machines are longer, and using fixed-position cameras would require a large number of cameras. Using a large number of cameras for yarn breakage detection would require a high-performance host computer with significant computing power, resulting in high deployment costs. Summary of the Invention

[0004] This invention primarily addresses the aforementioned problems by providing a segmented, cyclical warp knitting machine yarn breakage detection system and method. This system allows cameras to reciprocate and scan both sides of the warp knitting machine to acquire yarn images at the knitting mechanism. It utilizes a smaller number of cameras to cover a larger working area, reducing deployment costs and the computational power requirements of the host computer.

[0005] The technical solution adopted by this invention to solve its technical problem is a segmented circulating warp knitting machine yarn breakage detection system, including a display, a host computer, and several detection devices disposed on both sides of the warp knitting machine along its length. Both the detection devices and the display are electrically connected to the host computer. Each detection device includes: The slide mechanism is set along the length of the warp knitting machine; The camera component is mounted on the slide rail mechanism and is driven by the slide table mechanism to reciprocate within a preset length range to acquire yarn images at the warp knitting machine's weaving mechanism. The supplementary lighting component is mounted below the camera component via a mounting bracket and moves synchronously with the camera component.

[0006] As a preferred embodiment of the above solution, the slide mechanism is provided with a limit sensor for limiting the movement range of the camera assembly.

[0007] As a preferred embodiment of the above solution, a strip groove is fixedly provided below the slide mechanism, and a drag chain is provided in the strip groove, with the wiring harness of the camera assembly disposed in the drag chain.

[0008] As a preferred embodiment of the above solution, the mounting bracket includes a front bending plate and a rear bending plate. The upper ends of the front bending plate and the rear bending plate are respectively fixedly disposed on both sides of the slide of the slide mechanism, and the lower ends of the front bending plate and the rear bending plate are respectively fixed below the slide mechanism after passing around the slide mechanism and the sides of the strip groove.

[0009] As a preferred embodiment of the above solution, a first mounting plate extends vertically from the lower end of the front bending plate. The first mounting plate is provided with at least two vertically extending first straight slots. A lifting frame that moves up and down along the first straight slots is provided on the first mounting plate. The lighting assembly is rotatably mounted on the lifting frame.

[0010] As a preferred embodiment of the above solution, the lifting frame includes a base plate slidably disposed on a first mounting plate and first mounting ears disposed on both sides of the base plate. The upper end of the lighting assembly is provided with a second mounting ear. Both the first mounting ear and the second mounting ear are provided with through holes, and bolts are inserted through the through holes.

[0011] As a preferred embodiment of the above solution, the slide of the slide assembly is provided with two vertically extending mounting strips arranged opposite each other, and the mounting strips are provided with a second straight groove. The lower rear side of the camera assembly is provided with a rotating plate, and the rotating plate is provided with a through hole for bolts to pass through.

[0012] As a preferred embodiment of the above scheme, the position coordinates of the camera components in the two opposite detection devices on both sides of the warp knitting machine at each current relative moment are different from the position coordinates at the previous relative moment.

[0013] Correspondingly, the present invention also provides a segmented circulating warp knitting machine yarn breakage detection method, which adopts the above-mentioned segmented circulating warp knitting machine yarn breakage detection system, including: the host computer controls each camera component to move back and forth to acquire yarn images at the knitting mechanism of the warp knitting machine; During the process of acquiring yarn images, the position coordinates of the two camera components when they face each other again are calculated based on the moving speed and moving direction of the two camera components on opposite sides of the warp knitting machine and are called the facing coordinates; when the camera component passes the facing coordinates and the opposite camera component is not near the facing coordinates, the images of the facing coordinates and several frames before and after the facing coordinates are acquired as a reference image set. For a yarn image captured by a camera component, determine whether the camera component and the opposite camera component are both located near the same coordinates when the yarn image is captured. If so, shadow exclusion is performed; otherwise, the yarn image is input into a pre-trained detection model for yarn breakage detection. When the detection model detects a yarn breakage, the coordinates of the yarn breakage point are displayed on the monitor.

[0014] As a preferred embodiment of the above solution, the shadow exclusion includes: Obtain the difference map between the current yarn image and the corresponding reference frame image in the reference image set, and perform Gaussian filtering; The dynamic threshold T = μ + k × σ is calculated based on the mean μ and standard deviation σ of the difference plot, where k is the threshold coefficient. Regions exceeding the dynamic threshold T in the difference map are marked as shadow interference regions, and morphological opening and closing operations are performed on the shadow interference regions. Input the current yarn image into the detection model to obtain candidate yarn breakage regions. Calculate the ratio of the intersection area of ​​each candidate yarn breakage region and the shadow interference region to the area of ​​the candidate yarn breakage region. If the ratio is greater than a preset value, the candidate yarn breakage region is determined to be a shadow interference region and removed. If several ratios are less than the preset value, the detection model determines whether there is a yarn breakage in the candidate yarn breakage region.

[0015] The advantages of this invention are: by having the camera components reciprocate to scan and acquire yarn images at the weaving mechanism on both sides of the warp knitting machine, a larger working area of ​​the warp knitting machine can be covered with a smaller number of camera components, reducing deployment costs and the computing power requirements of the host computer. A shadow elimination algorithm is included to effectively eliminate shadow interference generated on the opposite side by the respective supplementary lighting components of the two camera components when they are facing each other. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the detection device.

[0017] Figure 2 for Figure 1 A magnified view of a portion of region A in the middle.

[0018] Figure 3 for Figure 1 A magnified view of a portion of region B in the middle.

[0019] 1-Slide mechanism 2-Strip groove 3-Fill light assembly 4-Mounting bracket 5-Camera assembly 6-Mounting strip 7-First mounting plate 8-Lifting frame 9-Second mounting ear Detailed Implementation

[0020] The technical solution of the present invention will be further described below through embodiments and in conjunction with the accompanying drawings.

[0021] Example: This embodiment discloses a segmented circulating warp knitting machine yarn breakage detection system, including a display, a host computer, and several detection devices disposed on both sides of the warp knitting machine along its length. Both the detection devices and the display are electrically connected to the host computer. Figures 1 to 3As shown, each detection device includes a sliding table mechanism 1, a camera assembly 5, and a supplementary lighting assembly 3. The host computer acquires images captured by the camera assembly 5 and performs yarn breakage detection using a yarn breakage detection model based on YOLOv8-nano or MobileNetV3.

[0022] The sliding table mechanism 1 is arranged along the length of the warp knitting machine. The camera assembly is fixedly mounted on the upper sliding table of the sliding table mechanism 1. The host computer moves the camera assembly back and forth within a preset length range via the sliding table mechanism. During the movement, the camera assembly acquires real-time images of the yarn at the weaving mechanism of the warp knitting machine. The supplementary lighting assembly 3 is mounted below the camera assembly via a mounting bracket 4 and moves synchronously with the camera assembly to provide supplementary lighting. In this embodiment, the position coordinates of the camera assemblies 5 in the two opposing detection devices on both sides of the warp knitting machine are different at each current relative moment from the position coordinates at the previous relative moment. Specifically, the moving speed, moving direction, or initial position of the two opposing camera assemblies are different. This arrangement reduces the number of times the two camera assemblies face each other and ensures that their positions are different each time they face each other, minimizing the shadow cast by the supplementary lighting assembly 3 on the opposite side that could affect the yarn breakage detection results of the host computer.

[0023] Furthermore, limit sensors are provided at both ends of the slide rail of the sliding table mechanism. These limit sensors are used to limit the movement range of the camera assembly. In this embodiment, the host computer can obtain the specific position of the slide table on the slide rail through the sliding table mechanism. When the slide table moves to the limit sensor, the host computer receives the limit sensor signal, changes the movement direction of the slide table, and recalibrates the position of the slide table. The setting of the limit sensor ensures that the host computer can always accurately obtain the position of the camera assembly, avoiding the continuous accumulation of slide table position detection errors during long-term operation of the sliding table mechanism.

[0024] Furthermore, mounting beams are fixedly installed at both ends below the slide mechanism 1, and the other end of the mounting beam is fixedly installed on the warp knitting machine frame. A strip groove 2 is fixedly installed below the mounting beam through a connector. A drag chain is installed in the strip groove 2, and the signal wires, power wires and other wire harnesses of the camera component are installed in the drag chain.

[0025] Furthermore, the mounting frame 4 includes a front bending plate and a rear bending plate. The upper ends of the front bending plate and the rear bending plate are respectively fixedly mounted on the front and rear sides of the slide mechanism 1. The lower ends of the front bending plate and the rear bending plate pass around the front and rear sides of the slide mechanism 1 and the strip groove 2 and are fixed together below the slide mechanism. A first mounting plate 7 extends vertically from the lower end of the front bending plate. The first mounting plate 7 is provided with at least two vertically extending first straight slots. A lifting frame 8 that moves up and down along the first straight slots is provided on the first mounting plate 7. The lighting assembly 3 is rotatably mounted on the lifting frame 8. Specifically, the lifting frame 8 includes a base plate that slides on the first mounting plate 7 and first mounting ears provided on both sides of the base plate. The upper end of the lighting assembly 3 is provided with a second mounting ear 9. Both the first mounting ear and the second mounting ear are provided with through holes, through which bolts are inserted. In this embodiment, the first mounting ear 9 and the first mounting ear can rotate within a certain angle range around the bolt as an axis. By adjusting the tightness of the bolt, the illumination angle of the supplementary lighting assembly 3 can be adjusted.

[0026] Furthermore, the slide assembly 1 has two vertically extending mounting strips 6 arranged opposite each other. Each mounting strip 6 has a second straight groove. A rotating plate is located at the rear of the lower end of the camera assembly 5, and the rotating plate has through holes for bolts to pass through. In this embodiment, the camera assembly 5 can be height-adjusted along the second straight groove. In addition, the camera assembly 5 can also be adjusted for forward and backward rotation angles.

[0027] Correspondingly, this embodiment also provides a method for detecting yarn breakage on a segmented circulating warp knitting machine, employing the aforementioned segmented circulating warp knitting machine yarn breakage detection system, including: The host computer controls the reciprocating movement of each camera component to acquire yarn images at the weaving mechanism of the warp knitting machine. During the acquisition of yarn images, the position coordinates of the two camera components facing each other on both sides of the warp knitting machine are calculated based on their moving speed and direction, and are designated as the facing coordinates. When a camera component passes the facing coordinate and the opposite camera component is not near the facing coordinate, images of the facing coordinate and several frames before and after the facing coordinate are acquired as a reference image set. Each pair of opposing camera components has a reference image set. After each camera component faces each other, a new reference image set is generated. The images in the reference image set need to be detected by a detection model to ensure that there are no broken yarns.

[0028] For a yarn image captured by a camera component, determine whether the camera component and the opposite camera component are both located near the same coordinates when the yarn image is captured. If so, shadow exclusion is performed; otherwise, the yarn image is input into a pre-trained detection model for yarn breakage detection. When the detection model detects a yarn breakage, the coordinates of the yarn breakage point are displayed on the monitor.

[0029] Furthermore, shadow exclusion includes: Obtain the difference map D(x,y) between the current yarn image and the corresponding reference frame image in the reference image set. D(x,y) = |I(x,y) - I ref (x,y)|, where D represents the difference map, I represents the current frame image, and I ref The reference frame image is represented by (x, y), which represents the pixel coordinates. Absolute value operations are used to eliminate the influence of positive and negative differences, preserving the amplitude information of all grayscale variations. Subsequently, a Gaussian filter is applied to the difference image to eliminate noise and small fluctuations. The dynamic threshold T = μ + k × σ is calculated based on the mean μ and standard deviation σ of the difference map, where k is the threshold coefficient. Regions in the difference map that exceed the dynamic threshold T are marked as shadow interference regions. Morphological opening and closing operations are performed on the shadow interference regions to optimize the boundaries of the shadow regions and eliminate small noise. Input the current yarn image into the detection model to obtain candidate yarn breakage regions. Calculate the ratio of the intersection area of ​​each candidate yarn breakage region and the shadow interference region to the area of ​​the candidate yarn breakage region. If the ratio is greater than a preset value, the candidate yarn breakage region is determined to be a shadow interference region and removed. If several ratios are less than the preset value, the detection model determines whether there is a yarn breakage in the candidate yarn breakage region.

[0030] The segmented cyclic scanning warp knitting machine yarn breakage detection system and method in this embodiment uses camera components to perform segmented cyclic scanning of the warp knitting machine. This allows a smaller number of camera components to cover a longer working range of the warp knitting machine, reducing hardware costs and the computational requirements of the host computer. Furthermore, by having opposing camera components run at different speeds, directions, or initial positions, the number of times the camera components and supplementary lighting components on both sides face each other is reduced. The system also ensures that adjacent facing positions are different, and by combining this with a shadow elimination algorithm, it minimizes the interference of shadows cast by the supplementary lighting components on the opposite side when they are facing each other, thus ensuring detection accuracy.

[0031] The specific embodiments described herein are merely illustrative of the spirit of the invention. Those skilled in the art to which this invention pertains may make various modifications or additions to the described specific embodiments or use similar methods to substitute them, without departing from the spirit of the invention or exceeding the scope defined by the appended claims.

Claims

1. A segmented circulating warp knitting machine yarn breakage detection system, characterized in that: The system includes a display, a host computer, and several detection devices positioned on both sides of the warp knitting machine along its length. Both the detection devices and the display are electrically connected to the host computer. Each detection device includes: The slide mechanism is set along the length of the warp knitting machine; The camera component is mounted on the slide rail mechanism and is driven by the slide table mechanism to reciprocate within a preset length range to acquire yarn images at the warp knitting machine's weaving mechanism. The supplementary lighting component is mounted below the camera component via a mounting bracket and moves synchronously with the camera component.

2. The segmented circulating warp knitting machine yarn breakage detection system according to claim 1, characterized in that: The slide mechanism is equipped with a limit sensor for limiting the range of movement of the camera assembly.

3. The segmented circulating warp knitting machine yarn breakage detection system according to claim 1, characterized in that: A strip groove is fixedly provided below the slide mechanism, and a drag chain is provided in the strip groove. The wiring harness of the camera assembly is arranged in the drag chain.

4. The segmented circulating warp knitting machine yarn breakage detection system according to claim 3, characterized in that: The mounting bracket includes a front bending plate and a rear bending plate. The upper ends of the front bending plate and the rear bending plate are respectively fixedly installed on both sides of the slide of the slide mechanism. The lower ends of the front bending plate and the rear bending plate pass around the slide mechanism and the two sides of the strip groove and are fixed together below the slide mechanism.

5. The segmented circulating warp knitting machine yarn breakage detection system according to claim 4, characterized in that: The lower end of the front bending plate has a vertically extending first mounting plate. The first mounting plate has at least two vertically extending first straight slots. The first mounting plate is provided with a lifting frame that moves up and down along the first straight slots. The lighting assembly is rotatably mounted on the lifting frame.

6. The segmented circulating warp knitting machine yarn breakage detection system according to claim 5, characterized in that: The lifting frame includes a base plate slidably disposed on a first mounting plate and first mounting ears disposed on both sides of the base plate. The upper end of the lighting assembly is provided with a second mounting ear. Both the first mounting ear and the second mounting ear are provided with through holes, and bolts are inserted through the through holes.

7. The segmented circulating warp knitting machine yarn breakage detection system according to claim 1, characterized in that: The slide assembly has two vertically extending mounting strips arranged opposite each other. The mounting strips have a second straight groove. The camera assembly has a rotating plate at the lower rear side, and the rotating plate has a through hole for bolts to pass through.

8. The segmented circulating warp knitting machine yarn breakage detection system according to claim 1, characterized in that: The position coordinates of the camera components in the two opposing detection devices on both sides of the warp knitting machine at each current relative moment are different from the position coordinates at the previous relative moment.

9. A method for detecting yarn breakage on a segmented circulating warp knitting machine, employing the segmented circulating warp knitting machine yarn breakage detection system as described in any one of claims 1-8, characterized in that it comprises: The host computer controls the reciprocating movement of each camera component to acquire yarn images at the weaving mechanism of the warp knitting machine; During the process of acquiring yarn images, the position coordinates of the two camera components when they face each other again are calculated based on the moving speed and moving direction of the two camera components on opposite sides of the warp knitting machine and are called the facing coordinates; when the camera component passes the facing coordinates and the opposite camera component is not near the facing coordinates, the images of the facing coordinates and several frames before and after the facing coordinates are acquired as a reference image set. For a yarn image captured by a camera component, determine whether the camera component and the opposite camera component are both located near the same coordinates when the yarn image is captured. If so, shadow exclusion is performed; otherwise, the yarn image is input into a pre-trained detection model for yarn breakage detection. When the detection model detects a yarn breakage, the coordinates of the yarn breakage point are displayed on the monitor.

10. The method for detecting yarn breakage on a segmented circulating warp knitting machine according to claim 9, characterized in that: The shadow exclusion includes: Obtain the difference map between the current yarn image and the corresponding reference frame image in the reference image set, and perform Gaussian filtering; The dynamic threshold T = μ + k × σ is calculated based on the mean μ and standard deviation σ of the difference plot, where k is the threshold coefficient. Regions exceeding the dynamic threshold T in the difference map are marked as shadow interference regions, and morphological opening and closing operations are performed on the shadow interference regions. Input the current yarn image into the detection model to obtain candidate yarn breakage regions. Calculate the ratio of the intersection area of ​​each candidate yarn breakage region and the shadow interference region to the area of ​​the candidate yarn breakage region. If the ratio is greater than a preset value, the candidate yarn breakage region is determined to be a shadow interference region and removed. If several ratios are less than the preset value, the detection model determines whether there is a yarn breakage in the candidate yarn breakage region.

Images