Method for detecting knitting needle defects of automatic knitting machine, corresponding system and computer program

Abstract

A method, system and computer program for detecting knitting needle defects of an automatic knitting machine wherein a digital camera captures digital image frames of a set of knitting needles in an operating state; performing automatic image recognition to determine a first pattern co-defined by the needle and / or a second pattern defined by the length of the yarn interacting with the needle; automatically deriving at least one parameter from the first pattern or the second pattern; automatically calculating a deviation between the parameter and a predetermined reference value; and automatically determining whether a defect exists in the knitting needle group according to the deviation.

Description

Technical Field

[0001] The invention belongs to the field of automatic knitting machines.

[0002] More specifically, the present invention relates to a method for detecting defects in knitting needles of an automatic knitting machine, wherein the automatic knitting machine is of a type that includes multiple yarn feeders, multiple movable knitting needles, each of which is suitable for sequentially hooking and releasing yarn provided by one of the yarn feeders when the knitting needle moves, and an actuator device for automatically moving the knitting needles according to a preset pattern to use the yarn provided by the yarn feeders to manufacture knitted fabrics.

[0003] The invention also relates to a system for detecting defects in the needles of an automatic knitting machine of this type, and to a corresponding computer program for carrying out the method. Background Art

[0004] Automatic knitting machines are commonly used to produce knitted fabrics on an industrial scale. These machines are designed to operate intensively. They have numerous components, including numerous needles and the mechanisms that move them, which must work with extreme precision to knit each stitch flawlessly. The machines are typically equipped with sensors and monitored by human knitters, who can halt the production process if a problem occurs. Both the machine components and the knitted fabric being produced can be monitored. When a machine is stopped due to a detected defect in the fabric, this interruption is costly, as the human knitter must inspect the fabric, discard the defective part, resolve the problem, and restart the machine. Therefore, early detection of defects in machine components that could lead to fabric defects is crucial.

[0005] Needle defects are a major cause of fabric defects. As needles continuously reciprocate to form loops, they come into contact with the yarn, which is under tension. This interaction with the yarn gradually causes needle wear and misalignment, which can lead to fabric defects. Needles can also break due to interaction with the yarn or the mechanism that moves the needles.

[0006] US Patent No. 6,035,669A discloses a method for detecting broken hooks in knitting machine needles. Each needle is slidably received in a cam system with a groove segment formed therein. This allows the shank of a broken needle to be pushed into the groove segment, while the shank of an intact needle is biased against entering the groove segment. A sensor in the groove segment identifies the broken needle shank.

[0007] Chinese patent CN103437061A discloses a method for detecting defects in knitting machine needles. Moving needles sequentially pass in front of the end of an optical fiber emitting LED light. Light reflected from the needle surface is transmitted through the optical fiber to a photodetector, which generates a pulse signal. Changes in the needle's shape or relative position cause a change in the pulse signal's frequency, which is detected and interpreted as a needle defect.

[0008] Chinese patent CN109881356A discloses a method for detecting defects in knitting machine needles using a digital camera. The digital camera captures and processes image frames of a group of knitting needles to crop images of each individual needle and compare their shapes with predetermined shapes for normal, broken, and bent needles.

[0009] International patent WO2020079493A1 discloses a circular knitting machine in which a digital camera is fixed to a rotating drum for continuously monitoring the knitted fabric produced by the machine. The digital camera takes digital images of the knitted fabric continuously produced by the machine. The knitted fabric is illuminated by a lighting device and rotates with the rotating drum. Before being wound into the drum under the machine, the flattened knitted fabric section is photographed by the digital camera. The digital image is processed to automatically identify defective patterns or unevenness in the knitted fabric. Automatic recognition is performed by calculating Gaussian filters, local binary pattern (LBP) algorithms or machine learning techniques, which are adjusted according to the type of fabric produced (single needle, double rib, etc.). Summary of the Invention

[0010] The object of the present invention is to provide a method for detecting defects in the needles of an automatic knitting machine of the type described in the "Technical Field" section, which method can be easily and cheaply implemented without modifying the knitting machine and can automatically and reliably detect various defects in the needles.

[0011] The object of the present invention is achieved by a method for detecting defects in knitting needles of an automatic knitting machine of this type, characterized in that a digital camera is provided to capture digital image frames of a group of knitting needles, in which the knitting needles are adjacent to each other, while the automatic knitting machine is operating to produce knitted fabric; and the method comprises the following steps:

[0012] [a] acquiring, from the digital camera, an image frame containing at least part of the knitting needles in the group while the automatic knitting machine is operating to produce the knitted fabric;

[0013] [b] performing automatic image recognition in the image frames to determine at least one of a first pattern collectively defined by at least some of the knitting needles in the set and a second pattern defined by a length of at least one yarn interacting with at least one of the knitting needles in the set;

[0014] [c] automatically deriving from said first or second pattern at least one parameter related to the presence, position or shape of needles or some of the needles in said group;

[0015] [d] automatically calculating the deviation of the at least one parameter from a predetermined reference value;

[0016] [e] Automatically determine whether there is a defect in the group of knitting needles based on the deviation.

[0017] The term " knitting needle " used in this article refers to a complete knitting needle, including all its parts. For example, when the knitting needle is a latch needle, the term " knitting needle " refers to a knitting needle including a latch needle.

[0018] As will be seen in the detailed description of the embodiments, the method according to the present invention can be easily implemented by simply installing a digital camera and using software to perform automatic image recognition and needle defect determination. While the automatic knitting machine is operating, it is generally not a problem to install the camera in a position where it can capture digital image frames of the appropriate needle group. The camera can be mounted on the knitting machine itself or on an external frame.

[0019] It should be noted that the method according to the present invention is not based on identifying the shape of individual knitting needles, but rather on identifying a first pattern defined by at least some of the needles in a group, and / or a second pattern defined by the length of at least one yarn interacting with at least one needle in the group, contained in an image frame captured by a digital camera. This provides a faster and more reliable process, as there is no need to analyze the needles individually or know their exact shape. Furthermore, the method enables more reliable detection of needle defects, as the pattern can be simple and deviations are easily determined. Furthermore, the method can be easily adapted to various knitting machines equipped with different needle types (e.g., latch needles, compound needles, and proprietary needles) without requiring any component disassembly. The method can be adapted to a specific machine using software by initially selecting a pattern and parameters from an image frame captured by a digital camera while the machine is operating and all needles are known to be free of defects, and setting predetermined reference values ​​for the parameters in that image frame. A key aspect of the present invention is that it not only detects needle defects that could lead to loop defects in the fabric being produced by the knitting machine, but also detects needle defects before the needles cause such loop defects in the fabric. For example, slight deviations of the hook or latch of a knitting needle, or initial wear of certain parts of the needle due to friction with the yarn, which initially do not lead to stitch defects, can be detected by the method according to the invention before they become serious and lead to yarn defects. The present invention thus makes it possible to carry out preventive maintenance in an efficient manner by replacing needles that have been identified as potentially causing yarn defects. The present invention also makes it possible to determine the severity of the defect and issue a corresponding warning signal. A first type of warning signal indicates that the knitting machine must be stopped immediately because a serious defect has been detected in a certain needle, which may have already caused a stitch defect. Another type of warning signal indicates that the knitting machine should be stopped as soon as possible and a certain needle should be checked and eventually replaced because a minor defect has been detected in this needle, which may soon lead to a stitch defect.

[0020] In a preferred embodiment, the first pattern determined in step [b] comprises a plurality of light spots in the image frame, each of which corresponds to light reflections from a portion of a needle in the group. This provides a particularly quick and simple method for determining the first pattern and determining parameters, predetermined reference values, and deviations, since the groups of light spots are easily recognized by the software. This solution is particularly suitable for knitting machine needles, since these needles are typically made of metal with a highly reflective surface and have various curvatures that can form reflective light spots in the image frame. Furthermore, the method according to this solution can be transferred from one knitting machine to another with only minor adjustments, since the first pattern of the group of light spots generated by equivalent portions of the needles can be very similar even if the needles have different shapes. Another advantage of this solution is that it makes it easier to identify which needle is defective.

[0021] The term "flare" as used herein should be interpreted in its ordinary sense: a small area that is noticeably different in color or finish from the surrounding area. A spot can have any shape: it does not necessarily have to be a circular dot.

[0022] Preferably, the parameter automatically derived from the first pattern in step [c] includes at least one of the following: the number of light spots in the first pattern, the shape of the light spots, the area of ​​the light spots, the position of the light spots in the image frame, and the relative distance between two of the light spots. This enables a simple and efficient way to identify significant deviations in the parameters, and thus a defect in a particular knitting needle. For example, when the parameter is the number of light spots in the first pattern, the absence of one of the light spots is easily identifiable and is a clear indicator of a defect in the corresponding knitting needle.

[0023] Preferably, each light spot corresponds to a light reflection on the curved surface of a portion of each knitting needle. These light spots are particularly well defined and separated from each other.

[0024] Preferably, the curved portion of the needle that reflects the light spot is designated as a hook. The hook is adapted to sequentially pick up and release the yarn as the needle moves. The hook is one of the parts of the needle that is most susceptible to wear due to friction with the yarn and is most susceptible to defects during use. Excessive wear on the hook can alter the shape, position, or presence of the light spot, making it easily detectable.

[0025] Preferably, the curved surface portion of the needle that reflects light corresponding to the light spot is located at the free end of the needle. This is the portion of the needle that typically forms the hook, making it easier for a digital camera to capture images of this portion during operation of the knitting machine. Furthermore, when the defect is a tilted or bent needle, the free end is the portion of the needle that experiences the greatest positional variation, resulting in a significant change in the shape, position, or presence of the corresponding light spot, thus enabling better detection of the defect.

[0026] In some preferred embodiments, the knitting needle is a latch needle, and the curved surface portion of the needle that reflects light corresponding to the light spot is selected as a portion of the needle latch. Directly detecting defects in the needle latch makes this method more reliable. Latch failure is a common defect that can prevent the needle from picking up yarn.

[0027] In a preferred embodiment, the second pattern determined in step [b] comprises lines corresponding to light reflections along the length of the yarn, the lines having one or more discontinuities, each corresponding to a section of the yarn passing through the hooks of the needles in the group. Similar to the light spots in the first pattern, the lines in the second pattern provide a particularly quick and easy method for determining the second pattern and determining parameters, predetermined reference values, and deviations, as lines with one or more discontinuities are easily recognized by the software. This solution is also particularly suitable for knitting machine needles, as yarns are generally highly reflective and form clearly recognizable lines. As with the light spots in the first pattern, the method according to this solution can be transferred from one knitting machine to another with only minor adjustments, as the second pattern can be very similar even with different needle distributions. A particular advantage of this solution is that it can reliably identify needle defects that prevent the yarn from passing through the hooks of the needles. For example, in latch needles commonly used in automatic knitting machines, such defects are often caused by a broken or damaged latch that cannot close with the hook end of the needle to form a loop.

[0028] Preferably, in step [b], both the first pattern and the second pattern are determined by automatic image recognition, and in step [c], at least one first parameter is automatically derived for the first pattern, and at least one second parameter different from the first pattern is automatically derived for the second pattern. This provides high-reliability detection of needle defects, as a single defect (e.g., a broken needle) may cause deviations in both the first and second parameters.

[0029] Preferably, the parameter automatically derived from the second pattern in step [c] comprises at least one of the following: the number of discontinuities in a line, the length of said discontinuities, and the position of said discontinuities in the image frame. This allows for a simple and efficient determination of significant deviations in the parameter, and thus of a defect in a particular needle. For example, when the parameter is the number of discontinuities in a line, the absence of one of the discontinuities is easily identifiable and a clear indicator of a defect in the corresponding needle.

[0030] Preferably, before step [b], the image frame (6) obtained in step [a] is converted into a monochrome image, thereby making the automatic recognition process faster. This is an important advantage because it solves the problem of how to complete the first pattern or second pattern recognition process quickly enough in the needle portion contained in the image frame before generating the next image frame. Converting the image into a monochrome image means losing most of the information contained in the image, but this does not weaken the robustness of the method because it is not necessary to know the exact shape of the light spot.

[0031] Preferably, the controlled illumination is focused on said group of needles so that the light reflections producing the light spot are independent of the ambient light at the location of the machine.

[0032] Preferably, the number of knitting needles at least part of which is included in the image frame is 2 to 50, preferably 2 to 30, more preferably 5 to 15. These ranges are optimal for determining the pattern, the first parameter or the second parameter and the predetermined reference value.

[0033] Preferably, when a defect in the group of knitting needles is automatically detected in step [e], the defective needle is automatically identified based on the parameters automatically derived in step [c]. This allows the problem to be solved by directly replacing the defective needle. It also allows inference of which part of the knitted fabric may be defective due to the defective needle, thereby enabling more efficient inspection of the knitted fabric.

[0034] While the method according to the present invention is applicable to various automatic knitting machines, in a preferred embodiment, the automatic knitting machine is a circular knitting machine in which the knitting needles are arranged in a rotating cylinder so that they travel along a circumference coaxial with the rotating cylinder, and the digital camera is arranged statically so that it does not rotate with the rotating cylinder, and the digital image frame includes a portion of the circumference. The digital camera is arranged statically, so it does not rotate with the knitting needle cylinder, and the digital image frame includes a portion of the circumference. In such applications, the method according to the present invention is particularly advantageous. Mounting the digital camera is easy, and capturing a fixed pattern in the image frame is also easy. Depending on the structure of the automatic knitting machine, the digital camera can be mounted inside or outside the rotating cylinder of the knitting needles.

[0035] The present invention also includes a corresponding system for detecting defects in knitting needles of an automatic knitting machine, comprising an automatic knitting machine having a plurality of yarn feeders, a plurality of movable knitting needles, each of the knitting needles being adapted to sequentially pick up and release yarn provided by one of the yarn feeders as the knitting needle moves, and an actuating device for automatically moving the knitting needles according to a preset pattern to produce a knitted fabric using the yarn provided by the yarn feeders; characterized in that it also includes:

[0036] a digital camera configured to capture digital image frames of the groups of knitting needles, the knitting needles being adjacent to each other in the groups, while the automatic knitting machine is operating to produce the knitted fabric;

[0037] a processor connected to the digital camera;

[0038] A computer program comprising instructions, which, when executed by a processor, causes the processor to perform the following steps:

[0039] [a] acquiring, from the digital camera, an image frame containing at least part of the knitting needles in the group while the automatic knitting machine is operating to produce the knitted fabric;

[0040] [b] performing automatic image recognition in the image frames to determine at least one of a first pattern collectively defined by at least some of the needles in the set and a second pattern defined by a length of yarn interacting with at least one needle in the set;

[0041] [c] automatically deriving from said first or second pattern at least one parameter related to the presence, position or shape of needles or parts thereof in said set;

[0042] [d] automatically calculating the deviation of the at least one parameter from a predetermined reference value;

[0043] [e] Based on the deviation, automatically determine whether there is a defect in the group of knitting needles.

[0044] The system optionally has structural features according to the preferred embodiments described above for the method, and the computer program optionally includes instructions for executing the method steps according to the preferred embodiments.

[0045] The invention also comprises the above-mentioned computer program defined in the system description.

[0046] The invention also includes other features related to the details shown in the detailed description and drawings of embodiments of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Advantages and features of the invention appear from the following description, in which a preferred embodiment is described in a non-limiting manner with reference to the accompanying drawings with respect to the scope of the main claims.

[0048] Figure 1 is a schematic diagram of a first embodiment of a system according to the invention, the automatic knitting machine being of the circular type, of the type having a set of vertical needles and a set of sinkers.

[0049] Figure 2 It is a schematic diagram of a group of knitting needles in working condition in an automatic knitting machine.

[0050] Figure 3 View of the upper part of the knitting needle with the latch in the open position.

[0051] Figure 4 Here is a view of the upper part of the same knitting needle with the latch in the closed position.

[0052] Figure 5 A more detailed view of a group of working needles in an automatic knitting machine, from which a digital camera captures image frames; all needles are free of defects.

[0053] Figure 6 corresponds to Figure 5Schematic diagram of a processed image frame showing light spots and lines used to determine whether there are defects in the pattern and knitting needles.

[0054] Figure 7 is equivalent to Figure 5 view, but one of the knitting needles is broken.

[0055] Figure 8 is equivalent to Figure 6 Schematic diagram of the processed image frame, but corresponding to Figure 7 A broken knitting needle is shown.

[0056] Figure 9 is equivalent to Figure 5 view, but the latch of one of the knitting needles is broken.

[0057] Figure 10 is equivalent to Figure 6 Schematic diagram of the processed image frame, but corresponding to Figure 9 A knitting needle latch is shown broken.

[0058] Figure 11 Captured by a digital camera Figure 5 Image frame of knitting needles set.

[0059] Figure 12 corresponds to Figure 11 A processed monochrome image frame of an image frame in which all needles are free of defects is equivalent to Figure 6 Schematic diagram of .

[0060] Figure 13 yes Figure 12 A magnified portion of an image frame that is processed by software to detect defects.

[0061] Figure 14 corresponds to Figure 11 A processed monochrome image frame of an image frame in which one knitting needle is broken is equivalent to Figure 8 Schematic diagram of .

[0062] Figure 15 This relates to a second embodiment of the system according to the present invention. The automatic knitting machine is a circular dial and cylindrical weft knitting machine of the type having a set of vertical and horizontal needles. The figure shows a detailed view of a set of needles in operation, from which a digital camera captures image frames; all needles are free of defects.

[0063] Figure 16 Captured by a digital camera Figure 15 The processed monochrome image frame of the image frame of the knitting needle group.

[0064] Figure 17 Equivalent to Figure 15, which shows an indication of two rectangular areas that are cropped to form an image frame, which is processed by the software to detect defects.

[0065] Figure 18 yes Figure 17 An enlarged view of the upper rectangular area indicated in Figure 15 A magnified portion of an image frame is processed by software to detect defects in the horizontal needle groups.

[0066] Figure 19 yes Figure 17 An enlarged view of the lower rectangular area indicated in Figure 15 A magnified portion of an image frame is processed by software to detect defects in the vertical needle groups.

[0067] Figure 20 is a block diagram of the main steps of the method according to the present invention. DETAILED DESCRIPTION

[0068] Figure 1-14 The present invention relates to a first embodiment of the system and method according to the present invention. In this first embodiment, the automatic knitting machine 1 is a circular knitting machine having a set of vertical needles 3 and a set of sinkers 16 for producing a single jersey tubular fabric 17.

[0069] Figure 1 is a schematic diagram of the system, in which a knitting machine 1 is shown in a schematic cross-sectional view. The system comprises the automatic knitting machine 1, a digital camera 4, a controlled lighting device 13, and a computer system including a processor 15 connected to the digital camera 4. The processor 15 executes software including an automatic image recognition algorithm.

[0070] The automatic knitting machine 1 includes a plurality of yarn feeders 2, a plurality of movable knitting needles 3, each of the knitting needles 3 being adapted to sequentially pick up and release a yarn 7 provided by one of the yarn feeders 2 as the knitting needle 3 moves, and an actuating device for automatically moving the knitting needles 3 according to a preset pattern to manufacture a knitted fabric using the yarn 7 provided by the yarn feeder 2.

[0071] More specifically, the following Figures 1 to 14The automatic knitting machine 1 used in the test is a single-needle circular knitting machine of the CANMARTEX-JUMBERCA brand, equipped with 1,728 knitting needles, a diameter of 30 inches, and a needle gauge of 18 (needles per inch). The knitting needles 3 are arranged in a rotating cylinder 14 so that they move along a circle coaxial with the rotating cylinder 14. The knitting needles 3 interact with a cam statically arranged on a needle disk around the rotating cylinder 14. When the rotating cylinder 14 continues to rotate, the cam causes each knitting needle 3 to move vertically up and down to form a coil. The rotating cylinder 14 driven by a motor and the static cam are the above-mentioned actuators for automatically moving the knitting needles 3. Because the cam is static, each knitting needle 3 that moves along the circumference has a unique position at each point on the circumference. The sinker 16 is arranged between the knitting needles 3 and moves horizontally to control the movement of the fabric when the machine is knitting. Figure 2 Schematic diagram of a group of knitting needles 3 in working state in an automatic knitting machine 1. Figure 3 and Figure 4 As shown, all needles 3 are identical and are latch needles. Each needle 3 has a hook 9 and a pivotable latch 10 at its free top end. As the needle 3 moves up and down, the latch 10 is pushed by the yarn in the loop, causing it to sequentially close and open the hook 9. The hook 9 is adapted to sequentially pick up and release the yarn as the needle 3 moves. Figure 3 and Figure 4 The knitting needle 3 is shown with the needle latch 10 in the open and closed positions. The operation of such a circular knitting machine with needles and sinkers, and the movement of the latch needle to form a stitch, will not be described in detail here, as it is already known to those skilled in the art.

[0072] The digital camera 4 is statically positioned so that it does not rotate with the rotating drum 14. The digital camera 4 is configured to capture digital image frames 5 of a portion of the circumference of the knitting needles 3 while the automatic knitting machine 1 is operating to produce a knitted fabric 17. The image frames 5 captured by the digital camera 4 include a group 6 of adjacent knitting needles 3. The lighting device 13 is a statically positioned lamp that emits light focused on the group 6 of knitting needles 3. The relative position of the lamp with respect to the digital camera 4 and the intensity of the light emitted by the lamp are adjusted to obtain a suitable light reflection on the knitting needles 3 (and preferably also on the length of yarn 7), thereby allowing for pattern recognition in the processed image frames 5 as described below. In the exemplary embodiment shown in the figures, the digital camera 4 and lighting device 13 are fixed to a static support located outside a virtual cylinder extending axially relative to the rotating drum 14. In other embodiments, the digital camera 4 and / or lighting device 13 may be positioned differently, such as by being fixed to a static support inside the virtual cylinder.

[0073] The digital camera used in the following tests is the TIS-DMK-33UX264 model video digital camera, commercialized by the German company The Imaging Source Europe GmbH. It is equipped with a Sony IMX264 sensor, has a resolution of 2448 × 2048 pixels, and can capture video at a frame rate of 35 frames per second (FPS). This sensor uses a global shutter CMOS image capture method, which allows all data to be collected simultaneously without shutter lag. Figure 11 A raw, unprocessed image frame 5 of a group 6 of knitting needles 3 captured by a digital camera 4 is shown.

[0074] The method according to the present invention comprises the following main steps:

[0075] [a] acquiring, from a digital camera 4, an image frame 5 of at least some of the knitting needles 3 of a group 6 of knitted fabric while the automatic knitting machine 1 is operating to produce a knitted fabric;

[0076] [b] performing automatic image recognition in the image frame 5 to determine at least one of a first pattern collectively defined by at least some of the knitting needles 3 in the group 6 and a second pattern defined by a length of at least one yarn 7 interacting with at least one of the knitting needles 3 in the group 6;

[0077] [c] automatically deriving from the first pattern or the second pattern at least one parameter related to the presence, position or shape of the needles 3 or some of the needles 3 in the group 6;

[0078] [d] automatically calculating the deviation of the at least one parameter from a predetermined reference value;

[0079] [e] Based on the deviation, it is automatically determined whether there is a defect in the group 6 of knitting needles 3.

[0080] Figures 5 to 10 is a schematic idealized illustration of an image frame during these steps.

[0081] Figure 5 and Figure 6 The following are respectively shown: an unprocessed image frame 5 of a group 6 of knitting needles 3 in an automatic knitting machine 1 in working condition, taken by a digital camera 4 in step [a], and a corresponding processed image frame 5 used in the subsequent steps of the method. The image frame 5 contains a total of 10 knitting needles 3, but as will be described below, only 8 of them form the first pattern. For clarity, only some of the knitting needles are marked with the number 3 in the figure. It is known that all the knitting needles are free of defects. Before step [b], the knitting needles obtained in step [a] are processed. Figure 5 The image frame 5 is processed by oversaturating it and converting it into a monochrome image. Figure 6 shown.

[0082] In an initial step, when the automatic knitting machine 1 is running and all needles 3 are known to be free of defects, the analysis is similar to Figure 6 The processed image 5 is schematically shown in FIG. Through this initial analysis, the first pattern and / or second pattern and one or more parameters are selected and the corresponding predetermined reference values ​​are set. These first and / or second patterns, parameters and predetermined reference values ​​are imported into the software as settings.

[0083] exist Figure 6 A first pattern for detecting defects in the knitting needles 3 is identified in FIG. This first pattern consists of a plurality of light spots 8, each of which corresponds to the reflection of the light emitted by the lighting device 13 on a portion of the knitting needles 3 of the group 6. In the exemplary embodiment shown in the figures, each light spot 8 corresponds to the reflection of light on the curved surface of the hook 9 at the free end of the knitting needle 3. The first pattern consists of Figure 6 The eight aligned continuous light spots 8 starting from the rightmost light spot 8 are composed of. At least one parameter is selected from the first pattern and Figure 6 The parameter is set to a predetermined reference value, wherein all needles are free of defects. This parameter can be, for example, one of the following:

[0084] Number of light spots 8: The preset reference value is 8 (there must be 8 light spots);

[0085] The shape of the light spot 8: The predetermined reference value is an elliptical shape with the major axis oriented vertically, such as Figure 6 As shown;

[0086] The area of ​​the light spot 8: The predetermined reference value is Figure 6 The area of ​​the light spot 8 measured in the middle;

[0087] The position of the light spot 8: the predetermined reference value is Figure 6 XY coordinates of the center point of the middle light spot 8;

[0088] The relative distance between two light spots 8, for example, the relative distance between each pair of adjacent light spots 8 in the arrangement of light spots 8: the predetermined reference value is Figure 6 The measured spot 8 spacing.

[0089] exist Figure 6 A second pattern for detecting defects in the knitting needles 3 is also identified in FIG. This second pattern is a line 11 corresponding to the reflection of light on the yarn 7, said line 11 having one or more discontinuities 12, each discontinuity 12 corresponding to a section of said yarn 7 passing through the hook 9 of the knitting needles 3 in the group 6. In the exemplary embodiment shown in the figures, the line 11 is Figure 6 The short dashed line in the right area has two discontinuous parts 12. At least one parameter is selected from the second pattern and the Figure 6 The parameter is set to a predetermined reference value, wherein all needles are free of defects. This parameter can be, for example, one of the following:

[0090] Number of discontinuous parts 12: the predetermined reference value is 2 (there must be 2 discontinuous parts);

[0091] The length of the discontinuous portion 12: The predetermined reference value is Figure 6 The length of the discontinuous portion 12 measured in;

[0092] Position of the discontinuous portion 12: The predetermined reference value is Figure 6 The XY coordinates of the center point of the discontinuous portion 12.

[0093] While the automatic knitting machine 1 is operating to produce knitted fabric, the software performs steps [a] through [e] of the method on subsequent image frames 5 captured by the digital camera 4. Steps [b] and [c] are performed by an image recognition algorithm included in the software. The image frames 5 are captured at a suitable temporal frequency, adjusted based on the rotational speed and needle length of the rotating drum 14, to ensure that all knitting needles 3 are in the same position in each image frame 5. The temporal frequency of capturing the image frames 5 is also adjusted so that every knitting needle 3 of the knitting machine 1 is included in at least one image frame 5 and appears in at least one of the first pattern or the second pattern.

[0094] Figure 7 and Figure 8 Respectively Figure 5 and Figure 6 is equivalent, but in this case, as Figure 7 As shown, one of the knitting needles 3 is broken, and the broken knitting needle is the fourth from the left. Figure 7 In the first pattern automatically detected in the processed image frame, the light spot 8 corresponding to the broken needle is missing. The software that performs steps [a]-[e] automatically recognizes this situation and automatically determines in step [e] that there is a defect in the group 6 of needles 3. For example, if the parameters selected in the initial step are the number and position of the light spots 8, then in step [d] the deviation is manifested by the number of light spots being 7 instead of 8, and the fourth light spot 8 from the left is not in its normal position; Figure 7-8 In the example, the position of the light spot 8 completely disappears. The faulty needle 3 can be identified based on the deviation in the position of the light spot 8. On the other hand, the nature and severity of the defect can be inferred from the nature of the deviation. The missing position of the fourth light spot 8 indicates that the fourth needle is defective, and the defect is likely to be a broken needle or a height deviation of the needle. Therefore, the defect of this needle is a critical defect. The software will issue an alarm, prompting the knitting machine to stop immediately and indicate the number of the needle with the critical defect. The needle number refers to the position of the needle in the knitting machine, which is calculated by the software based on the position of the needle in the image frame 5 and the count of the needles 3 passing through the image frame 5. It can also automatically calculate and indicate the area of ​​the knitted fabric that may be affected by the faulty needle.

[0095] The same process can be used to detect defects in the tongue 10 of the knitting needle 3. To this end, the curved portion of the knitting needle 3 where light reflection occurs and corresponds to the light spot 8 is selected as the portion of the tongue 10 of the knitting needle 3. The light intensity emitted by the lighting device 13 and the absolute and relative positions of the lighting device 13 and the digital camera 4 can be adjusted to obtain the light spot 8 corresponding to the light reflection of the portion of the knitting needle tongue 10.

[0096] Figure 9 and Figure 10 Respectively Figure 5 and Figure 6 is equivalent, but in this case, as Figure 9 As shown, the needle latch 10 of the eighth needle from the left is missing. Figure 10 In the processed image frame, the corresponding discontinuous portion 12 in the line 11 of the second pattern is missing. Figure 5-Figure 6 The same procedure is used for the treatment of broken needles in the knitting machine. The absence of the discontinuity 12 is automatically detected as a deviation of the parameter from a predetermined reference value. As a result of step [e], the software issues an alarm, prompting the knitting machine to stop immediately and indicating the number of the needle with the critical defect.

[0097] Non-critical defects on needles 3 can also be automatically detected. For example, if the position, area, or shape of light spot 8 differs slightly from a predetermined reference value, the software will automatically determine in step [e] that the corresponding needle 3 has a non-critical defect and display a message prompting that the needle should be inspected as soon as possible. These non-critical defects could include, for example, moderate wear on the hook end, moderate deflection of the entire needle, or moderate deflection of the latch.

[0098] Figure 11-14 is an example of an actual image frame 5 acquired from a digital camera 4 and processed according to steps [a]-[e] above. Figure 11 is with Figure 5 Equivalent image frame 5. Figure 12 and 14 is with Figure 6 and Figure 8 Equivalent processed monochrome image frame 5. Figure 13 yes Figure 12 Steps [b] and [c] are preferably performed on this magnified area of ​​the image frame 5, which is focused on the area containing the first pattern and / or the second pattern. Figure 13 In the example shown, the enlarged area contains a first pattern formed by eight aligned light spots 8 and a second pattern formed by a line 11 having two discontinuous portions 12 .

[0099] Figures 15 to 20 This relates to a second embodiment of the system according to the invention. This automatic knitting machine is also of the circular type, but of the type having a set of vertical needles and a set of horizontal needles, and is used to produce interlock knitted tubular fabrics. More specifically, the following Figures 15 to 20 The automatic knitting machine used in the test is a double rib circular knitting machine of the CANMARTEX-JUMBERCA brand, equipped with 852×852 needles, a diameter of 17 inches, and a needle gauge of 16G (number of needles per inch). The vertical and horizontal knitting needles 3 are arranged in a rotating cylinder so that they move along a circle coaxial with the rotating cylinder. The knitting needles 3 interact with the cams statically arranged on the needle disk around the knitting needle cylinder, so that each vertical knitting needle 3 moves up and down and each horizontal knitting needle 3 moves forward and backward. The knitting needle cylinder driven by a motor and the static cam are the above-mentioned actuators for automatically moving the knitting needles 3. Since the cam is static, each vertical knitting needle 3 and horizontal knitting needle 3 that moves along the circumference has a unique position at each point of the circumference. All knitting needles 3 are the same and are the above-mentioned latch needles, such as Figure 3 and Figure 4 The operation of this circular knitting machine with vertical and horizontal needles, as well as the movement of the vertical and horizontal latch needles to form the loops, will not be described in detail here, as they are already well known to those skilled in the art.

[0100] This method and system are equivalent to the method and system of the first embodiment described above. The digital camera 4 and controlled lighting device 13 are also statically positioned to capture image frames 5 of the group 6 of knitting needles 3 from the digital camera 4. Steps [a]-[e] are essentially the same, with the only difference being that the group 6 of knitting needles 3 includes both vertical and horizontal needles. The focal axis of the digital camera 4 is preferably located in the 45° plane, which bisects the 90° angle formed by the vertical and horizontal needles.

[0101] Figure 15 is a schematic idealized illustration of an image frame 5 acquired from the digital camera 4 in step [a], with all needles being free of defects. Figure 16 is the corresponding processed image frame 5 used in the subsequent steps of the method. Figure 16 The image frame 5 is obtained from the digital camera 4 in step [a] before step [b]. Figure 15 The equivalent image frame is oversaturated and converted to monochrome. Figure 17 Each of the two rectangular areas shown performs steps [b]-[e] of the method. The upper and lower rectangular areas contain the free hook ends of the horizontal and vertical needles 3, respectively, including the ends of the latches 10. Figure 18 and 19 The enlarged views of the upper and lower rectangular areas are shown respectively. As shown in these figures, the points corresponding to the free hook ends of the knitting needles 3 are clearly distinguishable, so the first pattern can be identified. The method of automatically identifying defects in the knitting needles 3 from the first pattern is equivalent to the method of the first embodiment described above. Figure 18 and 19In the example, the end portion of the latch 10 can also be identified as a smaller light spot adjacent to the larger light spot corresponding to the end of the needle. Both the end portion of the latch 10 and the end of the needle 3 can be included in the first pattern, thereby allowing latch defects to be detected through this first pattern. The second pattern can also be identified by focusing the digital camera 4 on different areas (not shown) in the group 6 of needles 3 where the yarn 7 is fed into the needles 3.

[0102] As described above with respect to the two exemplary embodiments, the system and method according to the present invention can automatically detect at least the following defects in knitting needles:

[0103] Broken knitting needles;

[0104] The hook end of the knitting needle wears out from friction with the yarn;

[0105] Lateral or axial displacement of the hook end of the needle due to needle bending;

[0106] The needle latch of the knitting needle is broken or missing;

[0107] Lateral or axial displacement of the needle latch due to bending of the needle latch or pivotal movement of the needle latch.

[0108] These defects are automatically identified because they cause changes in the light reflection on the needle or yarn, which in turn causes changes in the first pattern or the second pattern recognized by the image recognition software. As described above, the system and method according to the present invention allow for automatic identification of whether a defect is critical (a faulty needle that is likely to cause a defect in the knitted fabric) or non-critical (a defect on an identified needle that may not have caused a defect in the knitted fabric but requires inspection as soon as possible).

[0109] The present invention is not limited to the large diameter circular knitting machines (producing weft-knitted tubular fabrics of equal width and continuous length) described in the above two exemplary embodiments, but is also applicable to other types of automatic knitting machines, such as small diameter circular weft knitting machines and flat knitting machines that produce garment length sequences of knitted fabrics.

Claims

1. A computer-implemented method for detecting defects in knitting needles of an automatic knitting machine (1), the automatic knitting machine (1) comprising a plurality of yarn feeders (2), a plurality of movable knitting needles (3), each of the knitting needles (3) being adapted to sequentially pick up and release yarn supplied by one of the yarn feeders (2) as the knitting needle (3) moves, and an actuating device for automatically moving the knitting needles (3) according to a preset pattern to manufacture a knitted fabric using the yarn supplied by the yarn feeders (2); It is characterized in that a digital camera (4) arranged to capture digital image frames (5) of a group (6) of knitting needles (3) when the automatic knitting machine (1) is in operation to produce the knitted fabric, the knitting needles (3) being adjacent to each other in the group (6); The method includes the following steps performed by a processor: [a] acquiring, from the digital camera (4), an image frame (5) containing at least some of the knitting needles (3) in the group (6) while the automatic knitting machine (1) is operating to produce the knitted fabric; [b] performing automatic image recognition in the image frame (5) to determine at least one of a first pattern collectively defined by at least some of the knitting needles (3) in the group (6) and a second pattern defined by a length of at least one yarn (7) interacting with at least one of the knitting needles (3) in the group (6); [c] automatically deriving from the first pattern or the second pattern at least one parameter related to the presence, position or shape of the knitting needles (3) or some of the knitting needles (3) in the group (6); [d] automatically calculating the deviation of the at least one parameter from a predetermined reference value; [e] Based on the deviation, automatically determine whether there is a defect in the group (6) of knitting needles (3).

2. A computer-implemented method according to claim 1, wherein the first pattern determined in step [b] comprises a plurality of light spots (8) in the image frame (5), each of the light spots (8) corresponding to light reflections of a portion of the knitting needles (3) in the group (6).

3. The computer-implemented method according to claim 2, wherein the parameters automatically derived from the first pattern in step [c] include at least one of the following groups: the number of the light spots (8) in the first pattern, the shape of the light spots (8), the area of ​​the light spots (8), the position of the light spots (8) in the image frame (5), and the relative distance between two of the light spots (8).

4. The computer-implemented method according to any one of claims 2 to 3, wherein each of the light spots (8) corresponds to a reflection of light on a curved surface of a portion of each of the knitting needles (3).

5. The computer-implemented method according to claim 4, wherein the curved portion of the knitting needle (3) corresponding to the light spot (8) for light reflection is selected as a needle hook (9), and the needle hook (9) is suitable for hooking and releasing the yarn in sequence when the knitting needle (3) moves.

6. The computer-implemented method according to any one of claims 4 or 5, wherein the curved surface portion of the knitting needle (3) corresponding to the light spot (8) where light reflection occurs is selected to be located at the free end of the knitting needle (3).

7. The computer-implemented method according to claim 4, wherein the knitting needle (3) is a latch needle, and the curved surface portion of the knitting needle (3) corresponding to the light spot (8) for light reflection is selected as a part of the needle latch (10) of the knitting needle (3).

8. A computer-implemented method according to any one of claims 1 to 7, wherein the second pattern determined in step [b] comprises a line (11) corresponding to the reflection of light over the length of the yarn (7), the line (11) having one or more discontinuous portions (12), each of the discontinuous portions (12) corresponding to a section of the yarn (7) passing through the hook (9) of the knitting needle (3) in the group (6).

9. A computer-implemented method according to claim 8, wherein the parameters automatically derived from the second pattern in step [c] include at least one of the following groups: the number of the discontinuous portions (12), the length of the discontinuous portions (12) and the position of the discontinuous portions (12) in the image frame (5).

10. A computer-implemented method according to any one of claims 1 to 9, wherein prior to step [b], the image frame (6) acquired in step [a] is converted into a monochrome image.

11. Computer-implemented method according to any one of claims 1 to 10, wherein the controlled lighting means (13) is focused on the group (6) of knitting needles (3).

12. A computer-implemented method according to any one of claims 1 to 11, wherein when a defect in the group (6) of knitting needles (3) is automatically detected in step [e], the knitting needles (3) having the defect are automatically identified based on the parameters automatically derived in step [c].

13. A computer-implemented method according to any one of claims 1 to 12, wherein the automatic knitting machine (1) is a circular knitting machine, the knitting needles (3) are arranged in a rotating cylinder (14) so ​​that the knitting needles (3) move along a circumference coaxial with the rotating cylinder (14), and the digital camera (4) is statically arranged so that the digital camera (4) does not rotate with the rotating cylinder (14), and the digital image frame (5) includes a portion of the circumference.

14. A system for detecting defects in knitting needles of an automatic knitting machine (1), comprising an automatic knitting machine (1), the automatic knitting machine (1) having a plurality of yarn feeders (2), a plurality of movable knitting needles (3), each of the knitting needles (3) being adapted to sequentially pick up and release yarn supplied by one of the yarn feeders (2) as the knitting needle (3) moves, and an actuating device for automatically moving the knitting needles (3) according to a preset pattern to manufacture a knitted fabric using the yarn supplied by the yarn feeders (2); It is characterized in that Also includes: a digital camera (4) configured to capture digital image frames (5) of a group (6) of knitting needles (3) when the automatic knitting machine (1) is in operation to produce the knitted fabric, the knitting needles (3) being adjacent to each other in the group (6); a processor (14) connected to the digital camera (4); A computer program comprising instructions which, when executed by the processor (14), cause the processor (14) to perform the following steps: [a] acquiring, from the digital camera (4), an image frame (5) containing at least some of the knitting needles (3) in the group (6) while the automatic knitting machine (1) is operating to produce the knitted fabric; [b] performing automatic image recognition in the image frame (5) to determine at least one of a first pattern collectively defined by at least some of the knitting needles (3) in the group (5) and a second pattern defined by a length of at least one yarn (7) interacting with at least one of the knitting needles (3) in the group (5); [c] automatically deriving from the first pattern or the second pattern at least one parameter related to the presence, position or shape of the knitting needles (3) or some of the knitting needles (3) in the group (6); [d] automatically calculating the deviation of the at least one parameter from a predetermined reference value; [e] Based on the deviation, automatically determine whether there is a defect in the group (6) of knitting needles (3).

15. A system according to claim 14, wherein the instructions of the computer program are configured so that the first pattern determined in step [b] includes a plurality of light spots (8) in the image frame (5), each of the light spots (8) corresponding to light reflections of a portion of the knitting needles (3) in the group (6).

16. The system according to claim 15, wherein the instructions of the computer program are configured so that the parameters automatically derived from the first pattern in step [c] include at least one of the following groups: the number of the light spots (8) in the first pattern, the shape of the light spots (8), the area of ​​the light spots (8), the position of the light spots (8) in the image frame (5), and the relative distance between two of the light spots (8).

17. The system according to any one of claims 15 to 16, wherein the instructions of the computer program are configured to cause each of the light spots (8) to correspond to light reflection on a curved surface of a portion of each of the knitting needles (3).

18. A system according to claim 17, wherein the instructions of the computer program are configured so that the curved surface portion on the knitting needle (3) corresponding to the light spot (8) for light reflection is selected as the needle hook (9), and the needle hook (9) is suitable for hooking and releasing the yarn in sequence when the knitting needle (3) moves.

19. A system according to any one of claims 17 or 18, wherein the instructions of the computer program are configured so that the curved surface portion on the knitting needle (3) corresponding to the light spot (8) for light reflection is selected to be located at the free end of the knitting needle (3).

20. A system according to claim 17, wherein the instructions of the computer program are configured to make the knitting needle (3) a latch needle, and the curved surface portion of the knitting needle (3) corresponding to the light spot (8) for light reflection is selected as a part of the needle latch (10) of the knitting needle (3).

21. A system according to any one of claims 14 to 20, wherein the instructions of the computer program are configured so that the second pattern determined in step [b] includes a line (11) corresponding to the reflection of light along the length of the yarn (7), the line (11) having one or more discontinuous portions (12), each of the discontinuous portions (12) corresponding to a section of the yarn (7) passing through the hook (9) of the knitting needle (3) in the group (6).

22. A system according to claim 21, wherein the instructions of the computer program are configured so that the parameters automatically derived from the second pattern in step [c] include at least one of the following groups: the number of the discontinuous portions (12), the length of the discontinuous portions (12) and the position of the discontinuous portions (12) in the image frame (5).

23. The system according to any one of claims 14 to 22, wherein the instructions of the computer program are configured to convert the image frame (6) acquired in step [a] into a monochrome image before step [b].

24. System according to any one of claims 14 to 23, comprising controlled lighting means (13) focused on the group (6) of knitting needles (3).

25. A system according to any one of claims 14 to 24, wherein the instructions of the computer program are configured to automatically identify the knitting needle (3) having the defect based on the parameters automatically derived in step [c] when a defect in the group (6) of knitting needles (3) is automatically detected in step [e].

26. A system according to any one of claims 14 to 25, wherein the automatic knitting machine (1) is a circular knitting machine, the knitting needles (3) are arranged in a rotating cylinder (14) so ​​that the knitting needles (3) move along a circumference coaxial with the rotating cylinder (14), and the digital camera (4) is statically arranged so that the digital camera (4) does not rotate with the rotating cylinder (14), and the digital image frame (5) includes a portion of the circumference.

27. A computer program for detecting defects in the needles of an automatic knitting machine (1), the computer program comprising instructions which, when executed by a processor (15), cause the processor (15) to perform the following steps: [a] acquiring, from a digital camera (4), image frames (5) of at least some of the knitting needles in a group (6) of knitting needles (3) of an automatic knitting machine (1), wherein the image frames are taken by the digital camera (4) while the automatic knitting machine (1) is operating to produce a knitted fabric; [b] performing automatic image recognition in the image frame (5) to determine at least one of a first pattern collectively defined by at least some of the knitting needles (3) in the group (5) and a second pattern defined by a length of at least one yarn (7) interacting with at least one of the knitting needles (3) in the group (5); [c] automatically deriving from the first pattern or the second pattern at least one parameter related to the state of the knitting needles (3) or some of the knitting needles (3) in the group (6); [d] automatically calculating the deviation of the at least one parameter from a predetermined reference value; [e] Based on the deviation, automatically determine whether there is a defect in the group (6) of knitting needles (3).

Images