Circular knitting machine that can indicate the status of the knitting machine in real time based on the status of the fabric surface

Abstract

A circular knitting machine that can indicate the status of the knitting machine in real time based on the status of the fabric surface, comprising a needle cylinder, a camera module, an information processing unit and an encoder. The camera module can take a picture of a fabric when it is falling. A camera head of the camera module does not rotate with the needle cylinder, and the camera shooting timing is controlled by multiple shooting signals. The information processing unit receives multiple image data generated by the camera module and compares the images. When two consecutive images on the same vertical line of the image data differ, the knitting machine status is indicated. The encoder generates multiple pulse signals when the needle cylinder rotates. The encoder outputs these pulse signals to the camera module or the information processing unit, and the receiver counts these pulse signals to generate these shooting signals.

Description

Technical Field

[0001] The invention relates to a circular knitting machine capable of judging the state of the knitting machine, and in particular to a circular knitting machine capable of prompting the state of the knitting machine in real time based on the state of a fabric surface. Background Art

[0002] The knitting quality of a circular knitting machine refers to whether there are defects in the surface state of the fabric, and the generation of defects is closely related to the pin states of multiple knitting needles on the circular knitting machine. When the lower edge of the pin of one of these knitting needles is worn too much, the push-up height of one of these knitting needles will not be able to reach the expected height, resulting in the following situations: the needle tongue cannot be fully opened; the old yarn loop is not cleared (the knitting needle knits with double yarn errors); the needle hook cannot hook the new yarn (holes appear on the fabric). On the other hand, when the upper edge of the pin of one of these knitting needles is worn too much, the following situations will occur: the old yarn loop cannot be completely removed (in addition to holes in the fabric, abnormal line textures can be visually found on the fabric after weaving); the knitted yarn loop is smaller than the yarn loop knitted in the previous period (abnormal line textures can be visually found on the fabric).

[0003] Although there are many technical solutions for fabric quality detection, such as those disclosed in CN102778414A, CN102967606A, CN103451846A, CN103604809A, CN108364291A, CN108921819A, CN109696442A, CN110389130A and other patents, the time point for the quality detection of knitting in the previous technology is after the circular knitting machine completes knitting. Even if the fabric is found to have defects, it is impossible to judge which part of the knitting needles of the circular knitting machine needs to be replaced. As a result, after the current producers learn that the fabric has defects, the general practice is to eliminate and replace all the knitting needles on the circular knitting machine. The general number of knitting needles on the circular knitting machine is 1500 to 2640 needles, and the number of knitting needles eliminated and replaced at one time is too large, and it is not in line with the time cost to conduct a status assessment of all the knitting needles. This will not only cause a waste of resources, but also increase the production costs of producers in disguise.

[0004] In addition to the above, although there are also technical solutions for quality inspection of fabrics on circular knitting machines, such as those disclosed in CN105044121A, CN110188806A, and CN111812108A, the above technologies still only focus on the inspection of fabric quality, and do not associate the inspection results with the state of the loom, resulting in the manufacturer still having to adopt the method of completely eliminating and replacing knitting needles to solve the problem of abnormal knitting quality. Furthermore, the technical solutions disclosed in the above patents all require the calculator that mainly performs identification to perform feature learning on the state of the fabric surface before making a judgment. In fact, circular knitting machines are not designed to weave only a single type of fabric. Feature learning is only performed at the beginning of the production of a type of fabric. It is likely that feature learning will not be completed until the entire production of the fabric is completed, which is obviously inconsistent with actual needs. Furthermore, if the plan is to complete feature learning before the production of the fabric, it means that the circular knitting machine must first weave a large amount of fabric before actual production for the computer to perform feature learning. For the producer, a large amount of woven fabric will be regarded as waste, and the production cost is bound to increase significantly, which does not meet actual needs.

[0005] Although weaving factories nowadays use manual inspections on knitting machines, the aforementioned weaving defects cannot be discovered through simple inspections. Instead, the weaving defects may occur but the knitting machines continue to knit defective fabrics. The weaving factories have to wait until the fabrics are dropped and transported in rolls to the quality inspection department for inspection before they can discover the weaving defects. At this time, the weaving factories can only regard the fabrics with weaving defects as waste, resulting in a waste of resources. Summary of the invention

[0006] The main purpose of the present invention is to solve the problem that a circular knitting machine cannot know the specific position to be repaired from the fabric surface state detection result.

[0007] The secondary purpose of the present invention is to solve the problem that defect detection cannot be performed synchronously during the cloth dropping process of a conventional circular knitting machine.

[0008] To achieve the above-mentioned purpose, the present invention provides a circular knitting machine that can prompt the status of the knitting machine in real time based on the status of the fabric surface. The circular knitting machine has a corner support and a needle cylinder that is driven to rotate relative to the corner support. When the circular knitting machine is started, a fabric woven is dropped from the side of the needle cylinder away from the corner support. The circular knitting machine includes a camera module, an information processing unit and an encoder. The camera module can shoot the fabric when it is dropped and generate multiple image data. A camera of the camera module does not rotate with the needle cylinder and the camera is controlled by multiple shooting signals to shoot the fabric when it is dropped. The information processing unit receives these image data and compares each of these image data with another image data located on the same vertical line among the image data shot by the needle cylinder at least one previous rotation. When there is a difference, the knitting machine status is prompted. The encoder generates multiple pulse signals when the syringe rotates, and the encoder outputs these pulse signals to the camera module or the information processing component. The receiver counts these pulse signals to generate these shooting signals. The count value used by the receiver is a factor of the total number of pulse signals generated when the syringe rotates one circle.

[0009] In one embodiment, the value of the number of shots taken by the camera module can evenly divide the central angle of the circle and can evenly divide the total number of knitting needles included in the needle cylinder.

[0010] In one embodiment, the image data include the same number of transverse loop images, and each of the transverse loop images corresponds to one of the plurality of knitting needles on the circular knitting machine.

[0011] In one embodiment, the image data include the same number of longitudinal loop images, and the total length of the longitudinal loop images added together is proportional to the amount of fabric dropped when the needle cylinder rotates one circle.

[0012] In one embodiment, the circular knitting machine includes a mounting arm that suspends the camera within the area surrounded by the needle cylinder.

[0013] In one embodiment, the mounting arm is offset from the center of the area surrounded by the syringe.

[0014] In one embodiment, the information processing unit compares each color pixel on the image data.

[0015] According to the aforementioned implementation of the present invention, compared with the commonly used one, it has the following characteristics: after the camera module of the present invention is installed, the camera included therein will not rotate with the needle cylinder, and the shooting trigger of the camera is based on the shooting signals generated by the counting of these pulse signals by the camera module or the information processing unit, and the camera shoots the fabric in the process of dropping the fabric to generate these image data. The information processing unit does not need to perform the deep learning required for the defect judgment of the fabric in advance, but only compares each of these image data with another image data located on the same vertical line in the image data taken by the needle cylinder at least one previous rotation, which greatly reduces the computing power required by the information processing unit, which means a reduction in hardware requirements. However, although the present invention reduces the required computing power, it does not reduce the judgment efficiency. On the contrary, it can judge the defects of the fabric more quickly than the prior art, so that the operator can immediately eliminate the cause. In addition, the present invention can prompt the operator which part of the circular knitting machine has a problem and should be replaced based on only one of the abnormalities in the image data, specifically solving the problem of resource waste caused by the prior art that the knitting needles included in the circular knitting machine cannot be checked one by one and can only be replaced in batches. Furthermore, the present invention can detect defects on the fabric in real time, allowing the operator to stop the circular knitting machine and prevent the continuous production of defective fabrics in real time. Production can be resumed only after the cause of the defect is eliminated, avoiding the current implementation process in which defect judgment is only performed after the fabric is completed, resulting in excessive waste. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of the partial structure of a circular knitting machine according to an embodiment of the present invention.

[0017] Figure 2 It is a cross-sectional schematic diagram of a partial structure of a circular knitting machine according to an embodiment of the present invention.

[0018] Figure 3 Schematic diagram of an implementation unit of a circular knitting machine according to an embodiment of the present invention.

[0019] Figure 4 Schematic diagram of an implementation unit of a circular knitting machine according to another embodiment of the present invention.

[0020] Figure 5 Schematic diagram of a fabric according to an embodiment of the present invention.

[0021] Figure 6 FIG. 4 is a schematic diagram of image data containing fabric defects according to an embodiment of the present invention. DETAILED DESCRIPTION

[0022] The detailed description and technical contents of the present invention are now described as follows with reference to the accompanying drawings:

[0023] See also Figures 1 to 4The present invention provides a circular knitting machine 20, which can prompt the knitting machine status in real time based on the fabric surface status of a fabric 30 generated during knitting. Here, the implementation basis of the circular knitting machine 20 is first described. The circular knitting machine 20 has a corner support 21, an encoder 22, and a needle cylinder 23. The basic configuration of the corner support 21 and the needle cylinder 23 is circular. The needle cylinder 23 is arranged on the inner circle of the corner support 21. The needle cylinder 23 is driven by a driving member (not shown) to rotate relative to the corner support 21. When the needle cylinder 23 rotates, the multiple knitting needles 231 on the needle cylinder 23 are guided by the multiple corner groups 211 on the corner support 21 to generate a knitting action. In addition, when the circular knitting machine 20 is started, the fabric 30 woven is doffed from the side of the needle cylinder 23 away from the corner support 21. The encoder 22 generates a plurality of pulse signals 221 when the needle cylinder 23 rotates. The pulse signals 221 are generated regularly and continuously. The pulse signals 221 generated by the encoder 22 when the needle cylinder 23 rotates one circle are fixed. In the prior art of the circular knitting machine 20, the pulse signals 221 are only used to determine the rotation angle of the needle cylinder 23.

[0024] As mentioned above, the circular knitting machine 20 of the present invention further comprises a camera module 24 and an information processing unit 25. The camera module 24 can take a picture of the fabric 30 during the fabric drop and generate a plurality of image data 241. More specifically, a camera 242 of the camera module 24 is arranged at a position where the fabric 30 during the fabric drop can be photographed. In one embodiment, the circular knitting machine 20 comprises a mounting arm 26 that allows the camera 242 to be suspended in the area surrounded by the needle cylinder 23. The mounting arm 26 can be arranged on a yarn feeding ring (not shown in the figure) of the circular knitting machine 20, or on an external structure that does not belong to the circular knitting machine 20. Considering the shooting distance and photographic quality of the camera 242, the mounting arm 26 is not limited to being located at the center of the area surrounded by the needle cylinder 23, but can be arranged off the center. The camera 242 of the present invention does not rotate with the needle cylinder 23, that is, the camera 242 is fixed at one place and photographs objects passing by the camera 242. The camera 242 of the present invention does not take pictures of the fabric 30 all the time, so no video data is generated. The camera 242 takes pictures of the fabric 30 in the process of falling cloth only after receiving a plurality of shooting signals 27 .

[0025] Please refer again Figure 3 to Figure 4Before explaining the shooting signal 27, it is necessary to explain that the camera module 24, in one embodiment, has a computing unit 243, and the computing unit 243 can perform computing on the received information based on the memorized computing program. On the other hand, the information processing unit 25 can be implemented using the central control module of the circular knitting machine 20. The information processing unit 25 is connected to the camera module 24 in a wired or wireless manner to receive the image data 241 generated by the camera module 24. Back to the description of the shooting signal 27, the pulse signal 221 generated by the encoder 22 is not only used for the original work control of the circular knitting machine 20, but also output to the camera module 24 or the information processing unit 25. It should be noted that the prerequisite for selecting the output to the camera module 24 is that the camera module 24 itself has the computing unit 243. The camera module 24 or the information processing unit 25 serves as a receiver of the pulse signals 221, that is, it counts the pulse signals 221 to generate the shooting signal 27, and the count value used by the receiver is a factor of the total number of pulse signals 221 generated when the syringe 23 rotates one circle. For example, the total number of pulse signals 221 generated by the encoder 22 when the syringe 23 rotates one circle is 2640, and the count value can be selected to be a value that is sufficient to divide 2640, such as 88. In this way, the camera module 24 or the information processing unit 25 as a receiver will generate the shooting signal 27 every time it receives 88 of the pulse signals 221. Therefore, the number of shots of the camera module 24 is the quotient obtained by dividing the count value by the total number of pulse signals 221. For the example disclosed in the previous text, it is 30 shots. However, in order to make the camera module 24 take the same number of shots in each rotation of the needle cylinder 23, the number of shots needs to be implemented with the value of the average central angle. It is preferred that the number of shots can be divided by the total number of the knitting needles 231 included in the needle cylinder 23, so that the image data 241 evenly include the knitting needles 231, and the information processing unit 25 can prompt the operator which part of the knitting needles 231 is abnormal according to one of the abnormalities in the image data 241. To continue with the above example, the total number of the knitting needles 231 is 2640, and the camera module 24 takes 30 shots, which is equivalent to the camera module 24 taking a shot once every 12 degrees of rotation of the needle cylinder 23, so the total number of the image data 241 generated by one rotation of the needle cylinder 23 is 30. Thus, each of the image data 241 includes 88 of the knitting needles 231. That is, if one of the image data 241 includes the 1st to the 88th needles of the knitting needles 231, the next image data 241 will include the 89th to the 176th needles of the knitting needles 231. Figure 5, the number of transverse loop images 601 included in these image data 241 is the same, and each of these transverse loop images 601 corresponds to one of the knitting needles 231 on the circular knitting machine 20. Furthermore, the number of longitudinal loop images 602 included in these image data 241 is the same, and the total length of these longitudinal loop images 602 added together is proportional to the amount of fabric dropped when the needle cylinder 23 rotates one circle. In this way, when inspecting in the future, the object to be repaired can be quickly found based on one of the image data 241 that is judged to be different.

[0026] In addition, the information processing unit 25 performs image comparison on the image data 241 based on a pre-stored execution program. The information processing unit 25 causes each of the image data 241 (eg Figure 5 The A) indicated in FIG. 1 is only the same as another one (such as Figure 5 B) indicated in the figure, in layman's terms, if one of the image data 241 to be compared is the 24th image data taken by the camera module 24 during the second rotation of the needle cylinder 23, the one to be compared with the aforementioned one of the image data 241 is the 24th image data taken by the camera module 24 during the first rotation of the needle cylinder 23. The aforementioned two of the image data 241 are continuous on the vertical line 50, indicating that the fabric 30 is continuously dropped during the knitting process. After the information processing unit 25 compares the aforementioned two, once a difference is found, it prompts the knitting machine status. Specifically, the comparison referred to in the present invention is to compare the color pixels of the two to be compared in the image data 241. Figure 6 For example, Figure 6 A state in which the fabric 30 is defective is Figure 6A circle of sparse areas 603 can be directly observed, and the rest are a circle of normal areas 604. The image of the circle of sparse areas 603 is different from the image of the circle of normal areas 604. These image data 241 are all electronic data, which means that the image data 241 has differences in the color pixels constituting the same part. For example, these image data 241 are three primary color light modes (RGB) that define the color of each color pixel. Under normal conditions, the RGB parameters of the color pixels on the same part of these image data are: R227, G23, B13. When the RGB parameters of the color pixels on the same part of one of these image data are changed to: R41, G36, B33, the information processing unit 25 can determine that there is a difference. It should be understood that the examples mentioned above are only for illustration. In fact, the information processing component 25 can use a wide range of image feature values ​​to perform approximate or identical comparison. The technical method of image comparison is not the focus of the present invention and can be implemented by any existing technology that can achieve the same effect.

[0027] Furthermore, if Figure 6 The latest shooting information of the camera module 24 is obtained by the information processing unit 25. Figure 6 Another image data 241 (not shown in the figure) located on the same vertical line 50 and taken when the needle cylinder 23 rotates in the previous circle is compared. If the aforementioned other image data 241 is normal knitting, there will be no loop sparse area 603. However, the color pixels of the loop sparse area 603 are different from the color pixels of the loop normal area 604. After the information processing unit 25 compares all parts of the aforementioned two of these image data 241, if the difference (color difference) between the two of these image data 241 is greater than the allowable value, the information processing unit 25 determines that the fabric 30 has quality defects, and then prompts the knitting machine status. The prompt referred to in this article can be achieved by various technical solutions, such as driving a prompt light or a prompt bell on the circular knitting machine 20 to produce an action, or the information processing unit 25 transmits a signal to an information device 28 via wired or wireless means for an operator to read. The aforementioned information device can be a terminal device held by the operator, or a server that can be connected to the terminal device for communication. In addition to the above, the information processing unit 25 can also be set to immediately request the encoder 22 to stop sending a signal to drive the needle cylinder 23 when an abnormality is detected in the fabric state, so as to temporarily shut down the circular knitting machine 20 to avoid the poor quality fabric 30 from being continuously woven.

[0028]

Description of Figure Numbers

[0029] 20.............Circular knitting machine

[0030] 21.............Mountain corner bearing

[0031] 211.............Shanjiao Group

[0032] 22.............Encoder

[0033] 221.............Pulse signal

[0034] 23.............Syringe

[0035] 231.............knitting needles

[0036] 24.............Camera module

[0037] 241.............Image data

[0038] 242.............Camera

[0039] 243.............Operational unit

[0040] 25........Information Processing

[0041] 26.............Mounting arm

[0042] 27.............Shooting signal

[0043] 28.............Information equipment

[0044] 30.............Fabric

[0045] 50.............vertical line

[0046] 601.............Horizontal ring image

[0047] 602........Vertical ring image

[0048] 603.............Ring sparse area

[0049] 604.............Ring Normal Area

[0050] A, B.............Comparator

Claims

1. A circular knitting machine for indicating the state of the knitting machine in real time based on the state of the fabric surface, the circular knitting machine having a corner support and a needle cylinder driven to rotate relative to the corner support, a fabric woven by the circular knitting machine when it is started falls from the side of the needle cylinder away from the corner support, characterized in that: The circular knitting machine comprises: A camera module, which can shoot the fabric when it is falling and generate a plurality of image data, wherein a camera of the camera module does not rotate with the needle cylinder and the camera is controlled by a plurality of shooting signals to shoot the fabric when it is falling; an information processing unit, receiving the plurality of image data, and making each of the plurality of image data be compared with only another image data located on the same vertical line among the image data taken by the needle cylinder at least one previous rotation, and prompting the knitting machine status when a difference occurs; and An encoder generates a plurality of pulse signals when the syringe rotates. The encoder outputs the plurality of pulse signals to the camera module or the information processing component. A receiver counts the plurality of pulse signals to generate the plurality of shooting signals. The count value used by the receiver is a factor of the total number of pulse signals generated when the syringe rotates one circle.

2. The circular knitting machine for indicating the status of the knitting machine in real time based on the status of the fabric surface according to claim 1, characterized in that: The numerical value of the number of shots taken by the camera module can evenly divide the central angle of a circle and can evenly divide the total number of knitting needles included in the needle cylinder.

3. The circular knitting machine for indicating the knitting machine status in real time based on the fabric surface status according to claim 1 or 2, characterized in that: The plurality of image data contain the same number of transverse loop images, and each of the transverse loop images corresponds to one of the plurality of knitting needles on the circular knitting machine.

4. The circular knitting machine for indicating the status of the knitting machine in real time based on the status of the fabric surface according to claim 3, characterized in that: The plurality of image data contain the same number of longitudinal loop images, and the total length of these longitudinal loop images added together is proportional to the amount of cloth dropped when the needle cylinder rotates one circle.

5. The circular knitting machine for indicating the status of the knitting machine in real time based on the status of the fabric surface according to claim 4, characterized in that: The circular knitting machine includes a mounting arm that suspends the camera within an area surrounded by the needle cylinder.

6. The circular knitting machine for indicating the status of the knitting machine in real time based on the status of the fabric surface according to claim 5, characterized in that: The mounting arm is offset from the center of the area surrounded by the syringe.

7. The circular knitting machine for indicating the status of the knitting machine in real time based on the status of the fabric surface according to claim 1, characterized in that: The circular knitting machine includes a mounting arm that suspends the camera within an area surrounded by the needle cylinder.

8. The circular knitting machine for indicating the status of the knitting machine in real time based on the status of the fabric surface according to claim 7, characterized in that: The mounting arm is offset from the center of the area surrounded by the syringe.

9. The circular knitting machine for indicating the status of the knitting machine in real time based on the status of the fabric surface according to claim 7 or 8, characterized in that: The information processing unit compares each color pixel on the plurality of image data.

Images