EP3926084A1

EP3926084A1 - Knitting machine and defect detection system

Info

Publication number: EP3926084A1
Application number: EP21178777.5A
Authority: EP
Inventors: Takahiro Kobayashi
Original assignee: Shima Seiki Mfg Ltd
Current assignee: Shima Seiki Mfg Ltd
Priority date: 2020-06-18
Filing date: 2021-06-10
Publication date: 2021-12-22
Anticipated expiration: 2041-06-10
Also published as: KR20210156781A; KR102573656B1; JP2021195698A; EP3926084B1; CN113818143B; CN113818143A

Abstract

A knitting machine 1 includes a knitting needle 11 configured to knit a knitted fabric K, an image capturer 22 configured to image a stitch held on the knitting needle 11, a determiner 60 configured to compare a state of the stitch imaged by the image capturer 22 with a state of the stitch in a normal state and perform a quality determination of the stitch, and a position storage 60 configured to store a position of the stitch imaged by the image capturer 22.

Description

TECHNICAL FIELD

The disclosure relates to a technique of a knitting machine and a defect detection system capable of detecting defective knitting of a knitted fabric.

BACKGROUND ART

A convention technique of detecting defective knitting of a knitted fabric is known. The technique is disclosed in, for example, JP S57-183469 A .
JP S57-183469 A discloses a technique of comparing an image of a knit fabric acquired using an image capturer with an image of a normal knit fabric stored in advance in a storage and detecting presence or absence of occurrence of defective knitted fabric (knitting defect).

SUMMARY OF INVENTION

However, the technique described in JP S57-183469 A is configured to detect a defect of a completed knit fabric, and thus unnecessary cost (power, yarn, knitting time, and the like of a knitting machine that has knitted the knit fabric) occurs.
Furthermore, for example, in a knitted fabric that contracts in a width direction such as a rib knitted fabric, a defective knitted fabric may be difficult to detect even if the fabric is imaged.
The disclosure has been made in view of the above circumstances, and an object of the disclosure is to provide a knitting machine and a defect detection system capable of detecting defective knitting at an early stage and with high accuracy.
An object of the disclosure is as described above, and hereinafter, measures to take in order to achieve this object will be described.
A knitting machine of the disclosure includes a knitting needle configured to knit a knitted fabric, an image capturer configured to image a stitch held on the knitting needle, a determiner configured to compare a state of the stitch imaged by the image capturer with a state of the stitch in a normal state and perform a quality determination of the stitch, and a position storage configured to store a position of the stitch imaged by the image capturer.
This configuration allows defective knitting to be detected at an early stage and with high accuracy.
The knitting needle may be disposed at each of needle beds disposed so as to face each other across a needle bed gap, and the image capturer may image a stitch held on the knitting needle from above the needle bed gap of the needle beds disposed to face each other.
This configuration allows stitches to be easily imaged, and eventually, defective knitting to be accurately detected.
Further, the image capturer may capture, from a first side of the needle beds disposed to face each other, an image of a stitch held on the knitting needle disposed at a second side of the needle beds.
This configuration allows stitches to be more easily imaged and allows defective knitting to be more accurately detected.
Further, the image capturer may be provided so as to be movable along the needle beds.
This configuration allows stitches at a plurality of places can be imaged by one image capturer.
The image capturer may be fixed at such a position as to image a stitch held on the knitting needle, and may image the stitch in association with a timing of formation of the stitch with the knitting needle.
This configuration can simplify the configuration. Further, an image of the stitches can be acquired at an appropriate timing.
Further, a defect detection system of the disclosure includes an image capturer configured to image a stitch held on the knitting needle configured to knit a knitted fabric, a determiner configured to compare a state of the stitch imaged by the image capturer with a state of the stitch in a normal state and perform a quality determination of the stitch, and a position storage configured to store a position of the stitch imaged by the image capturer.
This configuration allows defective knitting to be detected at an early stage and with high accuracy.
As an effect of the disclosure, an effect of being able to detect defective knitting at an early stage and with high accuracy is obtained.

BRIEF DESCRIPTION OF DRAWINGS

Fig. 1 is a front view illustrating an overall configuration of a flat knitting machine according to one embodiment of the disclosure;
Fig. 2 is a side view illustrating a configuration of needle beds and carriages;
Fig. 3 is a plan view illustrating the configuration of the needle beds and the carriages;
Fig. 4 is a block diagram illustrating a configuration related to control of the flat knitting machine;
Fig. 5 is a schematic plan view illustrating a cam mechanism provided in the carriage;
Fig. 6 is a schematic plan view of the carriages illustrating arrangements of image capturers and illuminators;
Fig. 7A is a plan view illustrating a configuration of the image capturer; Fig. 7B is a side view illustrating the configuration of the image capturer;
Fig. 8 is a partial sectional plan view illustrating a configuration of the illuminator;
Fig. 9 is a plan view illustrating a relative positional relationship between the image capturer and the illuminators;
Fig. 10A is a diagram illustrating the image capturers operating when the carriages move to the right; Fig. 10B is a diagram illustrating the image capturers operating when the carriages move to the left;
Fig. 11 is a schematic view illustrating processing of a controller for knitting each course;
Fig. 12 is a diagram illustrating an example of an image imaged by the image capturer and a normal image;
Fig. 13A is a diagram illustrating an example in which the image capturers are disposed below the needle beds; and Fig. 13B is a diagram illustrating an example in which the image capturers are disposed above and below the needle beds.

DESCRIPTION OF EMBODIMENTS

Hereinafter, directions indicated by arrows U, D, F, B, L, and R in the drawings are defined as up, down, front, rear, left, and right, respectively, and descriptions will be made accordingly. In the drawings, for convenience of description, illustration of some members may be omitted appropriately.
First, an overall configuration of a flat knitting machine 1 according to one embodiment of the disclosure will be described.
As illustrated in Figs. 1 to 4, the flat knitting machine 1 mainly includes a needle bed 10, a carriage 20, a yarn guide rail 30, a servomotor 40, a yarn stand base 50, and a controller 60.
The needle beds 10 illustrated in Figs. 1 to 3 are disposed so as to face each other across a needle bed gap S at the front and rear. The front and rear needle beds 10 are disposed in an inverted V shape in a side view so as to be inclined upward toward a center in the front and rear direction (direction facing each other) (see Fig. 2). Each of the needle beds 10 is provided with a large number of knitting needles 11 disposed side by side along a longitudinal direction (left and right direction) of each of the needle beds 10. The front and rear needle beds 10 can relatively move to the left and right when delivering and receiving (transferring) of a stitch are carried out between the front and rear needle beds 10.
A pair of front and rear carriages 20 are disposed to face the front and rear needle beds 10 from above. The front and rear carriages 20 are connected by a bridge 20a disposed to straddle the plurality of yarn guide rails 30. The carriages 20 can reciprocate along the longitudinal direction of the needle beds 10 by the servomotor 40 (see Fig. 4). The carriages 20 are provided with a needle selecting mechanism (not shown) and a cam mechanism 21 for selectively operating the knitting needles 11 of the needle beds 10. Each of the carriages 20 is also provided with an image capturer 22 and an illuminator 23 for imaging stitch held on the knitting needles 11.
The plurality of yarn guide rails 30 shown in Figs. 1 and 2 are disposed above the needle bed gap S so as to extend along the longitudinal direction of the needle beds 10. A yarn carrier 31 (see Fig. 1) configured to feed a knitting yarn Y is supported on the yarn guide rails 30 so as to be movable.
The yarn stand base 50 illustrated in Fig. 1 is provided with a yarn cone 51 around which the knitting yarn Y is wound. The knitting yarn Y from the yarn cone 51 is fed to the yarn carrier 31 through an appropriate yarn feeding path.
The controller 60 shown in Fig. 4 is configured to control an operation of the flat knitting machine 1. The controller 60 includes a calculation processor such as a CPU (central processing unit) and a storage such as a RAM (random access memory) and a ROM (read only memory). The storage of the controller 60 stores various information, programs, and the like used for controlling the flat knitting machine 1. The controller 60 is provided at an appropriate location (for example, in a body of the flat knitting machine 1 (below the rear needle bed 10) of the flat knitting machine 1.
The controller 60 is connected to the servomotor 40 and can control an operation of the servomotor 40. The controller 60 can arbitrarily move the carriages 20 by controlling the operation of the servomotor 40. The controller 60 can detect a position of the carriages 20 on the basis of a rotational speed of the servomotor 40.
The controller 60 is connected to the carriages 20 (specifically, the cam mechanism 21, the image capturers 22, and the illuminators 23) and can control an operation of the carriages 20.
The controller 60 controls each part of the flat knitting machine 1 on the basis of knitting data created in advance and the like. Specifically, the controller 60 can reciprocate the carriages 20 along the longitudinal direction of the needle beds 10 by controlling the operation of the servomotor 40. In this case, knitting operations such as knit, tuck, miss, and the like, and the transfer of the stitches between the front and rear needle beds 10 can be performed by moving the knitting needles 11 forward and rearward with respect to the needle bed gap S by the cam mechanism 21 and the like mounted on the carriages 20. A knitted fabric K is knitted by repeating such reciprocating movement of the carriages 20.
The controller 60 can image stitches held on the knitting needles 11 with the image capturers 22. The controller 60 can detect a defective knitting on the basis of an image (shape of the stitch) imaged by the image capturers 22.
Hereinafter, a configuration of the carriages 20 will be described more specifically.
In the following description, among the front and rear needle beds 10, the front needle bed 10 may be referred to as a front needle bed 10F, and the rear needle bed 10 may be referred to as a rear needle bed 10B. In the following description, among the front and rear carriages 20, the front carriage 20 may be referred to as a front carriage 20F, and the rear carriage 20 may be referred to as a rear carriage 20B.
First, a configuration of the cam mechanism 21 will be described with reference to Fig. 5.
As illustrated in Fig. 5, the front carriage 20F is provided with three cam mechanisms 21 for advancing and retracting the knitting needles 11. Specifically, a first transfer cam mechanism 21A, a knit cam mechanism 21B, and a second transfer cam mechanism 21C are disposed along a moving direction (left and right direction) of the carriages 20.
The knit cam mechanism 21B is configured to form stitches using the knitting yarn Y fed from the yarn carrier 31. The first transfer cam mechanism 21A and the second transfer cam mechanism 21C are configured to transfer stitches between the knitting needles 11 of the front and rear needle beds 10. In the embodiment, the first transfer cam mechanism 21A, the knit cam mechanism 21B, and the second transfer cam mechanism 21C are disposed in that order from left to right.
Each of the cam mechanisms 21 can guide butts of the knitting needles 11 selected on the basis of knitting data along an advancing and retracting track L to advance and retract the knitting needles 11. Thus, the stitches using the knitting yarn Y can be formed and transferred.
For example, in the embodiment, when the carriages 20 move to the right as indicated by an arrow in Fig. 5, the knit cam mechanism 21B serves as a preceding system and forms stitches. In this case, the first transfer cam mechanism 21A serves as a succeeding system and transfers the stitches formed by the knit cam mechanism 21B. In this case, the second transfer cam mechanism 21C does not transfer the stitches.
On the contrary, when the carriages 20 move to the left, the knit cam mechanism 21B serves as the preceding system and forms the stitches. In this case, the second transfer cam mechanism 21C serves as a succeeding system and transfers the stitches formed by the knit cam mechanism 21B. In this case, the first transfer cam mechanism 21A does not transfer the stitches.
However, such an operation of each cam mechanism 21 is an example, and stitches can be transferred, formed, and then transferred sequentially using, for example, three cam mechanisms 21.
The front carriage 20F has been described with reference to Fig. 5, but as illustrated in Fig. 6, the rear carriage 20B is substantially longitudinally symmetric with the front carriage 20F (includes the first transfer cam mechanism 21A and the like similarly to the front carriage 20F), and thus a detailed description of the rear carriage 20B will be omitted.
Next, a configuration of the image capturers 22 will be described with reference to Fig. 7.
As illustrated in Fig. 7, each of the image capturers 22 mainly includes a support member 22a, a camera 22b, a center prism 22c, and side prisms 22d.
The support member 22a supports the camera 22b, the center prism 22c, and the side prisms 22d. The support member 22a includes a plate-like member. The support member 22a is formed in an appropriate shape so as to be able to support the camera 22b, the center prism 22c, and the side prisms 22d at predetermined positions.
The camera 22b is a device for acquiring an image. The camera 22b is fixed to the support member 22a with a lens 22e oriented in a predetermined direction.
The center prism 22c guides light from two directions to the lens 22e of the camera 22b. The center prism 22c is disposed in front of the lens 22e (in a direction toward which the lens 22e is directed). The center prism 22c reflects light from sides (an up and down direction in Fig. 7A) and guides the light to the lens 22e.
The side prisms 22d guide light from predetermined directions to the center prism 22c. A pair of the side prisms 22d (a pair in the up and down direction in Fig. 7A) are provided with the center prism 22c interposed therebetween. The side prisms 22d reflect light from a front (the right in Fig. 7A) and guide the light to the center prism 22c.
In the image capturers 22 configured as described above, light from the front of the pair of side prisms 22d is appropriately guided by the side prisms 22d and the center prism 22c and enters the lens 22e of the camera 22b. As a result, the image capturers 22 acquire images of two different locations (front surfaces of the two side prisms 22d) by one camera 22b.
Next, arrangements of the image capturers 22 will be described with reference to Figs. 2 and 6.
As illustrated in Figs. 2 and 6, the image capturers 22 are disposed above the needle bed gap S of the needle beds 10. The front and rear carriages 20 are provided with a total of four image capturers 22. Specifically, two image capturers 22 are disposed on each of the front and rear with the needle bed gap S interposed therebetween. The image capturers 22 are fixed to the carriages 20 directly or indirectly (through the bridge 20a or another appropriate member). The image capturers 22 provided on the carriages 20 which are movable allows imaging to be performed with the image capturers 22 moving together with the carriages 20, and thus, imaging can be performed at a plurality of locations by a single image capturer 22, for example.
As for the two image capturers 22 disposed at the front of the needle bed gap S, the image capturers 22 are disposed above the front carriage 20F. The image capturers 22 are disposed such that an imaging direction (lens 22e) faces rear-downward. Thus, the image capturers 22 can image distal ends (specifically, stitches held on the knitting needles 11) of the knitting needles 11 of the rear needle bed 10B.
Thus, the shape of stitches can be easily and clearly imaged by imaging the knitting needles 11 of the rear needle bed 10B from above the needle bed gap S. In particular, as in the embodiment, by imaging the knitting needles 11 of the rear needle bed 10B from the opposing front needle bed 10F, the shape of the stitches can be easily imaged more clearly, and eventually, defective knitting described later can be accurately detected.
Further, the two image capturers 22 disposed in front of the needle bed gap S are disposed so as to correspond to positions in the left and right of the cam mechanism 21. Specifically, as illustrated in Fig. 6, the image capturers 22 are disposed at substantially the same position in the left and right direction as the first transfer cam mechanism 21A and the second transfer cam mechanism 21C. In addition, the side prisms 22d of the image capturers 22 are disposed to be located to the left and right of the cam mechanism 21, respectively.
As a result, as illustrated in Fig. 6, the two image capturers 22 can image four imaging points (rear first imaging point PB1, rear second imaging point PB2, rear third imaging point PB3, and rear fourth imaging point PB4) of the rear needle bed 10B.
The rear first imaging point PB1 is set to be located to the left of the first transfer cam mechanism 21A. The rear second imaging point PB2 is set so as to be located to the right of the first transfer cam mechanism 21A and to the left of the knit cam mechanism 21B. Further, the rear third imaging point PB3 is set to be located to the right of the knit cam mechanism 21B and to the left of the second transfer cam mechanism 21C. The rear fourth imaging point PB4 is set to be located to the right of the second transfer cam mechanism 21C.
Further, the two image capturers 22 disposed behind the needle bed gap S are disposed symmetrically in the front and rear direction with the image capturers 22 disposed at the front across the needle bed gap S, and thus a detailed description of the two image capturers 22 will be omitted. As illustrated in Fig. 6, the two image capturers 22 disposed behind the needle bed gap S can image four imaging points (front first imaging point PF1, front second imaging point PF2, front third imaging point PF3, and front fourth imaging point PF4) of the front needle bed 10F.
Left and right positions of the four front imaging points PF1 to PF4 are also similar to those of the four rear imaging points PB1 to PB4, and thus detailed description will be omitted.
As described above, in the embodiment, a total of eight imaging points can be imaged using the four image capturers 22.
Next, a configuration of the illuminators 23 will be described with reference to Fig. 8.
As illustrated in Fig. 8, each of the illuminators 23 mainly includes a light emitting diode (LED) 23a, condenser lenses 23b, and a guide 23c.
The LED 23a is a light source that emits light when a voltage is applied. The LED 23a can emit light in a predetermined direction (to the right in Fig. 8).
The condenser lenses 23b are configured to collect light emitted from the LED 23a. The condenser lenses 23b are disposed in front of the LED 23a (to the right in Fig. 8). In the embodiment, two condenser lenses 23b are disposed.
The guide 23c guides the light emitted from the LED 23a to a predetermined position. The guide 23c is formed with a material (acrylic or the like) having translucency. The guide 23c is formed in a longitudinal shape extending along an irradiation direction of light from the LED 23a (the left and right direction in Fig. 8). A distal end surface 23d (end surface opposite to the LED 23a) of the guide 23c is formed in an inclined surface cut obliquely. Specifically, the distal end surface 23d is formed so as to be inclined with respect to a direction perpendicular to the longitudinal direction of the guide 23c.
In the illuminator 23 configured as described above, when light is emitted from the LED 23a, the light enters the guide 23c through the condenser lenses 23b. The light entering the guide 23c travels along the longitudinal direction of the guide 23c, and is emitted from the distal end surface 23d to outside of the guide 23c. When the light is emitted from the distal end surface 23d, the light is refracted at an angle corresponding to an inclination angle of the distal end surface 23d.
Next, arrangements of the illuminators 23 will be described with reference to Figs. 2 and 6.
As illustrated in Figs. 2 and 6, the carriages 20 are provided with a total of eight illuminators 23. Specifically, four illuminators 23 are disposed at each of the front and rear with the needle bed gap S interposed therebetween. The illuminators 23 are fixed to the carriages 20 directly or indirectly (through the bridge 20a or another appropriate member).
As for the four illuminators 23 disposed in front of the needle bed gap S, the illuminators 23 are disposed such that the irradiation direction of the light (the distal end surface 23d of the guide 23c) faces rear-downward. Thus, the illuminators 23 can irradiate the distal ends (specifically, the stitches held on the knitting needles 11) of the knitting needles 11 of the rear needle bed 10B with light.
In addition, the four illuminators 23 disposed in front of the needle bed gap S are disposed so as to be able to irradiate the four imaging points PB1 to PB4 of the rear needle bed 10B with light.
Further, the four illuminators 23 disposed behind the needle bed gap S are disposed symmetrically in the front and rear direction with the illuminators 23 disposed at the front across the needle bed gap S, and thus a detailed description of the four illuminators 23 will be omitted. The four illuminators 23 disposed behind the needle bed gap S are disposed so as to be able to irradiate the four imaging points PF1 to PF4 of the front needle bed 10F with light as illustrated in Fig. 6.
Hereinafter, a relative positional relationship between the image capturers 22 and the illuminators 23 will be described with reference to Fig. 9.
Fig. 9 illustrates, as an example, the image capturer 22 that images the front first imaging point PF1 and the front second imaging point PF2, and the two illuminators 23 that irradiate the front first imaging point PF1 and the front second imaging point PF2 with light.
As illustrated in Fig. 9, the illuminators 23 are disposed to sides of a visual field of the image capturer 22 and irradiate the imaging points with light. Here, as described above, the distal end surfaces 23d of the illuminators 23 are formed so as to refract the light emitted from the guides 23c. In the embodiment, the illuminators 23 are disposed such that light is refracted toward the imaging points. As a result, the illuminators 23 can be disposed outside the visual field of the image capturer 22 (out of the visual field), and the illuminators 23 can be prevented from avoiding the imaging of the image capturer 22.
In the flat knitting machine 1 configured as described above, the controller 60 can detect defective knitting by performing a quality determination of the stitches on the basis of the image of the stitches imaged by the image capturers 22. Hereinafter, a method of detecting defective knitting by the controller 60 will be described.
The controller 60 carries out the imaging by the image capturers 22 in synchronization with the reciprocating movement of the carriages 20 during the knitting of the knitted fabric K. Specifically, when the knitting needles 11 to be imaged reach the imaging points by the movement of the carriage 20, the controller 60 causes the illuminators 23 to irradiate the stitches held on the knitting needles 11 with light and causes the image capturers 22 to image the stitches. The controller 60 can recognize the position of the carriages 20 from the rotational speed of the servomotor 40, and thus the imaging can be performed at an appropriate timing in accordance with the position of the stitches (knitting needles 11). The controller 60 images all the knitting needles 11 (stitches) during the knitting. At this time, the controller 60 stores the imaged position of the stitches.
Here, for example, as illustrated in Fig. 10A, in a case where the carriages 20 move to the right, stitches are formed by the knit cam mechanism 21B as a preceding system, and the stitches are transferred by the first transfer cam mechanism 21A as a succeeding system. That is, in this case, a defect associated with the knitting may occur to the knitting needles 11 after passing through the knit cam mechanism 21B and the knitting needles 11 after passing through the first transfer cam mechanism 21A.
Then, when the carriages 20 move to the right, the controller 60 images only four imaging points including the front first imaging point PF1, the front second imaging point PF2, the rear first imaging point PB1, and the rear second imaging point PB2. That is, when the carriages 20 move to the right, the controller 60 images the stitches using only the two image capturers 22 disposed on the left.
On the contrary, as illustrated in Fig. 10B, in a case where the carriages 20 move to the left, stitches are formed by the knit cam mechanism 21B as a preceding system, and the stitches are transferred by the second transfer cam mechanism 21C as a succeeding system. That is, in this case, a defect associated with the knitting may occur to the knitting needles 11 after passing through the knit cam mechanism 21B and the knitting needles 11 after passing through the second transfer cam mechanism 21C.
Then, when the carriages 20 move to the left, the controller 60 images only four imaging points including the front third imaging point PF3, the front fourth imaging point PF4, the rear third imaging point PB3, and the rear fourth imaging point PB4. That is, when the carriages 20 move to the left, the controller 60 images the stitches using only the two image capturers 22 disposed on the right.
However, such control of the image capturers 22 is an example, and all the imaging points can be imaged using all the four image capturers 22, for example. In this case, it is possible to alleviate a processing load of the controller 60 not by performing the quality determination of all the imaged images but by performing the quality determination of only necessary images appropriately.
In this way, the controller 60 images all the stitches of each course and stores the position of the imaged stitches while the knitted fabric K is being knitted, as shown in Fig. 11.
Further, the controller 60 performs the quality determination of the stitches of a previous course during the knitting of each course. Hereinafter, a method of performing the quality determination of a stitch will be described.
Fig. 12 illustrates an example of an image P of the stitches imaged by the image capturers 22. Fig. 12 illustrates an example of the image P obtained by imaging the front first imaging point PF1 and the front second imaging point PF2.
First, the controller 60 extracts a region (determination region R) where a stitch to be determined is shown from each of the front first imaging point PF1 and the front second imaging point PF2. At this time, the controller 60 can extract an appropriate determination region R on the basis of information of the image P. For example, a position of the distal end of each knitting needle 11 can be detected on the basis of luminance of the image P, and the determination region R can be extracted on the basis of the position of the distal end of each knitting needle 11.
Next, the controller 60 compares the shape of the stitch shown in the determination region R (actually imaged) with a shape of a stitch at a normal state stored in advance, and performs the quality determination of the stitch.
For example, the controller 60 has stored an image (normal image N) showing the shape of the stitch at the normal state in advance. The shape of the stitch at the normal state differs in accordance with a type of stitch (knit, miss, tuck, or the like), a type of knitting yarn Y, the adjacent stitches (for example, the type and presence or absence of the stitches that have been formed until the previous course and the stitches on the right and left), and the like. Therefore, the controller 60 has stored a large number of normal images N of various patterns in advance.
The controller 60 determines the shape of stitch that may be held on the knitting needle 11 of the imaged image in consideration of the type of the stitch, the type of the knitting yarn Y, the adjacent stitch, and the like described above on the basis of the knitting data of the knitted fabric K that is being currently knitted and the position of the imaged stitch. The controller 60 extracts a normal image N suitable for the quality determination (corresponding to the shape of the imaged stitch) from the stored various normal images N and uses the normal image N for the quality determination of the stitch.
The controller 60 performs the quality determination of the stitch by performing image processing using the image of the determination region R and the normal image N. For example, the controller 60 calculates a similarity between the shape of the stitch in the determination region R and the shape of the stitch shown in the normal image N, and when the similarity is high to some extent (greater than or equal to a predetermined threshold value), the controller can determine that the stitch in the determination region R is normal.
On the other hand, the controller 60 can determine that the stitch in the determination region R is abnormal when the similarity is low to some extent (less than the predetermined threshold value). Various states such as a state in which the knitting yarn Y is broken, a state in which a stitch (loop) is not formed, and a state in which the stitch is excessively clogged are assumed as the abnormality of the stitch.
The quality determination of the stitches (comparison of the shapes of the stitches) can be performed by comparing RGB values, luminance, and the like of the determination region R and the normal image N. The quality determination of the stitches can also be performed, for example, by extracting a core wire of the knitting yarn Y (line passing through a center of the knitting yarn Y) forming a stitch from an image and comparing (calculating similarity) a shape of the core wire and a shape of the core wire at a normal state without using the luminance or the like.
The controller 60 quickly performs the quality determination of the stitches as described above after imaging the image P. Specifically, as shown in Fig. 11, the controller 60 performs the quality determination of the image imaged in a certain course (for example, N course) during knitting of the next course (N + 1 course). Thus, if a defect occurs in the stitch, the defect can be detected immediately after (during the knitting of the next course).
For example, as shown in Fig. 11, if a defect occurs in the stitches during the knitting of N + 3 course, the defect of the stitches of the N + 3 course can be detected during knitting of N + 4 course. The controller 60 can quickly stop the operation of the carriages 20 and interrupt the knitting by the flat knitting machine 1 upon detection of the defect of the stitch. Timing of the quality determination of the stitches is not limited to this timing, and for example, the quality determination of the image imaged in a certain course can be performed in real time during the knitting of the course.
The controller 60 stores a position where the defect is detected (position of the knitting needles 11). The position can be recognized on the basis of the position of the carriages 20 when the image P is acquired. By storing the position, not only presence or absence of the occurrence of a defect but also an occurrence location of the defect can be recognized.
The controller 60 can also show the image of the stitch at the position where the defect is detected (or where a defect is suspected) on a display such as a monitor to allow an operator to confirm.
As described above, in the embodiment, the stitch held on the knitting needles 11 can be imaged, and a defect can be detected from the imaged image. That is, the presence or absence of a defect can be found during the knitting of the knitted fabric K, and thus defective knitting can be detected at an early stage. Thus, measures such as interruption of the knitting can be promptly taken, and unnecessary costs (power of flat knitting machine 1 for knitting the knitted fabric K, knitting yarn Y, knitting time, and the like) can be suppressed.
Further, the shape of each stitch can be more accurately checked, and eventually the defect can be accurately detected by performing the quality determination of the shape of each stitch held on the knitting needles 11 instead of performing the quality determination on the knitted fabric K after the knitting. In particular, the defect of the knitted fabric (rib knitted fabric, knitted fabric of stretch yarn, and the like) of which the defect is difficult to confirm in the state of the knitted fabric can also be accurately detected.
Note that the controller 60 according to the embodiment is one embodiment of a determiner and a position storage according to the disclosure.
Although the embodiment of the disclosure has been described above, the disclosure is not limited to the embodiment, and appropriate modifications can be made within the technical ideas of the disclosure described in the claims.
For example, in the embodiment, the flat knitting machine 1 has been described as an example of the knitting machine, but the disclosure is not limited to the flat knitting machine 1, and can be applied to other various knitting machines (for example, a circular knitting machine, a warp knitting machine, and the like). That is, the defective knitting can be detected by providing the image capturer capable of imaging the stitches held on the knitting needles of various knitting machines.
The image imaged by the image capturers 22 can also be used for purposes other than detection of defective knitting. For example, the knitted fabric K that is generally knitted is pulled down by a wind-down roller or the like, but can also be used for controlling the wind-down roller (controlling of a pull-down tension of the knitted fabric K).
Specifically, when the pull-down tension of the wind-down roller is weak, the knitted fabric K rises at the needle bed gap S. Then, presence or absence of rising of the knitted fabric K is determined from the image imaged by the image capturers 22, and the pull-down tension of the winding-down roller is increased if the knitted fabric K has risen. This can suppress the occurrence of defects. The flat knitting machine 1 can also be stopped upon determination that a defect obviously has occurred due to the rising of the knitted fabric K. The flat knitting machine 1 can also be stopped upon determination that a defect obviously has occurred by rising of the fabric.
Further, in the embodiment, an example in which the image capturers 22 are provided in the carriages 20 has been described, but the disclosure is not limited to this example, and the arrangements of the image capturers 22 are not limited. For example, the image capturers 22 can be provided on a moving body (moving body movable along the longitudinal direction of the needle beds 10) provided separately from the carriages 20. For example, the image capturers 22 may be provided on a moving body movable along the yarn guide rail 30.
Further, in the embodiment, the image capturers 22 are configured to be movable, but the disclosure is not limited to this configuration, and the image capturers 22 can be dispose (fixed) to be unmovable. For example, the image capturers 22 can be fixed at a position where the stitches can be imaged, such as above the needle bed gap S. In this case, an entire range of the needle beds 10 in the longitudinal direction may be imaged by one image capturer 22 or may be imaged at a plurality of places by using a plurality of image capturers 22.
In a case where the image capturers 22 are fixed, the stitches are desirably imaged with the image capturer 22 in association with timing of formation of the stitches with the knitting needle 11. Specifically, as described in the embodiment, the knitting needles 11 after passing through the preceding system and the succeeding system are desirably imaged by interlocking the timing of imaging with the position of the carriages 20.
Fixing the image capturers 22 in this way eliminates the need for a mechanism for movement and thus simplifies the configuration. The image of the stitch can be acquired at an appropriate timing (after the knitting is completed) by imaging the stitch in association with the timing of formation of the stitch by the knitting needles 11.
In the embodiment, the image capturers 22 are disposed above the carriages 20, and the stitches are imaged from above the needle beds 10, but the disclosure is not limited to this configuration. For example, as illustrated in Fig. 13A, the image capturers 22 can be disposed below the needle beds 10, and the stitches may be imaged from below the needle beds 10.
As illustrated in Fig. 13B, the image capturers 22 can be disposed above and below the needle beds 10, and the stitches may be imaged both from above and from below the needle beds 10. In this configuration, the shape of a stitch can be more accurately recognized and a detection accuracy of a defect can be improved by imaging the same stitch from a plurality of directions.
Similarly to the image capturers 22, the illuminators 23 can also be provided on a moving body other than the carriages 20. Further, the illuminators 23 can be also disposed (fixed) so as to be unmovable. The illuminators 23 can be disposed below (or both above and below) the needle beds 10, and light can be irradiated from below the needle beds 10 toward the stitches. Further, the illuminators 23 can be provided outside the carriages 20, and the carriages 20 can be provided with an appropriate reflecting member (mirror or the like) to guide the light emitted from outside of the carriages 20 to a desired imaging location.
Further, a type (color and intensity) of light irradiated by the illuminators 23 can be appropriately adjusted in accordance with the type and color of the knitting yarn Y.
As the normal image N for performing the quality determination of the stitches, not only an image obtained by actually imaging the normal stitches but also an image created by, for example, machine-learning a large number of images of normal stitches can be used. This makes it possible to perform more appropriate quality determination. In this case, the image of the stitches imaged by the image capturers 22 during the knitting by the flat knitting machine 1 can be further learned, and the normal image N can be updated as needed.
Further, an image created by combining a stitch shape and a yarn image obtained by a 3D simulation can be used as the normal image N.
In the embodiment, an example of performing the quality determination on the basis of the shape of the stitch has been described, but the disclosure is not limited to this example. For example, the controller 60 can determine the type of the imaged stitch, compare the type of the imaged stitch with the type information obtained from the knitting data, and perform the quality determination of the stitch. The controller 60 can also perform the quality determination of the stitch by determining the similarity between the imaged image of the stitch and the image of the normal stitch. Thus, the controller 60 can perform the quality determination of the stitch by appropriately comparing the imaged stitch with every state (shape, type, and the like) of the stitch in the normal state.
In the embodiment, the example in which the controller 60 stores the position where a defect is detected has been described, but the disclosure is not limited to this example. For example, the position where the defect is detected can be stored in an RFID (radio frequency identifier) (ID tag) provided in the knitted fabric K, a server that manages information related to the knitted fabric K, or the like.
In the embodiment, an example in which the flat knitting machine 1 is promptly stopped upon detection of a defect has been described, but the disclosure is not limited to this example. For example, the knitting can be continued without stopping the flat knitting machine 1 upon detection of a defect, and a defective location can be confirmed after the knitting of the knitted fabric K is completed. Further, if there is a stitch for which it is difficult to determine the quality that cannot be completely determined to be defective, the position of the stitch can be stored in the controller 60 or the like, and then the operator can visually confirm after the knitting of the knitted fabric K is completed.
In addition to the quality determination of the stitches as in the embodiment, the quality determination of the knitted fabric K that has been knitted can also be performed. For example, another image capturer different from the image capturer 22 according to the embodiment is provided below the needle bed 10 to image the knitted fabric K. The operator can check the image of the knitted fabric K of a location suspected of being defective by the quality determination of the stitches on the display, and the controller 60 can automatically determine the quality of the knitted fabric K. Thus, by combining the quality determination of the stitches and the quality determination of the knitted fabric K, a defect that cannot be determined only by the quality determination of the stitches can be detected from the image of the knitted fabric K, and a defect (oil streak, dirt, and the like) that can be detected only in the image of the knitted fabric K can be detected.
Further, by acquiring the image of the stitch as in the embodiment, not only the defect can be detected, but also a cause of the defect can be identified. Specifically, upon detection of a defect of the stitch, the cause of the defect can be specified by confirming the image. For example, when the knitting yarn Y is broken, a tension of the knitting yarn Y is too high, and when a stitch is not formed, a cause of a breakage of the knitting needle 11 or the like can be identified (estimated).
In the embodiment, an example in which a pair (two) of front and rear needle beds 10 are provided has been described, but the number of needle beds 10 is not limited to this example. For example, it is possible to provide two needle beds 10 on the upper and lower sides at each of the front and rear of the needle bed gap S (four in total).
Further, the cam mechanism 21 illustrated in the embodiment is an example, and the configuration (number, arrangement, operation, and the like) of the cam mechanism 21 can be arbitrarily changed. The disclosure can also be applied to a carriage-less knitting machine that does not include the cam mechanism 21 and drives knitting needles to advance and retract with an individual actuator.
In the embodiment, the configuration in which the position of the carriages 20 are detected on the basis of the rotational speed of the servomotor 40 has been exemplified, but the disclosure is not limited to this configuration. An appropriate sensor for detecting the position of the carriages 20 (furthermore, the position of stitch (knitting needle 11) to be imaged) can be separately provided.
Further, in the embodiment, an example in which various types of control and defect detection are performed by the controller 60 provided in the flat knitting machine 1 has been described, but the disclosure is not limited to this example. That is, some or all of functions of the controller 60 can be performed by a controller (for example, a personal computer or the like) provided separately from the flat knitting machine 1. For example, the quality determination of the stitches can be performed by a PC installed outside the flat knitting machine 1.
In the embodiment, an example in which the flat knitting machine 1 (the controller 60 provided in the flat knitting machine 1) has a function of detecting a defect has been described, but the disclosure is not limited to this example. That is, the disclosure can be applied not only to the knitting machine but also to a defect detection system that detects a defect of a knitted fabric knitted by the knitting machine. For example, in this case, it is sufficient that the defect detection system includes an image capturer capable of imaging a stitch held on a knitting needle, a determiner that performs a quality determination from a shape of the stitch imaged by the image capturer, and a position storage that stores a position of the imaged stitch. Such a defect detection system can be easily applied to an existing knitting machine by appropriately installing the image capturer in the existing knitting machine.

Claims

A knitting machine (1) comprising:
a knitting needle (11) configured to knit a knitted fabric (K);

an image capturer (22) configured to image a stitch held on the knitting needle (11);

a determiner (60) configured to compare a state of the stitch imaged by the image capturer (22) with a state of the stitch in a normal state and perform a quality determination of the stitch; and

a position storage (60) configured to store a position of the stitch imaged by the image capturer (22).
The knitting machine (1) as claimed in claim 1, wherein
the knitting needle (11) is disposed at each of needle beds (10) disposed so as to face each other across a needle bed gap (S), and
the image capturer (22) images a stitch held on the knitting needle (11) from above the needle bed gap (S) of the needle beds (10) disposed to face each other.
The knitting machine (1) as claimed in claim 2, wherein the image capturer (22) images, from a first side of the needle beds (10) disposed to face each other, a stitch held on the knitting needle (11) disposed at a second side of the needle beds (10).
The knitting machine (1) as claimed in claim 2 or 3, wherein the image capturer (22) is provided so as to be movable along the needle beds (10).
The knitting machine (1) as claimed in any one of claims 1 to 3, wherein the image capturer (22) is fixed at such a position as to image a stitch held on the knitting needle (11) and image the stitch in association with a timing of formation of the stitch with the knitting needle (11).
A defect detection system comprising:
an image capturer (22) configured to image a stitch held on a knitting needle (11) configured to knit a knitted fabric (K);

a determiner (60) configured to compare a state of the stitch imaged by the image capturer (22) with a state of the stitch in a normal state and perform a quality determination of the stitch; and

a position storage (60) configured to store a position of the stitch imaged by the image capturer (22).