CN114067453A

CN114067453A - Weaving defect detection method and defect inspection method

Info

Publication number: CN114067453A
Application number: CN202111279155.XA
Authority: CN
Inventors: 陆伟
Original assignee: Suzhou Jingsuo Intelligent Technology Co ltd
Current assignee: Suzhou Jingsuo Intelligent Technology Co ltd
Priority date: 2021-10-31
Filing date: 2021-10-31
Publication date: 2022-02-18

Abstract

The invention discloses a weaving defect detection method and a defect inspection method, which are used for detecting cloth woven by a weaving machine, and the detection method comprises the following steps: the method comprises the following steps of utilizing an image acquisition device to acquire an image of a weaving machine which is weaving, wherein the acquired image comprises a cloth area and a warp yarn area; and sending the acquired image to an AI recognition device for defect recognition, wherein if the image characteristic difference between the acquired image and the adjacent pixel in the cloth area exceeds a preset threshold value and the warp yarn area has a warp broken line, the AI recognition device outputs a recognition result that the cloth is defective, otherwise, the AI recognition device outputs a recognition result that the cloth is not defective. The defect feature identification standard provided by the invention can eliminate false defect misjudgment caused by dyeing threads and improve the accuracy of the detection result of the weaving defects.

Description

Weaving defect detection method and defect inspection method

Technical Field

The invention relates to the field of woven fabric detection, in particular to a woven fabric defect detection method and a defect inspection method.

Background

In the prior art, a movable vision system is proposed to detect defects in the fabric of multiple textile machines, such as in chinese patent application publication CN 113322653A.

However, practice shows that the weaving machine is stopped when the vision system judges that weaving defects exist, the weaving defects cannot be detected during manual inspection, the erroneous judgment of the vision system can seriously affect the production capacity, and a large amount of manual inspection is consumed.

Disclosure of Invention

The invention aims to provide a woven fabric defect detection method and a defect inspection method, which improve the accuracy of detection results.

In order to achieve the purpose, the technical scheme adopted by the invention is as follows:

a weaving defect detection method is used for detecting cloth woven by a weaving machine and comprises the following steps:

the method comprises the following steps of utilizing an image acquisition device to acquire an image of a weaving machine which is weaving, wherein the acquired image comprises a cloth area and a warp yarn area;

and sending the acquired image to an AI recognition device for defect recognition, wherein if the image characteristic difference between the acquired image and the adjacent pixel in the cloth area exceeds a preset threshold value and the warp yarn area has a warp broken line, the AI recognition device outputs a recognition result that the cloth is defective, otherwise, the AI recognition device outputs a recognition result that the cloth is not defective.

Further, the AI identification device is a pre-established cloth defect detection neural network model, which is pre-established by the following steps:

establishing a sample library: collecting a plurality of images of the woven fabric, wherein the images comprise adjacent fabric areas and warp yarn areas, and manually marking each image to be defective or non-defective, wherein if the difference of the image characteristics of the adjacent pixels in the fabric areas exceeds a preset threshold value and the warp yarn areas are broken, the image is marked to be defective, otherwise, the image is marked to be non-defective;

and performing machine learning and model verification by using the images in the sample library to obtain a cloth defect detection neural network model.

Further, the image acquisition device is configured to move along a track, and during the process that the image acquisition device shoots the same piece of cloth twice before and after, the image acquisition device moves to the position above other textile machines so as to acquire images of the cloth woven at the textile machines.

Furthermore, a plurality of position sensors corresponding to the plurality of looms are arranged on the track, and the main control module is used for analyzing the loom corresponding to the image acquisition device at present according to the position detection results of the plurality of position sensors;

and if the current output identification result of the AI identification device is that the cloth is defective, the main control module sends a stop instruction to the loom corresponding to the image acquisition device currently.

On the other hand, the invention also provides a weaving defect inspection method, which is based on weaving defect inspection equipment and performs visual inspection on the cloth at different looms in a certain sequence to identify whether the cloth has defects or not, and the weaving defect inspection method comprises the following steps:

the image acquisition device moves on an inspection circular track provided by the weaving defect inspection equipment so as to sequentially carry out wheel type image acquisition on the cloth weaved at different looms and send the acquired images to the AI identification device;

and if the AI identification device identifies that the cloth has defects, controlling the loom corresponding to the image acquisition device to stop working.

Further, the image acquired by the image acquisition device comprises a cloth area and a warp yarn area, if the image characteristic difference between the cloth area and the adjacent pixel exceeds a preset threshold value and the warp yarn area has a warp yarn broken line, the AI identification device outputs an identification result that the cloth is defective, otherwise, the AI identification device outputs an identification result that the cloth is defect-free.

Furthermore, the running speed of the image acquisition device at each position on the routing inspection circular track is configured, and the main control module is used for calculating the loom corresponding to the image acquisition device currently in real time according to the initial position and the moving starting time of the image acquisition device.

Furthermore, the circular inspection track comprises two linear tracks and two arc tracks, wherein the two linear tracks are arranged side by side, the arc tracks are respectively connected with the end parts of the two linear tracks to form a closed circular track, and the circular track is horizontally erected on the plurality of stand columns;

the annular track comprises an inner side sliding rail, an outer side sliding rail and a sliding table, wherein the inner side sliding rail and the outer side sliding rail are arranged in parallel, and the sliding table is configured to cover or partially cover the inner side sliding rail and the outer side sliding rail, so that the inner side sliding rail and the outer side sliding rail jointly limit the sliding table in the up-down direction;

steel balls are embedded on the inner side surface of the sliding table and can roll relative to the sliding table, and the sliding table is in point contact with the inner side sliding rail and the outer side sliding rail through the steel balls;

the side wall of the annular track is provided with a rack, the image acquisition device is arranged on the sliding table, the sliding table is also provided with a first driving mechanism, the first driving mechanism comprises a motor and a gear, and the gear is connected with the output shaft of the motor and is meshed with the rack; under the drive of the first driving mechanism, the sliding table can drive the image acquisition device to pass through the arc-shaped track on the annular track without speed reduction.

Furthermore, the weaving defect inspection equipment comprises two image acquisition devices and sliding tables corresponding to the two image acquisition devices one by one, wherein the two image acquisition devices comprise a first image acquisition device and a second image acquisition device;

the first image acquisition device and the second image acquisition device are configured to be separated by a distance equal to the sum of the lengths of a straight track and an arc track;

the image acquisition device is a non-linear scanning camera, and in the process from the shooting of one loom by the first image acquisition device to the shooting of the same loom by the second image acquisition device, the newly added weaving length of the loom is smaller than the width of the shooting area of the first image acquisition device or the second image acquisition device.

The technical scheme provided by the invention has the following beneficial effects:

the identification rule is determined to meet the condition that the difference between the image characteristics of adjacent pixels and the image characteristics of a cloth area exceeds a preset threshold value, and the condition that the warp yarn area has warp yarn breakage is met, so that the cloth defect is determined, the cloth defect detection accuracy is greatly improved, the misjudgment rate is reduced, and the productivity is improved under the condition of ensuring the cloth quality.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments described in the present application, and other drawings can be obtained by those skilled in the art without creative efforts.

Fig. 1 is a schematic structural diagram of a woven cloth defect inspection device according to an exemplary embodiment of the present disclosure;

FIG. 2 is an enlarged view of the middle structure of the circular track in FIG. 1;

FIG. 3 is an enlarged view of the end structure of the endless track of FIG. 1;

fig. 4 is a schematic structural diagram of a trolley power supply device according to an exemplary embodiment of the present disclosure;

FIG. 5 is a schematic cross-sectional view of a gantry rail provided in an exemplary embodiment of the present disclosure;

fig. 6 is a schematic perspective view of a gantry rail according to an exemplary embodiment of the present disclosure;

FIG. 7 is a schematic structural view of a rack on the inner side wall of a track provided by an exemplary embodiment of the present disclosure;

fig. 8 is a schematic flow chart of a fabric defect detection method according to an exemplary embodiment of the present disclosure;

FIG. 9 is a schematic diagram of different fabric defect detection results provided by an exemplary embodiment of the present disclosure;

wherein the reference numerals include: 11-upright post, 12-cross beam, 21-linear track, 22-arc track, 3-first visual detection device, 31-camera component, 32-control box, 4-second visual detection device, 51-conductive strip, 52-second driving mechanism, 53-crawler track, 54-connecting part, 55-carbon brush, 56-extension conductive wire, 61-inner slide rail, 62-outer slide rail, 63-slide rail, 64-steel ball, 65-inner side wall rack of arc track, and 66-inner side wall rack of linear track.

Detailed Description

In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

It should be noted that the terms "first," "second," and the like in the description and claims of the present invention and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, apparatus, article, or device that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or device.

In an embodiment of the present invention, a method for detecting a woven fabric defect is provided, and referring to fig. 8, the method for detecting a woven fabric defect includes the following steps:

the method comprises the following steps of utilizing an image acquisition device to acquire an image of a weaving machine which is weaving, wherein the acquired image comprises a cloth area and a warp yarn area, referring to fig. 9, the upper part is the warp yarn area, and the lower part is the cloth area;

and sending the acquired image to an AI recognition device for defect recognition, wherein if the image characteristic difference between the acquired image and the adjacent pixel in the cloth area exceeds a preset threshold value and the warp yarn area has a warp broken line, the AI recognition device outputs a recognition result that the cloth is defective, otherwise, the AI recognition device outputs a recognition result that the cloth is not defective. As shown in fig. 9, it is obvious that the space encircled by the left ellipse has a characteristic that the interval between two warps is relatively large, that is, the interval between the warps on the image is large due to the breakage of one corresponding warp, and it can be determined that the "wire drawing" in the cloth area is caused by the breakage of one warp, and if the loom is not stopped in time, the cloth is a defective product, which causes economic loss; in the area circled by the right ellipse in fig. 9, although there is a vertical line of "stringy" in the cloth area, the corresponding warp above it is not broken (the warp is uniformly distributed), so it should not be considered as a defect, in fact, it is caused by the dyeing line, and belongs to a non-defective product.

Specifically, the AI identification device is a pre-established cloth defect detection neural network model, which is pre-established by the following steps:

And inputting an image shot by a first visual detection device (namely an image acquisition device) into the cloth defect detection neural network model, and outputting an identification result by the cloth defect detection neural network model.

In this embodiment, the marking judgment standard in the step of pre-establishing the sample library is unique, and before the judgment standard is not found, practically, the false alarm caused by the wrong identification of the AI identification device is always found, that is, the textile machine is controlled to stop, the defect of the woven fabric is not found manually, and the reason of the false judgment of the AI identification device is found to be the dyeing line on the spool after long-time examination. The identification of different bobbins by dyeing threads of different colors is a traditional good in the textile industry, so that the bobbins in a textile workshop are as many as hundreds or even nearly ten thousand, therefore, the traditional identification method cannot be banned up until now, but the color generally used as the dyeing thread is different from the color of the bobbin, so that the dyeing thread is easy to cause the 'drawn thread' characteristic of suspected flaws visually when being woven into cloth, but actually the characteristic does not belong to the weaving defect characteristic.

The present application proposes that, when the suspected defect occurs, it is also necessary to determine whether the warp thread is broken, that is, if the adjacent distances of all warp threads are substantially equidistant, the warp thread is considered not to be broken, otherwise, if the distance between two warp threads is significantly greater than the distance between two other warp threads (e.g., in the oval area on the left side of fig. 9), the warp thread is considered to be broken. Only if the suspected defects (image features such as gray values and color differences between adjacent pixels) and the occurrence of warp thread breakage are simultaneously satisfied are the weaving defects considered to be true.

Different batches of woven fabrics have different colors and patterns, so that different artificially marked image samples need to be added in the sample library according to the increase of the batches of the woven fabrics.

In an embodiment of the present invention, there is provided a cloth defect inspection apparatus for circularly performing visual inspection on cloth at different looms in a certain order to identify whether there is a defect in the cloth, referring to fig. 1, the cloth defect inspection apparatus comprising:

the circular rail is horizontally erected on the plurality of columns 11 and comprises two linear rails 21 and two arc rails 22, wherein the two linear rails 21 are arranged side by side, and the arc rails 22 are respectively connected with the end parts of the two linear rails 21 to form a closed circular rail;

a first visual detection device 3 provided on the endless track and configured to move along the endless track by a first drive mechanism; referring to fig. 2, the first visual inspection device 3 includes a camera component 31 and a control box 32, wherein the camera component 31 is used for taking a downward image and sending the acquired image to an AI recognition device in the control box 32, and the AI recognition device is used for recognizing whether a woven fabric in the image has a defect or not, which is described in detail below.

In the present application, not only the first visual inspection device 3 is driven to move on the circular track without interruption, but also the moving first visual inspection device itself needs to be powered to work, so that powering an electric load moving on the circular track by an external power source is a difficult problem to be realized, and the present embodiment provides the following solutions for solving the problem,

the weaving defect inspection equipment of the embodiment further comprises a sliding contact power supply device, and referring to fig. 2 and 4, the sliding contact power supply device comprises a second driving mechanism 52, a sliding contact power supply, an elongated conductive bar 51 (such as copper foil), an elongated crawler 53, a connecting part 54 arranged on the crawler 53 and a carbon brush 55; wherein the caterpillar 53 is arranged side by side with the linear track 21, as shown in fig. 2, the caterpillar 53 can be carried by the beam 12, the elongated conductive strip 51 is electrically connected with the sliding contact power supply, the caterpillar 53 is configured to drive the connecting portion 54 and the carbon brush 55 to reciprocate under the forward and reverse alternate driving of the second driving mechanism 52, the carbon brush 55 is always in contact with the conductive strip 51, and the carbon brush 55 is electrically connected with the first visual inspection device 3 through the extended conductive wire 56 on the connecting portion 54;

a main control module electrically connected to the second driving mechanism 52, wherein the main control module is configured to control the second driving mechanism 52 to be intermittently activated, so that the connecting portion 54 stops moving during the movement of the first visual inspection device 3 on the arc-shaped track 22. The achievable solution is as follows:

a plurality of position sensors are arranged on the annular track, and are configured to detect the moving position of the first visual detection device 3 and send a detection signal to the main control module;

if the main control module determines that the first visual inspection device 3 moves to the arc-shaped track 22, sending a stop instruction to the second driving mechanism 52; if the main control module determines that the first visual inspection device 3 leaves the arc-shaped track 22, it sends a start operation instruction to the second driving mechanism 52.

That is, in the case that the length of the crawler 53 is configured to be substantially equal to the length of the linear track 21, when the first visual inspection device 3 makes a linear motion, the connecting portion 54 may move synchronously therewith, in this embodiment, the main control module is further electrically connected to the first driving mechanism, and under the control of the main control module, the first driving mechanism drives the first visual inspection device to move on the linear track at the same speed as the second driving mechanism drives the crawler to move horizontally, preferably at a relative motion, so that the length of the extension conductive wire 56 may be shortened; when the first visual inspection device 3 moves to the arc-shaped rail 22, the second driving mechanism 52 is controlled to stop, and the rotation of the track is stopped, that is, the connection portion 54 stops at one end of the track, because it is close to the arc-shaped rail 22, so that the first visual inspection device 3 can be always connected with the extension conductive wire 56, which has a certain elongation, preferably like a telephone cord, and can be flexibly pulled by the first visual inspection device 3, that is, can be flexibly turned to the other linear rail 21 from the direction toward the one linear rail 21, during the movement of the first visual inspection device 3 on the arc-shaped rail 22.

The inspection is that the cloth woven by a plurality of looms is inspected in turn by one visual inspection device without correspondingly arranging one visual inspection device for each loom, and the inspection can be realized because the woven cloth of the looms has certain speed, the detection area of the visual inspection device has certain width, and the width of the cloth newly woven by the looms is smaller than the width of the detection area of the visual inspection device within one cycle of the visual inspection device on the premise that the inspection result is not missed, otherwise, the missed inspection is caused.

Therefore, if only one visual detection device is adopted, the number of the textile machines which can be inspected by the visual detection device is small, such as two textile machines; in an embodiment of the present invention, two or even more than three visual inspection devices may be adopted, and taking fig. 1 as an example, two rows of looms may be arranged below the weaving defect inspection device, and each row may be provided with four looms (one loom may be arranged between every two upright posts 11 in the length direction in fig. 1). Correspondingly, the weaving cloth defect inspection equipment further comprises a second visual detection device 4 which is arranged on the annular track and is configured to move along the annular track under the driving of a third driving mechanism; in a preferred embodiment, the first visual detection device 3 and the second visual detection device 4 are arranged at a distance equal to the sum of the lengths of a linear track 21 and an arc-shaped track 22, i.e. after the first visual detection device 3 has passed the first textile machine and has made a further half-turn of the circular track, the second visual detection device 4 comes to the upper region of the first textile machine.

After having adopted second visual inspection device 4, having brought new challenge for the wiping power supply, if set up another track, connecting portion and carbon brush, then not only increased the cost to different carbon brushes reciprocate on same busbar 51, and the moving direction is opposite, and the structure that will design to prevent interfering is also not simple thing, and to this, the embodiment provides following solution:

the second visual inspection device 4 is electrically connected with the connecting portion 54 and the carbon brush 55 through a wire embedded in the caterpillar 53, and the wire embedded in the caterpillar 53 and the caterpillar 53 move synchronously, so that the relative positions of the second visual inspection device 4 and the leading-out end of the wire embedded in the caterpillar 53 are kept unchanged during the movement of the second visual inspection device 4 on the linear track 21, and the relative arrangement is also preferably adopted to reduce the length of the wire.

The first visual inspection device 3 and the second visual inspection device 4 are configured to be spaced apart by a distance equal to the sum of the lengths of one straight rail 21 and one curved rail 22, and there is also an advantage in that the second visual inspection device 4 must enter one curved rail 22 at the same time when the first visual inspection device 3 enters the other curved rail 22; when the first visual inspection device 3 leaves one arc-shaped rail 22, the second visual inspection device 4 necessarily leaves the other arc-shaped rail 22, so that the start and stop of the second driving mechanism 52 are not required to be controlled according to the position of the second visual inspection device 4.

In the prior art, an endless track is realized by arranging a chain on the track, and driving the chain to move by using a driving mechanism, so as to drive a target object to do endless motion along the track. However, the chain is required to realize arc-shaped turning movement, and the movement speed is limited. As the inspection principle is described above, the moving speed of the visual inspection device is crucial to the length of the inspection path, and if the inspection moving speed of the visual inspection device is slow, in order to ensure that the newly woven width of the cloth is smaller than the width of the visual inspection area within one week of inspection, the inspection path is shortened, for example, the inspection can be performed only for two or four textile machines; if the inspection moving speed of the visual detection device is increased, the path of one week of inspection is increased, and inspection can be carried out on eight or even more textile machines. In this embodiment, the improved circular track structure can increase the moving speed of the visual inspection device, and the scheme is as follows:

as shown in fig. 5 and 6, the circular track includes an inner slide rail 61, an outer slide rail 62 and a sliding platform 63 configured to cover or partially cover the inner slide rail 61 and the outer slide rail 62, which are arranged side by side, and as can be seen in fig. 6, the inner slide rail 61 and the outer slide rail 62 may be cylindrical columns, which are horizontally arranged under the support of the corresponding two vertical columns 11, and the sliding platform 63 covers approximately three quarters of the inner slide rail 61 and the outer slide rail 62, so that the sliding platform 63 is prevented from flying off the track; the manner of mounting the slide table 63 to the inner slide rail 61 and the outer slide rail 62 is: firstly, the sliding table 63 is sleeved on a section of track single body (the linear track 21 or the arc track 22) from the side surfaces of the inner side sliding track 61 and the outer side sliding track 62, and then all the tracks are spliced into a closed annular track. Moreover, the steel balls 64 are embedded on the inner side surface of the sliding table 63, the sliding table 63 is in point contact with the inner side sliding rail 61 and the outer side sliding rail 62 through the steel balls 64 respectively, and the sliding table 63 is in contact with the inner side sliding rail 61 and the outer side sliding rail 62 through different balls respectively, so that the linear speed of the steel balls 64 in point contact with the outer side sliding rail 62 is naturally and automatically greater than the linear speed of the steel balls 64 in point contact with the inner side sliding rail 61 when passing through the arc-shaped rail 22. In a preferred embodiment, the inner side surface of the sliding table 63 is provided with four sets of steel balls 64, the number of each set of steel balls 64 is more than two, the distribution positions of the four sets of steel balls 64 are shown in fig. 5, and the sliding table 63 covers about three quarters of the number of the steel balls 64, so that the steel balls 64 cannot be separated from the sliding table 63.

The first visual detection device 3 and the first driving mechanism are fixedly arranged on the sliding table 63; accordingly, in the case where a plurality of visual inspection devices are provided, the number of the slide table 63 is also plural and corresponds to one another.

The above-mentioned improvement that is made to the connection structure between slip table and the track of placing visual inspection device only for improving the translation rate, the difficult problem that needs to solve next is, abandoned the mode that the chain drags the target object to move on the circular orbit among the prior art, has proposed one kind and can realize driving slip table 63 and slide on the gantry track that inboard slide rail 61 and outside slide rail 62 are constituteed, especially involves the solution that moves on the arc track 22, specifically as follows:

the side wall of the circular track is provided with a rack, the first driving mechanism comprises a motor and a gear, the gear is connected with the output shaft of the motor and is meshed with the rack, specifically referring to fig. 3, the rack is respectively arranged on the inner side wall of the arc track and the inner side wall of the linear track, and then a new technical problem is encountered, namely how to make the gear smoothly pass through the arc track without speed reduction, the technical scheme provided by the embodiment is referring to fig. 7:

arranging a workbench capable of moving in an arc shape, wherein a cutter for machining the rack is arranged on the workbench;

fixing an arc-shaped track of a rack to be processed on a circumferential path;

the driving mechanism is started to drive the workbench to do arc motion, the track of the arc motion is determined by simulating the track of the gear moving on the arc track, the rack processing of the arc track is completed, the tooth form of the inner side wall rack 65 of the arc track manufactured in the way is consistent with that of the inner side wall rack 66 of the linear track, and further the first visual detection device 3 can do uniform motion under the driving of the first driving mechanism, and the speed reduction can be avoided even when the gear passes through the arc track 22 (for the gear of the first driving mechanism, the tooth form of the inner side wall rack 65 of the arc track manufactured in the way is consistent with that of the inner side wall rack 66 of the linear track).

By adopting the structure of the gantry rail (the inner side sliding rail 61 and the outer side sliding rail 62), the rail structure that the steel balls 64 are embedded in the sliding table 63 to be respectively in point contact with the inner side sliding rail 61 and the outer side sliding rail 62 and the manufacturing method of the inner side wall rack 65 of the arc-shaped rail, the speed of the visual detection device in the weaving defect inspection equipment in the embodiment can reach about 6m/s, and the speed can be reduced without being reduced when the weaving defect inspection equipment is bent, and the constant speed range of the movement is kept to be 1-6 m/s. In another embodiment, the visual inspection device may be arranged to decelerate to less than 1m/s, such as 0.6m/s or 0.3m/s, while traversing an arcuate path, and to accelerate to greater than 1m/s, such as 2.5m/s or 3.5m/s or 5.5m/s, while traveling on a linear path. The image acquisition device is a non-linear scanning camera, in the process that the image acquisition device shoots the same piece of cloth twice before and after, the newly added woven cloth is shorter than the width of a shooting area of the image acquisition device, and the circular track is only provided with one visual detection device to meet the inspection requirement by taking the weaving speed of a weaving machine as 20 cm/min, the time for the visual device to move for one circle as 18s and the detection area width (in the weaving extending direction) of the visual device as 8cm as examples; however, the larger the detection area width of the vision device is set, the higher the requirement on the image recognition capability is, so that the detection area width (in the fabric extending direction) of the vision device can be properly reduced, and taking the detection area width of the vision device as 5.5cm as an example, the inspection requirement cannot be met by only installing one vision detection device on the circular track; taking the width of the detection area of the vision device as 2.5cm as an example, only two vision detection devices are arranged on the circular track, which cannot meet the inspection requirement, at least 3 vision detection devices should be arranged, and the arrangement is preferably equal to the distance.

Compared with the chain driving mode in the prior art, the rack-and-pinion driving mode in the embodiment has the following advantages: first, each visual detection device moves independently, for example, three visual detection devices are arranged on an annular track, so that the speed of the other two visual detection devices cannot be reduced because one visual detection device experiences an arc track, and the two visual detection devices do not interfere with each other; the chain type driving mode can only be provided with two visual detection devices; secondly, as the sliding seat is in point contact with the side surface of the sliding rail by adopting steel balls, the service life is longer than that of a chain, and particularly, the circular motion of the chain can be greatly abraded mechanically; thirdly, in the embodiment, the slide seat can be driven to move by adopting the servo motor, so that the driving position is more accurate.

In still another aspect, the present invention provides a weaving system including a plurality of looms (not shown) arranged side by side in two rows, and a weaving defect inspection apparatus as described above, wherein the plurality of looms in a single row are arranged side by side below one linear guide rail of the weaving defect inspection apparatus. The track is provided with a plurality of position sensors corresponding to the plurality of looms, and the main control module is used for analyzing the looms corresponding to the image acquisition devices at present according to the position detection results of the position sensors; and if the current output identification result of the AI identification device is that the cloth is defective, the main control module sends a stop instruction to the loom corresponding to the image acquisition device currently.

In addition, the invention also provides a weaving defect inspection method based on the weaving system, which comprises the following steps:

starting a first visual detection device, starting a first driving mechanism to drive the first visual detection device to move along an annular track, detecting the moving position of the first visual detection device, and controlling the starting and stopping of a second driving mechanism according to the position detection result, wherein the method comprises the following steps: if the first visual detection device moves to the arc-shaped track, the second driving mechanism stops driving; and if the first visual detection device leaves the arc-shaped track, the second driving mechanism is started, and the driving directions of two adjacent driving mechanisms of the second driving mechanism are opposite.

In a further embodiment, a plurality of position sensors are distributed on the circular track of the weaving defect inspection device in advance, and the main control module of the weaving defect inspection device is configured to analyze the loom corresponding to the first visual detection device according to the position detection results of the position sensors;

and if the first visual detection device detects that the cloth in the current detection area has defects, the main control module sends a stop instruction to the corresponding loom.

Besides the mode of arranging the position sensor, the running speed of the image acquisition device at each position on the routing inspection circular track can be configured, and the loom corresponding to the image acquisition device at present is calculated in real time by using the main control module according to the initial position and the time for starting to move of the image acquisition device.

It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.

The foregoing is directed to embodiments of the present application and it is noted that numerous modifications and adaptations may be made by those skilled in the art without departing from the principles of the present application and are intended to be within the scope of the present application.

Claims

1. A weaving defect detection method is used for detecting cloth woven by a loom and is characterized by comprising the following steps:

2. The woven fabric defect detection method according to claim 1, wherein the AI recognition device is a pre-established fabric defect detection neural network model, and the fabric defect detection neural network model is pre-established by the following steps:

3. The fabric defect detection method of claim 1, wherein the image acquisition device is a non-linear scanning camera, and the length of the newly added fabric is smaller than the width of the shooting area of the image acquisition device in the process of shooting the same piece of fabric twice before and after the image acquisition device.

4. The cloth weaving defect detection method of claim 3, wherein the image acquisition device is configured to move along a rail, and during the process of shooting the same piece of cloth twice before and after, the image acquisition device moves to above other textile machines to acquire images of the cloth woven at the textile machines.

5. The woven cloth defect detection method of claim 4, wherein a plurality of position sensors corresponding to a plurality of looms are arranged on the track, and the main control module is used for analyzing the loom corresponding to the image acquisition device currently according to the position detection results of the plurality of position sensors;

6. A weaving defect inspection method is characterized in that visual detection is circularly performed on cloth of different looms in a certain sequence based on weaving defect inspection equipment to identify whether the cloth has defects, and the weaving defect inspection method comprises the following steps:

7. The weaving defect inspection method according to claim 6, wherein the image acquired by the image acquisition device includes a cloth area and a warp yarn area, if the image characteristic difference between the cloth area and the adjacent pixel exceeds a preset threshold value and the warp yarn area has a warp broken line, the AI recognition device outputs a recognition result that the cloth is defective, otherwise, the AI recognition device outputs a recognition result that the cloth is not defective.

8. The cloth weaving defect inspection method according to claim 6, wherein the running speed of the image acquisition device at each position on the inspection circular track is configured, and a main control module is used for calculating the loom corresponding to the image acquisition device currently in real time according to the initial position and the time for starting moving of the image acquisition device.

9. The fabric defect inspection method according to claim 8, wherein the inspection circular track comprises two linear tracks and two arc tracks, wherein the two linear tracks are arranged side by side, the arc tracks are respectively connected with the end parts of the two linear tracks to form a closed circular track, and the circular track is horizontally erected on a plurality of stand columns;

10. The weaving defect inspection method according to claim 9, wherein the weaving defect inspection equipment comprises two image acquisition devices and sliding tables corresponding to the two image acquisition devices one by one, wherein the two image acquisition devices comprise a first image acquisition device and a second image acquisition device;