CN114638779B

CN114638779B - Textile quality inspection system, method, device, computer equipment and storage medium

Info

Publication number: CN114638779B
Application number: CN202110865072.2A
Authority: CN
Inventors: 黄灼; 刘琰; 黄锡雄; 陈随夫
Original assignee: Gizwits Iot Technology Co ltd
Current assignee: Gizwits Iot Technology Co ltd
Priority date: 2021-07-29
Filing date: 2021-07-29
Publication date: 2023-09-29
Anticipated expiration: 2041-07-29
Also published as: CN114638779A

Abstract

The application relates to a textile quality inspection system, a method, a device, a computer device and a storage medium. The system comprises: quality inspection equipment, a big data server, a learning server and image acquisition equipment; the learning server is used for acquiring a quality inspection learning sample from the big data server, generating a quality inspection model according to the quality inspection learning sample, and sending the quality inspection model to the quality inspection equipment; the image acquisition equipment is used for acquiring the textile image to be inspected in the spinning process and transmitting the textile image to be inspected to the quality inspection equipment; the quality inspection equipment is used for acquiring real-time operation data in the spinning process; the quality inspection equipment is also used for inputting the textile image to be inspected and the real-time operation data into the quality inspection model in the weaving process to obtain the working condition environment output by the quality inspection model; the working condition environment comprises a flaw identification result of the textile to be inspected and an operation state identification result of the second loom. The system can improve the production quality of textiles and the production efficiency of the textiles.

Description

Textile quality inspection system, method, device, computer equipment and storage medium

Technical Field

The present application relates to the technical field of textile quality detection, and in particular, to a textile quality detection system, a textile quality detection method, a textile quality detection apparatus, a computer device, and a computer readable storage medium.

Background

The quality requirements of the industry are also improved while the textile processing capacity is improved, so that the requirements of users for quickly and accurately finding quality stains are stronger in the textile production process. However, the conventional textile production line mainly judges textile flaws by means of manual quality inspection.

The traditional manual quality inspection method is too dependent on personal experience of users, errors and missed inspection often occur, so that the stability of the large-scale textile production quality cannot be effectively ensured, and the increasingly improved textile production quality standard cannot be ensured. In addition, the manual quality inspection process needs to consume more time, so that the production efficiency of the textile fabric can be greatly influenced.

Therefore, the traditional textile quality inspection method has the problems that the textile production quality cannot be guaranteed and the textile production efficiency is affected.

Disclosure of Invention

In view of the foregoing, it is desirable to provide a textile quality inspection system, a textile quality inspection method and apparatus, a computer device, and a computer-readable storage medium.

In a first aspect, the present invention provides a textile quality inspection system comprising:

quality inspection equipment, a big data server, a learning server and image acquisition equipment; the big data server is used for collecting quality inspection learning samples; the quality inspection learning sample comprises a sample flaw image and sample operation data; the sample flaw image at least comprises an image obtained by shooting the flaw textile produced by the first loom; the sample operation data are used for reflecting the operation state of the first loom when the defective textile is produced;

the learning server is used for acquiring the quality inspection learning sample from the big data server, generating a quality inspection model according to the quality inspection learning sample, and sending the quality inspection model to the quality inspection equipment;

the image acquisition equipment is used for acquiring an image of the textile to be inspected in the spinning process and transmitting the image of the textile to be inspected to the quality inspection equipment; the textile image to be inspected is an image obtained by shooting the textile to be inspected produced by the second loom;

the quality inspection equipment is used for acquiring real-time operation data in the spinning process; the real-time operation data are used for reflecting the operation state of the second loom when the textile to be inspected is produced;

The quality inspection equipment is also used for inputting the textile image to be inspected and the real-time operation data into the quality inspection model in the spinning process to obtain the working condition environment output by the quality inspection model; the working condition environment comprises a flaw identification result of the textile to be inspected and an operation state identification result of the second loom; the operation state recognition result includes a judgment result of whether the second loom is in an abnormal operation state.

In a second aspect, the present invention provides a method of quality inspection of textile fabrics, comprising:

receiving a quality inspection model issued by a learning server; the learning server is used for acquiring a quality inspection learning sample from the big data server and generating the quality inspection model according to the quality inspection learning sample; the big data server is used for collecting the quality inspection learning sample; the quality inspection learning sample comprises a sample flaw image and sample operation data; the sample flaw image at least comprises an image obtained by shooting the flaw textile produced by the first loom; the sample operation data are used for reflecting the operation state of the first loom when the defective textile is produced;

receiving the textile image to be inspected acquired by the image acquisition equipment, and acquiring real-time operation data; the textile image to be inspected is an image obtained by shooting the textile to be inspected produced by the second loom; the real-time operation data are used for reflecting the operation state of the second loom when the textile to be inspected is produced;

Inputting the textile image to be inspected and the real-time operation data into the quality inspection model to obtain a working condition environment output by the quality inspection model; the working condition environment comprises a flaw identification result of the textile to be inspected and an operation state identification result of the second loom; the operation state recognition result includes a judgment result of whether the second loom is in an abnormal operation state.

In a third aspect, the present invention provides a textile quality inspection device comprising:

the model receiving module is used for receiving the quality inspection model issued by the learning server; the learning server is used for acquiring a quality inspection learning sample from the big data server and generating the quality inspection model according to the quality inspection learning sample; the big data server is used for collecting the quality inspection learning sample; the quality inspection learning sample comprises a sample flaw image and sample operation data; the sample flaw image at least comprises an image obtained by shooting the flaw textile produced by the first loom; the sample operation data are used for reflecting the operation state of the first loom when the defective textile is produced;

the image receiving module is used for receiving the textile image to be inspected acquired by the image acquisition equipment and acquiring real-time operation data; the textile image to be inspected is an image obtained by shooting the textile to be inspected produced by the second loom; the real-time operation data are used for reflecting the operation state of the second loom when the textile to be inspected is produced;

The working condition acquisition module is used for inputting the textile image to be inspected and the real-time operation data into the quality inspection model to obtain a working condition environment output by the quality inspection model; the working condition environment comprises a flaw identification result of the textile to be inspected and an operation state identification result of the second loom; the operation state recognition result includes a judgment result of whether the second loom is in an abnormal operation state.

In a fourth aspect, the present application provides a computer device comprising a memory storing a computer program and a processor implementing the textile quality inspection method of the second aspect described above when the computer program is executed by the processor.

In a fifth aspect, the present application provides a computer readable storage medium having stored thereon a computer program which when executed by a processor implements the method of quality inspection of textiles of the second aspect described above.

According to the textile quality inspection system, the method, the device, the computer equipment and the computer readable storage medium, the big data server collects the quality inspection learning sample, the learning server trains the quality inspection model based on the quality inspection learning sample, the quality inspection equipment inputs the textile image to be inspected acquired from the image acquisition equipment and the real-time operation data obtained from the loom into the quality inspection model, so that the working condition environment is obtained, a user can know the quality condition of textiles produced by the loom and the defect recognition result of whether serious defect problems exist, and the whole quality inspection process does not need to rely on manual quality inspection, so that the problems of common errors, omission and the like in the manual quality inspection are avoided, and meanwhile, the quality inspection speed is improved. Therefore, the production quality of the textile is improved, and meanwhile, the production efficiency of the textile is improved. In addition, the operation data of the loom and the corresponding flaw images are fused to be used as quality inspection learning samples, so that the model is trained based on two data with different properties of equipment state data and visual identification data. In addition, the textile quality inspection system carries out real-time quality inspection on textiles produced by the loom in the textile process, and the quality inspection on the produced textiles is ensured and the influence of quality inspection on the production efficiency is avoided by carrying out quality inspection on the textiles while producing the textiles.

Drawings

FIG. 1 is a block diagram of a textile quality inspection system according to one embodiment;

FIG. 2 is a schematic diagram of an exemplary textile quality inspection system in one embodiment;

FIG. 3 is a schematic view of a textile quality inspection system deploying a plurality of image acquisition devices, according to one embodiment;

FIG. 4 is a schematic diagram of an example continuous fault alerting procedure of one embodiment;

FIG. 5 is a schematic flow chart of a method of inspecting textile quality in accordance with one embodiment;

FIG. 6 is a block diagram of a textile quality inspection device according to one embodiment;

FIG. 7 is an internal block diagram of a computer device of one embodiment.

Detailed Description

The present application will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.

In one embodiment, as shown in fig. 1, a textile quality inspection system 100 is provided, and the textile quality inspection system 100 may include a quality inspection device 110, a big data server 120, a learning server 130, and an image acquisition device 140. In practical applications, the big data server 120 and the learning server 130 may be a single server or a server cluster.

The quality inspection device 110 may be a device with certain data processing capability and data communication capability for being deployed on a loom. Fig. 2 is a schematic diagram of an exemplary textile quality inspection system in one embodiment. As shown, quality inspection device 110 may communicate with image capture device 140, big data server 120, learning server 130 for processing of data interactions, receiving instructions, data synchronization, and the like. In practical applications, the quality inspection device 110 may be a robot with mobility, or a computer device fixed at a specific location. For the quality inspection device 110 in the form of a robot, it is possible to move between the respective looms, so that the quality inspection of the textile can be performed separately for the respective looms. Those skilled in the art may use different types of devices as the quality inspection device 110 according to actual needs. The textile can be an article produced by a weaving mode, and the material can be cotton, wool, silk, hemp and the like. In the production process, it is usually in the form of a cloth, and is therefore also referred to as cloth or cloth.

The big data server 120 may be a server for collecting, storing a large amount of data about the operation of textiles and equipment. The learning server 130 may be a server for performing machine learning, and the learning server 130 may communicate with the big data server 120 to acquire data for performing machine learning from the big data server 120.

The image pickup device 140 may be a device having an image capturing function, for example, an image pickup device for ultra-high definition industry of the order of hundred million pixels.

The quality inspection learning sample may include a sample flaw image and sample operation data, where the sample flaw image may be an image obtained by photographing a flaw textile produced by the first loom; the sample operation data is used for reflecting the operation state of the first loom when the first loom produces defective textiles.

The big data server 120 may communicate with a plurality of looms. In the course of producing textiles by the loom, the image capturing device 140 disposed around the loom may capture images of the textiles produced by the loom and transmit the captured images of the textiles to the quality inspection device 110. In addition, quality inspection device 110 may communicate with the loom to detect the operational status of the loom in producing the textile, forming operational data. The quality inspection device 110 may upload the received textile image and the operation data obtained by detecting the loom to the big data server 120. The big data server 120 may perform cleaning, classifying, etc. on the textile image and the operation data, to obtain a textile image reflecting the defective textile as a quality inspection learning sample, i.e. the sample defect image, and obtain operation data corresponding to the sample defect image as a quality inspection learning sample, i.e. the sample operation data. Thus, the big data server 120 can collect a large number of learning samples from a plurality of first looms.

In one embodiment, big data server 120 may collect hundreds of millions of operational data and hundreds of millions of flaw images for tens of thousands of looms by communicating with quality inspection device 110, and obtain a large number of sample operational data and sample flaw images by performing data cleaning, data sorting, etc. on the hundreds of millions of operational data and hundreds of millions of flaw images, forming a predictive maintenance database and a sample flaw image database for machine learning by learning server 130.

Wherein, the flaws of the textile can comprise broken warp, wrong reed, double warp, reed path, poor selvedge, wrong heald, tight warp, loose warp, different warp strips, greasy dirt, broken hole, broken weft, weft shrinkage, double weft, dense path, jump pattern, wave pattern and the like. The operating data may then be in particular data of the number of revolutions, pressure etc. of the weaving machine.

For the purpose of distinguishing the description, the loom corresponding to the sample flaw image and the sample operation data is named as a first loom.

The learning server 130 is configured to obtain a quality inspection learning sample from the big data server 120, generate a quality inspection model according to the quality inspection learning sample, and send the quality inspection model to the quality inspection device 110.

Specifically, the learning server 130 may obtain a large number of quality inspection learning samples including sample flaw images and sample operation data from the big data server 120, train the quality inspection model by using the association relationship between the sample flaw images and the sample operation data in the quality inspection learning samples, and then issue the trained quality inspection model to the quality inspection device 110.

The image acquisition device 140 is used for acquiring the textile image to be inspected in the spinning process and transmitting the textile image to be inspected to the quality inspection device 110; the textile image to be inspected is an image obtained by shooting the textile to be inspected produced by the second loom.

Wherein, the second loom is different from the first loom, and corresponds to the loom currently performing the quality inspection treatment of the textile fabric by the quality inspection device 110.

In particular, one or more image acquisition devices 140 may be deployed around the perimeter of the second loom. In the process of quality inspection by the quality inspection device 110, the image acquisition device 140 may photograph the textile to be inspected produced by the second loom to obtain the above image of the textile to be inspected, and transmit the image of the textile to be inspected to the quality inspection device 110.

The quality inspection device 110 is used for acquiring real-time operation data in the spinning process; the real-time operation data are used for reflecting the operation state of the second loom when the textile to be inspected is produced.

Specifically, during the quality inspection performed by the quality inspection device 110, the quality inspection device 110 may detect the second loom in operation, and obtain operation data reflecting the operation state of the second loom as the above-mentioned real-time operation data. Thus, the quality inspection device 110 obtains the textile image to be inspected and the corresponding real-time operation data transmitted by the image acquisition device 140.

The quality inspection device 110 is further configured to input the textile image to be inspected and the real-time operation data into the quality inspection model in the weaving process, so as to obtain a working condition environment output by the quality inspection model; the working condition environment comprises a flaw identification result of the textile to be inspected and an operation state identification result of the second loom; the operation state recognition result includes a judgment result of whether the second loom is in an abnormal operation state.

Specifically, the quality inspection device 110 may input the image of the textile to be inspected and the real-time operation data into the quality inspection model, and the quality inspection model may output the working condition environment correspondingly, where the working condition environment may specifically include whether the textile to be inspected produced by the second loom has a defect problem and a component in the second loom that may cause the defect problem. The user can judge whether maintenance measures such as maintenance or replacement of parts causing flaw problems in the second loom are required or not based on the working condition environment.

For example, defects such as a gear, shrinkage, etc. may be caused by mechanical failure or abnormal operation of the loom, and thus, the loom needs to be maintained in time, which would result in the production of a large amount of defective textiles. Therefore, when the working condition environment indicates that the textile to be inspected has a certain flaw and is caused by the mechanical failure of a certain part of the loom, a user can repair or replace the related part of the second loom according to the working condition environment.

In practical applications, the quality inspection device 110 may further determine whether a serious defect problem will continue to occur due to the severely failed component in the second loom according to the working condition environment, so that the second loom may be controlled to stop operating, and a large number of defective textiles are prevented from being produced continuously.

In practice, as shown in fig. 2, quality inspection device 110 of textile quality inspection system 100 may also communicate with a manufacturing execution system (Manufacturing Execution Systems, MES). Wherein, MES can carry out the optimization management to the whole production process from order to product completion through information transfer. The MES also provides critical task information about product behavior within the enterprise and throughout the product supply chain via two-way direct communication. The quality inspection device 110 is in communication with the MES to provide support for digital production data in the manufacturing process, and the MES data in the textile production process can be combined to continuously improve the cooperative operation capability of the devices, improve the productivity of the devices and reduce the internal loss.

On the one hand, in the textile production line, in the quality inspection process after the textile is produced, whether the color of the textile is qualified or whether the textile is knitted is qualified or not needs to be identified in real time for a long time, which is a very heavy work for quality inspectors. When the operation speed of the textile production line is high, the speed of manual quality inspection is required to be 15-20 meters of textile per minute, the visual inspection range of quality inspection staff is limited, the quality inspection of the textile with the width of 0.8-1 meter can be finished on average, and in all production links, the manual quality inspection link is the most easy bottleneck of the textile production flow. On the other hand, the manual quality inspection mode extremely depends on the quality inspection experience of quality inspection personnel, even if the quality inspection personnel are skilled, errors and missing inspection still occur frequently, and the quality inspection stability is poor, so that the quality of the produced textile is influenced.

Based on the two problems, the inventor provides a textile quality inspection system constructed based on the Internet of things technology, and the textile image acquired by the image acquisition device in real time and the operation data acquired by the loom in real time are input into the quality inspection model by the quality inspection device by interconnecting a big data server, a learning server, quality inspection equipment and image acquisition equipment and utilizing the capability of the big data server for collecting and processing big data and combining the learning server to perform machine learning capability so as to generate a quality inspection model, so that a working condition environment is obtained. The quality inspector can know the textile production condition and the information of whether the loom needs to be maintained or not according to the working condition environment.

In the textile quality inspection system, the quality inspection learning sample is collected by the big data server, the quality inspection model is trained by the learning server based on the quality inspection learning sample, the quality inspection equipment inputs the textile image to be inspected acquired from the image acquisition equipment and the real-time operation data obtained from the loom into the quality inspection model, so that the working condition environment is obtained, a user can know the quality condition of the textile produced by the loom and the defect recognition result of whether serious defect problems exist, and the whole quality inspection process is free from depending on manual quality inspection, so that the problems of common errors, missed inspection and the like in the manual quality inspection are avoided, and meanwhile, the quality inspection speed is also improved. Therefore, the production quality of the textile is improved, and meanwhile, the production efficiency of the textile is improved. In addition, the operation data of the loom and the corresponding flaw images are fused to be used as quality inspection learning samples, so that the model is trained based on two data with different properties of equipment state data and visual identification data. In addition, the textile quality inspection system carries out real-time quality inspection on textiles produced by the loom in the textile process, and the quality inspection on the produced textiles is ensured and the influence of quality inspection on the production efficiency is avoided by carrying out quality inspection on the textiles while producing the textiles.

Furthermore, quality inspection is performed by a quality inspection model obtained by training based on a large number of quality inspection learning samples, so that various flaws can be stably and accurately identified, the operation state identification result of the loom can be obtained based on the association relation between the flaws and the operation state of the loom, and a user can correspondingly maintain the loom according to the operation state identification result.

Further, a large number of quality inspection learning samples can be collected through the big data server, so that various types of flaws can be accurately identified by the generated quality inspection model, and the accuracy of quality inspection is improved.

In one embodiment, quality inspection device 110 is further configured to determine whether an abnormality exists in the operating environment, and when it is determined that an abnormality exists in the operating environment, perform an operating condition abnormality alarm and/or control the second loom to stop operating.

Specifically, quality inspection device 110 may also automatically perform an operation of a condition anomaly alarm and/or an operation of controlling the second loom to stop functioning according to the operating environment.

The quality inspection device 110 may determine whether the operation state recognition result in the working condition environment belongs to a serious abnormal operation condition that may continuously cause more defective textiles, for example, if the operation state recognition result is that a positioning component of the loom fails, the textile that is subsequently produced may have the same defect, so the quality inspection device 110 may perform the working condition abnormality alarm to notify the user to perform timely processing. The quality inspection device 110 can also directly control the second loom to stop running, so as to avoid continuously producing more textiles with serious flaw problems.

According to the textile quality inspection system, the quality inspection equipment automatically alarms for abnormal working conditions and/or controls the second loom to stop operating under the condition that the working conditions are abnormal, so that the abnormal loom is prevented from continuously producing textiles with serious flaw problems. Moreover, by controlling the loom to stop running, the use of the stop sheet can be reduced, and the production cost is saved.

In one embodiment, the quality inspection models are multiple, and the multiple quality inspection models correspond to different textile fabrics respectively; the quality inspection device 110 is specifically configured to identify a target textile fabric from the image of the textile fabric to be inspected, determine a quality inspection model corresponding to the target textile fabric, and input the image of the textile fabric to be inspected and the real-time operation data into the quality inspection model corresponding to the target textile fabric.

Specifically, for different textile fabrics, the learning server 130 may train the corresponding quality inspection model accordingly, and the quality inspection device 110 may perform quality inspection by using the corresponding quality inspection model according to the textile fabric of the textile fabric currently produced by the second loom, so that the quality inspection of the textile fabric may be performed more specifically and more accurately.

In one embodiment, image capture device 140 may include a first image capture device 141, a second image capture device 142, a third image capture device 143; the first image acquisition device 141 is used for shooting the front surface of the textile to obtain a front surface image of the textile; the second image acquisition device 142 is used for shooting the back surface of the textile to obtain a textile back surface image; the third image acquisition device 143 is used for shooting the edge of the textile to obtain an edge image of the textile; the textile image to be inspected comprises a textile front image, a textile back image and a textile edge image.

In particular, in textile quality inspection system 110, one or more image acquisition devices 140 may be deployed around the loom perimeter. Fig. 3 is a schematic diagram of a plurality of image acquisition devices of a textile quality inspection system according to one embodiment. As shown, three image capture devices 140 are deployed, a first image capture device 141, a second image capture device 142, and a third image capture device 143, respectively. Wherein the first image capturing device 141 captures the front side of the cloth produced by the loom, and the obtained image is a textile front side image. The second image pickup device 142 picks up the back side of the cloth produced by the loom, and the obtained image is a textile back side image. The third image capturing device 143 captures an edge of the cloth produced by the loom, and the obtained image is an edge image of the textile.

Of course, one skilled in the art may deploy at least one of the three image acquisition devices as desired. For example, only the first image pickup device 141 and the third image pickup device 143 are disposed for photographing the front and the edge of the cloth, respectively.

It should be noted that, the edge of the textile often has defects that are easy to miss, so that the edge of the textile is photographed by adding the third image collecting device 143 to obtain an image of the edge of the textile, so that the quality of the textile is more comprehensively checked, and missing defects are avoided.

According to the textile quality inspection system, the first image acquisition equipment, the second image acquisition equipment and the third image acquisition equipment are deployed to respectively shoot the front face, the back face and the edge of the textile, so that the quality inspection equipment can comprehensively inspect the front face image of the textile, the back face image of the textile and the edge image of the textile, and missing flaws are avoided.

In one embodiment, textile quality inspection system 100 may further include a remote data center; the quality inspection device 110 is further configured to record a flaw location when a flaw is identified in at least one of the textile front image, the textile back image, and the textile edge image, and upload the flaw location and the real-time operation data to a remote data center; the remote data center is used for generating a quality inspection report according to the flaw positions and the real-time operation data.

In particular, textile quality inspection system 100 may also deploy a remote data center for generating quality inspection reports. The quality inspection apparatus 110 may further locate the position of the flaw in the textile when the flaw is recognized from at least one of the textile front image, the textile back image, and the textile edge image at the time of textile quality inspection, and record as the flaw position. The quality inspection device 110 may upload the recorded flaw location and the corresponding real-time operation data to a remote data center, and the remote data center may generate a quality inspection report in which the flaw location and the corresponding real-time operation data are recorded, so that the user may download and review the quality inspection report from the remote data center.

In the textile quality inspection system, the quality inspection device can locate and record the flaw position when the flaw is identified in at least one textile image of the textile front image, the textile back image and the textile edge image, and upload the flaw position and corresponding real-time operation data to a remote data center to generate a quality inspection report, and a user can know the relevant operation state when the flaw problem of the loom occurs through the quality inspection report, so that the user can obtain more valuable information about the quality inspection result through the textile quality inspection system.

In one embodiment, learning server 130 is also configured to generate a decision model and send the decision model to quality inspection device 110; the quality inspection device 110 is further specifically configured to input a working condition environment into the decision model to obtain a target execution decision; the target execution decision comprises working condition abnormality alarming and/or controlling the second loom to stop operating.

Specifically, the learning server 130 may also generate a decision model according to the association between various operating conditions and various candidate decisions, and issue the decision model to the quality inspection device 110. The quality inspection device 110 may input the target execution decision to the decision model, so as to obtain a target execution decision which is output by the decision model and is adapted to the input working condition environment, and the quality inspection device 110 may execute the target execution decision, for example, perform a working condition abnormality alarm to a user or directly control the second loom to stop operating. Therefore, the textile quality inspection system can adopt corresponding decisions aiming at different working condition environments to ensure the production quality of textiles, and the flexibility of quality inspection is improved.

In one embodiment, textile quality inspection system 100 also includes an alarm device; the alarm device comprises at least one of an alarm lamp and an alarm loudspeaker; the image acquisition device 140 is further configured to acquire continuous multi-frame images of the textile to be inspected at the same detection position; the textile fabric image to be inspected has a corresponding detection position; quality inspection device 110 is specifically configured to:

Determining the textile image to be inspected, of which the textile flaws are identified, as a first flaw image; recording a detection position corresponding to the first flaw image as a flaw position; and triggering the alarm device to alarm when the detection position is detected to be recorded as a second flaw image of the flaw position.

Wherein the detection position is the position to be detected in the textile. The image acquisition device 140 may perform image acquisition for a specific position in the textile, so as to acquire an image of the textile to be inspected corresponding to the detected position.

Specifically, textile quality inspection system 100 may be provided with an alarm device that may be in wired/wireless communication with the quality inspection device, and the alarm device may alarm according to an indication of the quality inspection device to alert a user to timely handle when a quality inspection problem occurs. The alarm device may be embodied as a warning light, a warning loudspeaker or a combination thereof. The warning lamp can be turned on to give an alarm, and the warning megaphone can give an alarm to give an alarm.

The image acquisition device 140 can continuously acquire multiple frames of images aiming at the same detection position so as to obtain continuous multiple frames of textile images to be inspected. The corresponding detection position can be recorded for the textile image to be inspected.

It should be noted that, in the textile industry, if continuous flaws occur in the warp direction, it may result in that the whole roll of finished product (e.g. 300 m of cloth) is produced as a reject. Therefore, how to increase the detection rate of continuous flaws and reduce the false-negative rate of continuous flaws is a problem to be solved in the textile industry.

To avoid continuous flaws affecting product quality, quality inspection device 110 may trigger an alarm device to alarm based on a certain quality discrimination algorithm. When the quality inspection device 110 performs quality inspection, if a textile defect is identified in one of the textile images to be inspected, the textile image to be inspected may be used as the defect image. For the sake of distinguishing the description, it is named as the first defective image. Then, the detection position corresponding to the first flaw image may be recorded as a flaw position, but the alarm device may not be triggered to alarm. At this point, quality inspection device 110 may continue to inspect the quality of the other textile images to be inspected.

In the process of continuing to perform quality inspection on other textile images to be inspected, when detecting that a textile defect is identified in another textile image to be inspected, the quality inspection device 110 may first acquire a corresponding detection position, determine whether the detection position is recorded as a defect position, if so, indicate that the defect repeatedly occurs at the target detection position, and the textile image to be inspected is the second defect image, and may have an abnormal condition of continuous defects at present, so the quality inspection device 110 may trigger the alarm device to alarm.

After the quality inspection device 110 triggers the alarm device to alarm, the quality inspection of the images of other textiles to be inspected is continued.

According to the textile quality inspection system, the image acquisition equipment acquires continuous multi-frame textile quality inspection images at the same detection position, when textile flaws are identified in the textile quality inspection images, the textile quality inspection images are determined to be first flaw images, the detection positions corresponding to the first flaw images are recorded to be flaw positions, and when the second flaw images appear at the same detection position, the alarm equipment is triggered to alarm. Therefore, the potential continuous flaws are guaranteed to be timely alarmed, the detection rate of the continuous flaws is improved, and the missing report rate of the continuous flaws is reduced.

Moreover, alarm is carried out through warning light, warning megaphone, has guaranteed that the warning can in time be conveyed to the user, avoids the user to miss the warning of important continuous flaw.

In one embodiment, after triggering the alarm device to alarm, the quality inspection device 110 is further specifically configured to:

when the textile image to be inspected, which is detected to be defective in textile, does not exceed M frames in N frames of textile images to be inspected, which are continuous with the first defect image, triggering the alarm equipment to stop alarming; n is more than M;

When the textile image to be inspected, in which textile flaws are identified in the N frames of textile image to be inspected consecutive to the first flaw image, exceeds M frames, the quality inspection apparatus is further specifically configured to:

when the textile fabric images to be inspected are in P frames continuous with the first flaw image, and the textile fabric images to be inspected continuously are not more than Q frames, triggering the alarm equipment to stop alarming; p is more than N, M is more than or equal to Q.

In the alarm for the continuous flaws, not only the rate of missing the continuous flaws is required to be reduced, but also the rate of false alarms of the continuous flaws is required to be reduced, so that the production efficiency is prevented from being influenced by false alarms. In order to reduce the rate of continuous defect false alarm while reducing the rate of continuous defect false alarm, the quality inspection device 110 may trigger an alarm device to alarm based on a certain quality identification algorithm.

Specifically, after triggering the alarm device to alarm, the quality inspection device 110 may continue to track the defect recognition condition of the textile image to be inspected at the defect position. The quality inspection device 110 may acquire N images of the textile to be inspected, which are continuous with the first flaw image, and when the image of the textile to be inspected, in which the textile flaw is identified, does not exceed M frames, the continuous flaw at the flaw position is illustrated to disappear, so that the alarm device may be triggered to stop alarming. In addition, quality inspection device 110 may generate an alarm record for reference by the user.

And when textile flaws are identified in more than M frames of textile images to be inspected in the N frames of textile images to be inspected, the alarm device continuously alarms. The quality inspection device 110 continues to track the defect recognition condition of the textile image to be inspected at the defect position, and when the textile defect is recognized in the textile image to be inspected of the continuous P frames with the first defect image and the continuous textile image to be inspected does not exceed Q frames, the alarm device is triggered to stop alarming, and in addition, the quality inspection device 110 can generate an alarm record for reference of a user.

Correspondingly, when the textile fabric image to be inspected is detected in the P frames continuous with the first flaw image, and the continuous textile fabric image to be inspected exceeds Q frames, the alarm device continuously alarms.

From the alarm processing process, the whole quality inspection and alarm processing are in dynamic updating, the alarm can be triggered under the condition that the continuity flaw appears, and the alarm is stopped after the condition of the continuity flaw disappears, so that the abnormal condition can be guaranteed to be timely alarmed, and meanwhile, the alarm can be timely stopped after the abnormal condition disappears, and the influence on the normal production efficiency due to false alarm is avoided.

In practical applications, the specific value of M, N, P, Q can be set by those skilled in the art according to actual needs.

FIG. 4 is a schematic diagram of an example of a continuous fault alert procedure for one embodiment. As shown, first, when the quality inspection device 110 finds a textile flaw in an image acquired by the image acquisition device 140 for a certain detection position, recording is performed for that detection position. When the defect of the textile fabric is found again in the next frame image of the same detection position, the alarm device is triggered to alarm, and a plurality of subsequent frame images of the detection position are continuously detected. Triggering an alarm device to stop alarming when the image with textile flaws in the continuous 4 frames of images of the detection position is not more than 3 frames; and if not, triggering the alarm equipment to stop alarming.

In practical application, a processing confirmation button can be further arranged, and after the user gets an alarm, the user can click the processing confirmation button to trigger the alarm device to stop alarming.

According to the textile quality inspection system, after the alarm is triggered, continuous N frames of textile images to be inspected at the flaw position are continuously detected, the alarm is stopped when the textile images to be inspected at the flaw position of the textile are not more than M frames, and otherwise, the alarm is continuously performed; and in an alarm state, the quality inspection equipment continuously detects continuous P frames of textile images to be inspected at the flaw position, and when the textile flaws are identified and the continuous textile images to be inspected do not exceed Q frames, the alarm is stopped, otherwise, the alarm is continued. Therefore, the continuous flaws are guaranteed to be timely alarmed, the continuous flaw false alarm rate is reduced, meanwhile, the alarm is stopped under specific conditions, the continuous alarm is avoided under the condition that discontinuous flaws or continuous flaws disappear, and the continuous flaw false alarm rate is reduced.

In one embodiment, as shown in fig. 5, a method for inspecting quality of textile is provided, which is exemplified by the method applied to the quality inspection apparatus 110 in fig. 1, and includes the following steps:

s502, receiving a quality inspection model issued by a learning server; the learning server is used for acquiring a quality inspection learning sample from the big data server and generating the quality inspection model according to the quality inspection learning sample; the big data server is used for collecting the quality inspection learning sample; the quality inspection learning sample comprises a sample flaw image and sample operation data; the sample flaw image at least comprises an image obtained by shooting the flaw textile produced by the first loom; the sample operation data are used for reflecting the operation state of the first loom when the defective textile is produced;

S504, receiving the textile image to be inspected acquired by the image acquisition equipment, and acquiring real-time operation data; the textile image to be inspected is an image obtained by shooting the textile to be inspected produced by the second loom; the real-time operation data are used for reflecting the operation state of the second loom when the textile to be inspected is produced;

s506, inputting the textile image to be inspected and the real-time operation data into the quality inspection model to obtain a working condition environment output by the quality inspection model; the working condition environment comprises a flaw identification result of the textile to be inspected and an operation state identification result of the second loom; the operation state recognition result includes a judgment result of whether the second loom is in an abnormal operation state.

Since the relevant steps of the quality inspection device 110 have already been described in the above embodiments, they are not described in detail herein.

In the textile quality inspection method, the large data server collects the quality inspection learning sample, the learning server trains the quality inspection model based on the quality inspection learning sample, the quality inspection equipment inputs the textile image to be inspected acquired from the image acquisition equipment and the real-time operation data acquired from the loom into the quality inspection model, so that the working condition environment is obtained, a user can know the quality condition of the textile produced by the loom and the defect recognition result of whether serious defect problems exist, and the whole quality inspection process does not need to rely on manual quality inspection, so that the problems of common errors, missed inspection and the like in the manual quality inspection are avoided, and the quality inspection speed is improved. Therefore, the production quality of the textile is improved, and meanwhile, the production efficiency of the textile is improved. In addition, the application takes the operation data of the loom and the corresponding flaw image fusion as a quality inspection learning sample, so that the model is trained based on two data with different properties of equipment state data and visual identification data. In addition, the textile quality inspection system carries out real-time quality inspection on textiles produced by the loom in the textile process, and the quality inspection on the produced textiles is ensured and the influence of quality inspection on the production efficiency is avoided by carrying out quality inspection on the textiles while producing the textiles.

Furthermore, quality inspection is performed by a quality inspection model obtained by training based on a large number of quality inspection learning samples, so that various flaws can be stably and accurately identified, and an operation state identification result of the loom can be obtained based on the association relation between the flaws and the operation state of the loom.

In one embodiment, the method for inspecting textile quality may further include the steps of:

judging whether the working condition environment is abnormal or not; and when the working condition environment is judged to be abnormal, working condition abnormal alarm is carried out and/or the second loom is controlled to stop operating.

In one embodiment, step S506 may specifically include the following steps:

identifying a target textile fabric from the textile fabric image to be inspected; determining a quality inspection model corresponding to the target textile fabric; and inputting the textile image to be inspected and the real-time operation data into an inspection model corresponding to the target textile fabric, and obtaining the working condition environment output by the inspection model.

In one embodiment, the image capture device comprises a first image capture device, a second image capture device, and a third image capture device; the first image acquisition equipment is used for shooting the front surface of the textile to obtain a front surface image of the textile; the second image acquisition equipment is used for shooting the back surface of the textile to obtain a textile back surface image; the third image acquisition equipment is used for shooting the edge of the textile to obtain an edge image of the textile; the textile image to be inspected comprises the textile front image, the textile back image and the textile edge image, and the textile quality inspection method can further comprise the following steps:

recording a flaw position when a textile flaw is identified in at least one of the textile front image, the textile back image and the textile edge image; uploading the flaw location and the real-time operation data to a remote data center; and the remote data center is used for generating a quality inspection report according to the flaw positions and the real-time operation data.

Receiving a decision model sent by the learning server; inputting the working condition environment into the decision model to obtain a target execution decision; the target execution decision comprises working condition abnormality alarming and/or controlling the second loom to stop operating.

Since the above related steps of the quality inspection device 110 are already described in the above embodiments, they are not described in detail herein.

It should be understood that, although the steps in the flowchart of fig. 5 are shown in sequence as indicated by the arrows, the steps are not necessarily performed in sequence as indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in fig. 5 may include multiple sub-steps or stages that are not necessarily performed at the same time, but may be performed at different times, nor do the order in which the sub-steps or stages are performed necessarily performed in sequence, but may be performed alternately or alternately with at least a portion of the sub-steps or stages of other steps or other steps.

In one embodiment, as shown in fig. 6, there is provided a textile quality inspection device comprising:

the model receiving module 602 is configured to receive a quality inspection model issued by the learning server; the learning server is used for acquiring a quality inspection learning sample from the big data server and generating the quality inspection model according to the quality inspection learning sample; the big data server is used for collecting the quality inspection learning sample; the quality inspection learning sample comprises a sample flaw image and sample operation data; the sample flaw image at least comprises an image obtained by shooting the flaw textile produced by the first loom; the sample operation data are used for reflecting the operation state of the first loom when the defective textile is produced;

the image receiving module 604 is configured to receive the image of the textile to be inspected acquired by the image acquisition device, and acquire real-time operation data; the textile image to be inspected is an image obtained by shooting the textile to be inspected produced by the second loom; the real-time operation data are used for reflecting the operation state of the second loom when the textile to be inspected is produced;

the working condition acquisition module 606 is configured to input the textile image to be inspected and the real-time operation data to the quality inspection model, so as to obtain a working condition environment output by the quality inspection model; the working condition environment comprises a flaw identification result of the textile to be inspected and an operation state identification result of the second loom; the operation state recognition result includes a judgment result of whether the second loom is in an abnormal operation state.

For specific limitations of the textile quality inspection device, reference may be made to the limitations of the textile quality inspection method hereinabove, and no further description is given here. The various modules in the textile quality inspection device described above may be implemented in whole or in part by software, hardware, and combinations thereof. The above modules may be embedded in hardware or may be independent of a processor in the computer device, or may be stored in software in a memory in the computer device, so that the processor may call and execute operations corresponding to the above modules.

The textile quality inspection device provided by the above embodiment can be used for executing the textile quality inspection method provided by any embodiment, and has corresponding functions and beneficial effects.

In one embodiment, a computer device is provided, which may be a terminal, and the internal structure of which may be as shown in fig. 7. The computer device includes a processor, a memory, a network interface, a display screen, and an input device connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a method of indoor positioning of an air sensor. The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, can also be keys, a track ball or a touch pad arranged on the shell of the computer equipment, and can also be an external keyboard, a touch pad or a mouse and the like.

It will be appreciated by those skilled in the art that the structure shown in FIG. 7 is merely a block diagram of some of the structures associated with the present inventive arrangements and is not limiting of the computer device to which the present inventive arrangements may be applied, and that a particular computer device may include more or fewer components than shown, or may combine some of the components, or have a different arrangement of components.

In one embodiment, a computer device is provided comprising a memory and a processor, the memory having stored therein a computer program, the processor when executing the computer program performing the steps of:

In one embodiment, a computer readable storage medium is provided having a computer program stored thereon, which when executed by a processor, performs the steps of:

Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in embodiments provided herein may include non-volatile and/or volatile memory. The nonvolatile memory can include Read Only Memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link DRAM (SLDRAM), memory bus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), among others.

The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.

The above examples illustrate only a few embodiments of the application, which are described in detail and are not to be construed as limiting the scope of the application. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the application, which are all within the scope of the application. Accordingly, the scope of protection of the present application is to be determined by the appended claims.

Claims

1. A textile quality inspection system, comprising:

the quality inspection equipment is also used for inputting the textile image to be inspected and the real-time operation data into the quality inspection model in the spinning process to obtain the working condition environment output by the quality inspection model; the working condition environment comprises a flaw identification result of the textile to be inspected and an operation state identification result of the second loom; the operation state identification result comprises a judgment result of whether the second loom is in an abnormal operation state or not; the system further comprises an alarm device; the alarm equipment comprises at least one of a warning lamp and a warning loudspeaker; the image acquisition equipment is also used for acquiring continuous multi-frame images of the textile to be inspected at the same detection position; the textile image to be inspected has a corresponding detection position; the quality inspection device is specifically used for:

Determining the textile image to be inspected, of which the textile flaws are identified, as a first flaw image;

recording a detection position corresponding to the first flaw image as a flaw position;

triggering the alarm device to alarm when detecting that the detection position is recorded as a second flaw image of the flaw position;

after triggering the alarm device to alarm, the quality inspection device is further specifically configured to:

2. The textile quality inspection system of claim 1, wherein the quality inspection device is further configured to determine whether the operating environment is abnormal, and when it is determined that the operating environment is abnormal, perform an operating condition abnormality alarm and/or control the second loom to stop operating.

3. The textile quality inspection system of claim 1, wherein the quality inspection model has a plurality of quality inspection models, and the plurality of quality inspection models respectively correspond to different textile fabrics; the quality inspection equipment is specifically used for identifying a target textile fabric from the textile fabric image to be inspected, determining a quality inspection model corresponding to the target textile fabric, and inputting the textile fabric image to be inspected and the real-time operation data into the quality inspection model corresponding to the target textile fabric.

4. The textile quality inspection system of claim 1, wherein the image acquisition device comprises a first image acquisition device, a second image acquisition device, and a third image acquisition device; the first image acquisition equipment is used for shooting the front surface of the textile to obtain a front surface image of the textile; the second image acquisition equipment is used for shooting the back surface of the textile to obtain a textile back surface image; the third image acquisition equipment is used for shooting the edge of the textile to obtain an edge image of the textile; the textile image to be inspected comprises the textile front image, the textile back image and the textile edge image.

5. The textile quality inspection system of claim 4, further comprising a remote data center;

the quality inspection equipment is further used for recording the flaw position when at least one of the textile fabric front image, the textile fabric back image and the textile fabric edge image is identified, and uploading the flaw position and the real-time operation data to the remote data center;

and the remote data center is used for generating a quality inspection report according to the flaw positions and the real-time operation data.

6. The textile quality inspection system of claim 2, wherein the learning server is further configured to generate a decision model and send the decision model to the quality inspection device;

the quality inspection equipment is further specifically configured to input the working condition environment into the decision model to obtain a target execution decision; the target execution decision comprises working condition abnormality alarming and/or controlling the second loom to stop operating.

7. A method of quality testing a textile based on the textile quality testing system of any one of claims 1-6, comprising:

8. The method of quality inspection of textiles of claim 7, further comprising:

judging whether the working condition environment is abnormal or not;

and when the working condition environment is judged to be abnormal, working condition abnormal alarm is carried out and/or the second loom is controlled to stop operating.

9. The method for inspecting textile quality according to claim 8, wherein the inputting the image of the textile to be inspected and the real-time operation data into the quality inspection model to obtain the working condition environment output by the quality inspection model includes:

Identifying a target textile fabric from the textile fabric image to be inspected;

determining a quality inspection model corresponding to the target textile fabric;

and inputting the textile image to be inspected and the real-time operation data into an inspection model corresponding to the target textile fabric, and obtaining the working condition environment output by the inspection model.

10. The method of claim 7, wherein the image acquisition device comprises a first image acquisition device, a second image acquisition device, and a third image acquisition device; the first image acquisition equipment is used for shooting the front surface of the textile to obtain a front surface image of the textile; the second image acquisition equipment is used for shooting the back surface of the textile to obtain a textile back surface image; the third image acquisition equipment is used for shooting the edge of the textile to obtain an edge image of the textile; the textile image to be inspected comprises the textile front image, the textile back image and the textile edge image, and the method further comprises:

recording a flaw position when a textile flaw is identified in at least one of the textile front image, the textile back image and the textile edge image;

Uploading the flaw location and the real-time operation data to a remote data center; and the remote data center is used for generating a quality inspection report according to the flaw positions and the real-time operation data.

11. The method of quality inspection of textiles of claim 8, further comprising:

receiving a decision model sent by the learning server;

inputting the working condition environment into the decision model to obtain a target execution decision; the target execution decision comprises working condition abnormality alarming and/or controlling the second loom to stop operating.

12. A textile quality inspection device based on the textile quality inspection system of any one of claims 1-6, comprising:

13. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the textile quality inspection method of claim 7 when the computer program is executed.

14. A computer readable storage medium having stored thereon a computer program, characterized in that the computer program when executed by a processor carries out the steps of the method for quality inspection of textiles according to claim 7.