Knitted wool fabric production system based on artificial intelligence monitoring

Abstract

The invention relates to the field of spinning process monitoring, in particular to a knitted wool fabric production system based on artificial intelligence monitoring, which comprises a supervision and evaluation module, a monitoring and evaluation module, a monitoring and evaluation module and a monitoring and evaluation module, determining a production supervision state according to the reference reflection ratio index and the reference deformation ratio index; the execution analysis module is used for determining to carry out focus distribution analysis or focus formation analysis on the production supervision image according to the production supervision state; the distribution analysis module is used for determining whether to adjust a supervision process parameter or not according to a reference edge index according to a focus distribution evaluation coefficient or determining whether to adjust a spectrum composition or not according to a focus region color difference index; and the formation analysis module is used for determining whether to carry out equipment early warning for the image acquisition process according to the stage distribution extension index of each focus distribution area. According to the method, the effectiveness of the knitted fabric image for production optimization decision making is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of textile process monitoring, in particular to a knitted wool fabric production system based on artificial intelligence monitoring. BACKGROUND

[0002] In the traditional production process of knitted wool fabric, it is difficult to ensure comprehensive supervision of existing interference conditions, resulting in poor real-time risk warning of the fabric production process, and the problems existing in the production process cannot be analyzed in a timely and effective manner. By introducing artificial intelligence technology and combining real-time fabric images and other monitoring data, real-time analysis of the fabric production process can be performed to effectively ensure timely and effective analysis of problems existing in the fabric production process. However, there is still a lot of pressure on monitoring data analysis in the actual monitoring and analysis process of fabric production. Therefore, how to optimize the actual monitoring and analysis process based on actual production conditions to ensure effective regulation of the knitted wool fabric production process and further improve the production efficiency and quality of knitted wool fabric.

[0003] Chinese patent application publication No. CN117926496A discloses a knitted intelligent factory production management system and method, which relates to the field of intelligent monitoring. It uses artificial intelligence monitoring technology based on deep learning to identify faults by extracting features from vibration signals of different parts of the automatic flat knitting machine during operation and images of knitted fabrics knitted by the automatic flat knitting machine. However, the above-mentioned scheme has the following defects: it fails to consider the interference of the collection quality of the actual acquired knitted fabric image on the feature extraction process, resulting in low analysis quality of the problems existing in the actual knitted fabric production process, and further failing to ensure the effectiveness of the adjustment decisions made for the knitted fabric production process. SUMMARY

[0004] Therefore, the present application provides a knitted wool fabric production system based on artificial intelligence monitoring to overcome the problem that the prior art fails to consider the interference of the collection quality of the actual acquired knitted fabric image on the feature extraction process, resulting in low analysis quality of the problems existing in the actual knitted fabric production process, and further failing to ensure the effectiveness of the adjustment decisions made for the knitted fabric production process.

[0005] To achieve the above-mentioned purpose, the present application provides a knitted wool fabric production system based on artificial intelligence monitoring, comprising: The monitoring and evaluation module is used to periodically detect the monitoring quality parameters of each monitoring and evaluation period, and determine whether to perform effectiveness analysis on the production monitoring image according to the monitoring quality parameters, and determine the production monitoring state of the monitoring and evaluation period according to the reference reflection ratio index and the reference deformation ratio index; an execution analysis module connected with the supervision evaluation module, configured to determine focus distribution analysis or focus formation analysis for the production supervision image according to the production supervision state; a distribution analysis module connected with the execution analysis module, configured to determine a focus distribution evaluation coefficient according to the focus existing in each of the production supervision images obtained in the target evaluation period, and determine whether to adjust the supervision process parameter according to the reference edge index or adjust the spectral composition according to the focus area color difference index according to the focus distribution evaluation coefficient; a formation analysis module connected with the execution analysis module, configured to determine whether to give a device warning for the image acquisition process according to the stage distribution extension index of each focus distribution area.

[0006] Further, if the monitoring quality parameter of the target evaluation period is less than or equal to the preset monitoring quality parameter, it is determined that the effectiveness analysis is performed for the production supervision image; The monitoring quality parameter is determined according to the recognition evaluation index of each production supervision image obtained in each supervision evaluation period; The target evaluation period is the supervision evaluation period ending at the current time.

[0007] Further, the production supervision state includes a first preset production supervision state and a second preset production supervision state, wherein, The supervision evaluation period in the first preset production supervision state is the supervision evaluation period in which the reference reflection proportion index is greater than the preset reference reflection proportion index or the reference deformation proportion index is greater than the preset reference deformation proportion index; The supervision evaluation period in the second preset production supervision state is the supervision evaluation period in which the reference reflection proportion index is less than or equal to the preset reference reflection proportion index and the reference deformation proportion index is less than or equal to the preset reference deformation proportion index.

[0008] Further, if the target evaluation period is in the first preset production supervision state, it is determined that the distribution analysis module performs focus distribution analysis for the production supervision image.

[0009] Further, the focus distribution evaluation coefficient is determined according to the reference distribution index and the distribution overlap proportion index; The focus is a reflection pixel point and an unknown pixel point existing in the production supervision image.

[0010] Further, if the focus distribution evaluation coefficient is greater than the preset focus distribution evaluation coefficient, it is determined to adjust the output tension index or the light source input angle based on the reference edge index; when the reference edge index is greater than a preset reference edge index, the output tension index is adjusted in a decreasing manner according to the reference distribution coefficient; when the reference edge index is less than or equal to the preset reference edge index, the light source input angle is adjusted in a decreasing manner according to the distribution overlap proportion index; The monitoring process parameters include the output tension index and the light source input angle.

[0011] Further, if the focal point distribution evaluation coefficient is less than or equal to a preset focal point distribution evaluation coefficient, it is determined whether to adjust the spectral composition of the image acquisition light source according to the focal point area color difference index. If the focal point area color difference index is less than a preset area color difference index, the spectral composition of the image acquisition light source is adjusted.

[0012] Further, if the target evaluation period is in a second preset production monitoring state, the focal point formation analysis module determines to perform focal point formation analysis on the production monitoring image.

[0013] Further, the stage distribution extension index is determined according to the existing focal points in each focal point distribution combination in the formation analysis stage. The focal point distribution combination is a distribution analysis combination of existing focal points.

[0014] Further, if the stage distribution extension index of the existing focal point distribution combination is greater than a preset stage distribution extension index, a device warning is given for the image acquisition process.

[0015] To achieve the above-mentioned purposes, the present application has the beneficial effect that the present application periodically detects the monitoring quality parameters of the evaluation period to determine whether the obtained production monitoring image is abnormal. If the monitoring quality parameters of the target evaluation period are small, the confidence of the obtained fabric defects is poor, which interferes with the effectiveness of the optimization decision made for the actual production process. The effectiveness of the production monitoring image obtained by the target evaluation period is analyzed to determine the problems existing in the acquisition process of the production monitoring image, thereby improving the reliability of the obtained production monitoring image. The present application ensures the effectiveness of the obtained knitted fabric image for production optimization decision making.

[0016] Further, the present application determines the production supervision state of the target evaluation period according to the reference reflection proportion index and the reference deformation proportion index, determines whether the acquired production supervision image is in the case of image abnormality in a larger range by the reference reflection proportion index and the reference deformation proportion index, determines how to further analyze the abnormal point position existing in the acquired production supervision image according to the production supervision state of the target evaluation period, ensures that the analysis process of the production supervision image is more in line with the actual situation, and ensures the effectiveness of the analysis result and the adjustment means.

[0017] Further, the present application performs focus distribution analysis on the target evaluation period in the first preset production supervision state, since the production supervision image acquired by the target evaluation period generally exists a larger suspected light reflection area or a larger range of coil that cannot be effectively identified, the factors affecting the acquisition quality of the production supervision image are mainly concentrated in the light reflection situation and the fabric distortion situation with a larger influence range, by analyzing the difference between the actual distribution of the focus existing in the production supervision image acquired in the target evaluation period, the focus distribution analysis process is further distinguished, and the effectiveness of the analysis result and the adjustment means of the image abnormality situation is improved.

[0018] Further, the present application determines whether to adjust the supervision process parameters according to the reference edge index or to adjust the spectrum composition according to the focus area color difference index according to the focus distribution evaluation coefficient, the focus distribution evaluation coefficient represents the distribution concentration of the focus existing in the production supervision image acquired in the target evaluation period and the difference of the distribution range of the focus existing in each production supervision image, and then determines the main influencing factors, the present application improves the defect analysis efficiency of the acquired production supervision image and the effectiveness of the adjustment decision of the image acquisition process.

[0019] Further, the present application performs focus formation analysis on the target evaluation period in the second preset production supervision state, since the production supervision image acquired by the target evaluation period exists a smaller light reflection area and only a smaller range of coil that cannot be effectively identified, the reason for the low quality of the production supervision image at this time is focused on the interference factors with a smaller influence range, by analyzing whether the existence of the focus in a certain time range exists obvious aggravation and diffusion trend, it is determined whether the image acquisition device is affected by long-term dust accumulation, and then it is determined whether to perform equipment warning, the present application improves the defect analysis efficiency of the acquired production supervision image, reduces the useless data analysis situation and the consumption of human resources. BRIEF DESCRIPTION OF DRAWINGS

[0020] Fig. 1A module connection diagram of a knitted wool fabric production system based on artificial intelligence monitoring according to the present application; Fig. 2 A flowchart for determining whether to perform effectiveness analysis on a production supervision image according to a monitoring quality parameter according to the present application; Fig. 3 A flowchart for determining a production supervision state of a supervision evaluation period according to a reference reflection proportion index and a reference deformation proportion index according to the present application; Fig. 4 A flowchart for determining whether to perform focus distribution analysis or focus formation analysis on a production supervision image according to a production supervision state according to the present application. DETAILED DESCRIPTION

[0021] In order to make the objects and advantages of the present application clearer, the present application will be further described below in conjunction with embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application.

[0022] The preferred embodiments of the present application will be described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present application and are not used to limit the protection scope of the present application.

[0023] It should be noted that in the description of the present application, unless explicitly defined and limited, the terms "mounting", "connection" and "linking" should be understood broadly, for example, it can be fixed connection, detachable connection or integral connection; it can be mechanical connection or electrical connection; it can be direct connection or indirect connection through an intermediate medium; it can be internal communication of two elements. Those skilled in the art can understand the specific meaning of the above terms in the present application according to the specific circumstances.

[0024] Please refer to Figs. 1 to 4 The present application provides a knitted wool fabric production system based on artificial intelligence monitoring, which comprises: A supervision evaluation module is used to periodically detect the monitoring quality parameters of each supervision evaluation period, and to determine whether to perform effectiveness analysis on a production supervision image according to the monitoring quality parameters, and to determine the production supervision state of the supervision evaluation period according to a reference reflection proportion index and a reference deformation proportion index; An analysis execution module is connected with the supervision evaluation module, and is used to determine whether to perform focus distribution analysis or focus formation analysis on the production supervision image according to the production supervision state; a distribution analysis module connected with the execution analysis module, configured to determine a focus distribution evaluation coefficient according to the focus existing in each production supervision image obtained in a target evaluation period, and determine whether to adjust the supervision process parameter according to the reference edge index or adjust the spectral composition according to the focus area color difference index according to the focus distribution evaluation coefficient; a formation analysis module connected with the execution analysis module, configured to determine whether to give a device warning for the image acquisition process according to the stage distribution extension index of each focus distribution area.

[0025] In the present application, the production process of the knitted wool fabric is monitored in real time, and the problems existing in the production process are responded in time to ensure the production efficiency and production quality of the actual fabric production process. The currently monitored fabric production process is recorded as a target supervision stage, and the produced wool knitted fabric is recorded as a target production fabric. In the present application, an image acquisition device is arranged in the target supervision stage to acquire images in real time. The acquired images are recorded as production supervision images. In the present application, the specific model, specific position, quantity and interval time length of the image acquisition device are not limited, and the user can adaptively set according to the actual working scene. The production supervision images involved in the analysis process in the present application are all obtained from the image acquisition device arranged at the same position. In the present application, several production supervision records are applied. Any one of the production supervision records records at least one of the monitoring quality parameters, the reference reflection proportion index, the reference deformation proportion index, the distribution coincidence index, the focus area color difference index, the reflection evaluation parameter, the distribution edge index, the reference edge index, the focus distribution evaluation coefficient and the stage distribution extension index in the production supervision process of the knitted fabric production process. Each production supervision record corresponds to a qualified mark. The qualified mark records whether the effectiveness of the risk assessment made on the production process corresponding to the target supervision stage meets the user's demand. It can be understood that the user can determine whether the effectiveness of the risk assessment made on the production process corresponding to the target supervision stage meets the demand according to the self-set index. The self-set index includes but is not limited to: effective production index, effective production index = length of fabric produced in the target supervision stage but not meeting the quality requirements / length of fabric produced in the target supervision stage.

[0026] Specifically, if the monitoring quality parameter of the target evaluation period is less than or equal to the preset monitoring quality parameter, it is determined that the effectiveness analysis is performed on the production supervision image. The monitoring quality parameter is determined according to the recognition evaluation index of each production supervision image obtained in each supervision evaluation period. The target evaluation period is the supervision evaluation period with the current time as the end time.

[0027] Among them, a cycle of supervision evaluation period is applied in the application, the length of the supervision evaluation period can be determined by the user, the higher the user's requirement for the effectiveness of the risk assessment made by the user for the production process corresponding to the target supervision stage, the shorter the length of the supervision evaluation period, a supervision evaluation period of 10 minutes is provided, at the end of each supervision evaluation period, the monitoring quality parameter of the production supervision image obtained in the supervision evaluation period is detected; For a single supervision evaluation period, the monitoring quality parameter = (the average value of the identification evaluation index of each production supervision image obtained in the supervision evaluation period / the maximum value of the identification evaluation index of each production supervision image obtained in the supervision evaluation period) / (the standard deviation between the identification evaluation index of each production supervision image obtained in the supervision evaluation period / the average value of the identification evaluation index of each production supervision image obtained in the supervision evaluation period), if the standard deviation between the identification evaluation index of each production supervision image obtained in the supervision evaluation period is 0, the monitoring quality parameter = the average value of the identification evaluation index of each production supervision image obtained in the supervision evaluation period / the maximum value of the identification evaluation index of each production supervision image obtained in the supervision evaluation period, for a single production supervision image, the identification evaluation index = the number of complete loops identified for the production supervision image / the number of complete loops existing in the area of the knitted fabric corresponding to the collection range of the image collection device corresponding to the production supervision image, the pixel points existing in the image area of the production supervision image which do not identify complete loops are recorded as unknown pixel points, how to identify loops for the obtained production supervision image is a content mastered by those skilled in the art, which is not described here, the number of complete loops existing in the area of the knitted fabric corresponding to the collection range can be determined according to the actual area of the area of the knitted fabric corresponding to the collection range and the preparation requirements of the actually prepared knitted fabric, which is easily understood by those skilled in the art and is not described here; If the monitoring quality parameter of the target evaluation period is less than or equal to the preset monitoring quality parameter, it indicates that the production supervision images obtained in the target evaluation period are interfered to a certain extent and are disturbed as a whole, the value of the preset supervision quality parameter can be determined by the user according to the actual working scene, for example, the user can set it according to the production supervision record, the higher the user's requirement for the effectiveness of the risk assessment made by the user for the production process corresponding to the target supervision stage, the greater the value of the preset monitoring quality parameter, a method for determining the value of the preset monitoring quality parameter is provided, the maximum value of the monitoring quality parameter of the supervision evaluation period corresponding to the effectiveness analysis of the production supervision image in the production supervision record which meets the user's requirement for the effectiveness of the risk assessment made by the user for the production process corresponding to the target supervision stage is recorded as the preset monitoring quality parameter.

[0028] Specifically, the production supervision state includes a first preset production supervision state and a second preset production supervision state, wherein, The supervision evaluation period in the first preset production supervision state is a supervision evaluation period in which the reference reflection proportion index is greater than a preset reference reflection proportion index or the reference deformation proportion index is greater than a preset reference deformation proportion index; The supervision evaluation period in the second preset production supervision state is a supervision evaluation period in which the reference reflection proportion index is less than or equal to the preset reference reflection proportion index and the reference deformation proportion index is less than or equal to the preset reference deformation proportion index.

[0029] If the monitoring quality parameter of the target evaluation period is less than or equal to the preset monitoring quality parameter, the effectiveness of the production supervision image obtained in the target evaluation period is analyzed, and the production supervision state of the target evaluation period is determined according to the reference reflection proportion index and the reference deformation proportion index; For a single supervision evaluation period, the reference reflection proportion index is the average value of the reflection area proportion indexes of each production supervision image obtained in the supervision evaluation period, and the reference deformation proportion index = 1 / the average value of the identification evaluation indexes of each production supervision image obtained in the supervision evaluation period. For a single production supervision image, the reflection area proportion index = the number of reflection pixel points existing in the production supervision image / the number of pixel points existing in the production supervision image, and the reflection pixel point is a pixel point in the production supervision image whose reflection evaluation parameter is greater than a preset reflection evaluation parameter. For any pixel point existing in the production supervision image, the reflection evaluation parameter is the absolute value of the difference between the brightness value of the pixel point and the average value of the brightness values of the pixel points in the evaluation range of the pixel point. The evaluation range is a circular area with the pixel point as the center. The number of pixel points included in each evaluation range is consistent. The value of the number of pixel points included in the evaluation range can be determined by the user according to the actual working scene. The higher the user's requirement for the effectiveness of the risk assessment of the production process corresponding to the target supervision stage, the greater the value of the number of pixel points included in the evaluation range. A value of the number of pixel points included in the evaluation range is provided, which is 150. The value of the preset reflection evaluation parameter can be determined by the user according to the actual working scene. For example, the user can set it according to the production supervision record. The higher the user's requirement for the effectiveness of the risk assessment of the production process corresponding to the target supervision stage, the smaller the value of the preset reflection evaluation parameter. A method for determining the value of the preset reflection evaluation parameter is provided. The average value of the reflection evaluation parameters in the production supervision record that meets the user's requirement for the effectiveness of the risk assessment of the production process corresponding to the target supervision stage is taken as the preset reflection evaluation parameter. The user can determine the values of the preset reference reflection proportion index and the preset reference deformation proportion index according to the actual working scene. For example, the user can set the values according to the production supervision record. The higher the user's requirement for the effectiveness of the risk assessment made by the target supervision stage on the production process, the smaller the value of the preset reference reflection proportion index, and the smaller the value of the preset reference deformation proportion index. A method for determining the value of the preset reference reflection proportion index is provided. The maximum value of the reference reflection proportion index of the supervision evaluation period in the second preset production supervision state in the production supervision record that meets the user's requirement for the effectiveness of the risk assessment made by the target supervision stage on the production process is recorded as the preset reference reflection proportion index. A method for determining the value of the preset reference deformation proportion index is provided. The maximum value of the reference deformation proportion index of the supervision evaluation period in the second preset production supervision state in the production supervision record that meets the user's requirement for the effectiveness of the risk assessment made by the target supervision stage on the production process is recorded as the preset reference deformation proportion index.

[0030] Specifically, if the target evaluation period is in the first preset production supervision state, the focus distribution analysis module determines the focus distribution analysis of the production supervision image.

[0031] If the target evaluation period is in the first preset production supervision state, it indicates that the production supervision image obtained in the target evaluation period generally has a large suspected light reflection area or a large range of coil that cannot be effectively identified. At this time, the reason for the low quality of the obtained production supervision image is mainly focused on the reflection of light and the coil distortion caused by the fabric pulling in the actual production process. At this time, the obtained production supervision image cannot effectively reflect the preparation effect of the obtained wool knitted fabric. The interference of different regions is determined by focus distribution analysis to provide a reference for the preparation evaluation process of the wool knitted fabric.

[0032] Specifically, the focus distribution evaluation coefficient is determined according to the reference distribution index and the distribution coincidence proportion index. The focus is a reflection pixel point and an unknown pixel point existing in the production supervision image.

[0033] The focus distribution evaluation coefficient for the target evaluation period is the sum of the reference distribution index and the distribution coincidence proportion index. The reference distribution index is the average value of the focus distribution density index of each production supervision image obtained in the target evaluation period. For a single production supervision image, the focus distribution density index is the average value of the number of focus points in the evaluation range of each focus point existing in the production supervision image divided by the number of focus points existing in the production supervision image. The distribution coincidence proportion index is the number of distribution coincidence combinations divided by the number of distribution analysis combinations obtained by division. Each production supervision image is divided to obtain a plurality of rectangular regions with the same area, which are referred to as distribution analysis regions. The division process of each production supervision image is consistent. The distribution analysis regions with consistent division coordinates corresponding to each production supervision image are referred to as a distribution analysis combination. The division coordinates are the row number and column number corresponding to the distribution analysis region obtained after the production supervision image is divided. When determining the division coordinates of each distribution analysis region, the production supervision image does not undergo rotation processing. For a single distribution analysis combination, if the distribution coincidence index of the distribution analysis combination is less than a preset distribution coincidence index, the distribution analysis combination is determined to be a distribution coincidence combination. The distribution coincidence index is the maximum value of the difference between the number of focus points existing in each distribution analysis region in the distribution analysis combination. The value of the preset distribution coincidence index can be determined by the user according to the actual working scenario. For example, the user can set it according to the production supervision record. The higher the user's requirement for the effectiveness of the risk assessment made by the target supervision stage for the production process, the smaller the value of the preset distribution coincidence index. A method for determining the value of the preset distribution coincidence index is provided. The average value of the distribution coincidence index of the distribution coincidence combination in the production supervision record that meets the user's requirement for the effectiveness of the risk assessment made by the target supervision stage for the production process is referred to as the preset distribution coincidence index.

[0034] Specifically, if the focus distribution evaluation coefficient is greater than a preset focus distribution evaluation coefficient, the reference edge index is used to determine the adjustment of the output tension index or the light source input angle; When the reference edge index is greater than a preset reference edge index, the reference distribution coefficient is used to reduce the output tension index; When the reference edge index is less than or equal to the preset reference edge index, the distribution coincidence proportion index is used to reduce the light source input angle; The supervision process parameters include the output tension index and the light source input angle.

[0035] Among them, if the focus distribution evaluation coefficient is greater than the preset focus distribution evaluation coefficient, it indicates that the distribution of the focus (i.e., the pixel points that interfere with coil recognition) in the production supervision images obtained during the target evaluation period is generally concentrated and the distribution range of the focus in each production supervision image is relatively consistent, which further indicates that the production supervision images obtained during the target evaluation period are affected by the same factors. By analyzing the edge distribution of the focus distribution combination, the main influencing factors are analyzed; The value of the preset focus distribution evaluation coefficient can be determined by the user according to the actual working scenario. For example, the user can set it according to the production supervision record. A method for determining the value of the preset focus distribution evaluation coefficient is provided. The production supervision record for adjusting the output tension index or the light source input angle based on the reference edge index of the distribution overlap combination is recorded as the adjustment reference record. The minimum value of the briefing distribution evaluation coefficient in the adjustment reference record that meets the validity requirements of the risk assessment made by the user for the production process corresponding to the target supervision stage is recorded as the preset briefing distribution evaluation coefficient. The reference edge index = the number of edge distribution combinations corresponding to the target evaluation period / the number of focus distribution combinations corresponding to the target evaluation period. For a single focus distribution combination, if the distribution edge index of the focus distribution combination is greater than the preset distribution edge index, the focus distribution combination is determined to be an edge distribution combination. The distribution edge index = 1 / the shortest distance between the image area corresponding to the focus distribution combination and the position of the target production fabric and the edge of the target production fabric. The edge of the target production fabric is the straight line formed by the connection between the target production fabric and the pulling mechanism during the production process of the target production fabric. The value of the preset distribution edge index can be determined by the user according to the actual working scenario. For example, the user can set it according to the production supervision record. A method for determining the value of the preset distribution edge index is provided. The average value of the distribution edge index of the edge distribution combination in the production supervision record that meets the effectiveness requirements of the risk assessment made by the user for the production process corresponding to the target supervision stage is recorded as the preset distribution edge index. When the reference edge index is greater than the preset reference edge index, it indicates that the focal point distribution combination has a significant edge distribution trend, and the problem existing in the image acquisition process mainly focuses on the fabric distortion caused by the excessive force between the pulling mechanism and the target production fabric. According to the reference distribution coefficient, the output tension index is adjusted to be reduced, the output tension index is the value of the tensile force formed between the target production fabric and the pulling mechanism, and the reduction value of the output tension index is in positive correlation with the reference edge index. When the reference edge index is less than or equal to the preset reference edge index, it indicates that the focal point distribution combination does not have a significant edge distribution trend, and at this time the problem existing in the image acquisition process mainly focuses on the excessive reflection caused by the incident angle of the light source. According to the distribution overlap proportion index, the light source input angle is adjusted to be reduced, the light source input angle is the acute angle value formed by the incident light of the image acquisition light source and the target production fabric, and the reduction value of the light source input angle is in positive correlation with the distribution overlap proportion index. The image acquisition light source is a light source manually set for image acquisition of the target production fabric. How to adjust the light source input angle of the image acquisition light source and the output tension index is a content mastered by those skilled in the art, and will not be described here. The value of the preset reference edge index can be determined by the user according to the actual working scene. For example, the user can set it according to the production supervision record. A method for providing a value of the preset reference edge index is provided. The production supervision record in which the output tension index is adjusted to be reduced according to the reference distribution coefficient is recorded as the process optimization record. The minimum value of the reference edge index in the process optimization record that meets the effectiveness requirement of the user's risk assessment of the production process corresponding to the target supervision stage is recorded as the preset reference edge index.

[0036] Specifically, if the focal point distribution evaluation coefficient is less than or equal to the preset focal point distribution evaluation coefficient, it is determined whether to adjust the spectral composition of the image acquisition light source according to the focal point area color difference index. If the focal point area color difference index is less than the preset area color difference index, the spectral composition of the image acquisition light source is adjusted.

[0037] If the focal point distribution evaluation coefficient is less than or equal to a preset focal point distribution evaluation coefficient, it indicates that the distribution of the focal points (i.e., the pixel points that interfere with the identification of the coil) existing in the production supervision images obtained in the target evaluation period is generally more dispersed, and there is a large difference between the distribution ranges of the focal points existing in each production supervision image. At this time, the impact on the acquisition process of the production supervision image is not the stretching and distortion of the specific direction of the pile reflection and the characteristic region of the production process, but because the color of the knitted fabric produced at this time is relatively rich, the spectral composition of the image acquisition light source cannot effectively ensure the image acquisition effect of all regions, and then by analyzing the color of the region where the focal points exist, how to adjust the spectral composition of the image acquisition light source is determined. The evaluation range of the focal points existing in the production supervision images obtained in the target evaluation period is recorded as the focal point region, and the region color difference index = standard deviation between the average values of the gray values of the pixel points in each focal point region / 255. If the region color difference index is small, it indicates that the color depth of the fabric corresponding to the focal point region is consistent, and it indicates that the color depth of the fabric in some regions is affected by the current image acquisition light source. At this time, the spectral composition of the image acquisition light source needs to be adjusted. The image acquisition light source is a light source set to ensure the acquisition effect of the production supervision image. The spectral composition is the intensity ratio of the red, green, blue, and infrared four channels included in the image acquisition light source. How to adjust the spectral composition according to the gray values of the pixel points in each focal point region is easily understood by those skilled in the art. For example, if the average value of the gray values of the pixel points in each focal point region is less than or equal to a first preset evaluation gray value, it indicates that the fabric part corresponding to the focal point region is relatively deep. At this time, the intensity ratio of the red and infrared light channels needs to be increased. The increase in the intensity ratio is negatively correlated with the average value of the gray values of the pixel points in each focal point region. If the average value of the gray values of the pixel points in each focal point region is greater than a second preset evaluation gray value, it indicates that the fabric part corresponding to the focal point region is relatively shallow. At this time, the intensity ratio of the blue light channel needs to be increased. The increase in the intensity ratio is positively correlated with the average value of the gray values of the pixel points in each focal point region. This will not be elaborated or specifically limited. How to adjust the intensity ratio of the red, green, blue, and infrared four channels included in the image acquisition light source is a content that those skilled in the art have mastered. The first preset evaluation gray value and the second preset evaluation gray value are determined by the user according to the actual working scene, and a first preset evaluation gray value is provided, and the first preset evaluation gray value is 110; a second preset evaluation gray value is provided, and the second preset evaluation gray value is 150; the preset region color difference index is determined by the user according to the actual working scene, for example, the user can set it according to the production supervision record, a method for determining the preset region color difference index is provided, the production supervision record for adjusting the spectral composition of the image acquisition light source is recorded as the light source optimization record, and the average value of the focal region color difference index of the light source optimization record meeting the effectiveness requirement of the user's risk assessment on the production process corresponding to the target supervision stage is recorded as the preset region color difference index.

[0038] Specifically, if the target evaluation period is in the second preset production supervision state, the focal formation analysis module determines to perform focal formation analysis on the production supervision image.

[0039] If the target evaluation period is in the second preset production supervision state, it indicates that the production supervision image obtained in the target evaluation period has a small area of light reflection and only a small range of coil cannot be effectively identified, at this time, the reason for the low quality of the obtained production supervision image has little to do with the relationship between the actual production process and the fluff reflection on the fabric and the fabric pulling, and more focuses on the problems existing in the actual image acquisition process. At this time, the obtained production supervision image cannot effectively reflect the preparation effect of the obtained wool knitted fabric, and the time sequence of the formation of the focus in the similar range between the obtained growth images is determined by the focal growth analysis to determine the interference condition of different regions, thereby providing a reference for the preparation evaluation process of the wool knitted fabric.

[0040] Specifically, the stage distribution extension index is determined according to the focus existing in each focal distribution combination in the formation analysis stage. The focal distribution combination is a distribution analysis combination of the focus.

[0041] Specifically, if the stage distribution extension index of the focal distribution combination is greater than the preset stage distribution extension index, a device warning is given for the image acquisition process.

[0042] The stage distribution extension index of the focus distribution combination is the sum of a region growth amplitude index and a region extension proportion index, the region growth amplitude index = (focus distribution index of the target evaluation period - minimum value of the focus distribution index of each regulatory evaluation period in the formation analysis stage except the target evaluation period) / minimum value of the focus distribution index of each regulatory evaluation period in the formation analysis stage, the focus distribution index of a single regulatory evaluation period is the average value of the number of focuses determined for each distribution analysis region in the focus distribution combination in the regulatory evaluation period, and the region extension proportion index = (number of focus distribution combinations adjacent to the focus distribution combination in the target evaluation period - number of focus distribution combinations adjacent to the focus distribution combination in the first regulatory evaluation period in the formation analysis stage) / number of focus distribution combinations adjacent to the focus distribution combination in the first regulatory evaluation period in the formation analysis stage, the end time of the formation analysis stage is the end time of the target evaluation period, the length of the formation analysis stage, and the user can determine the length of the formation analysis stage according to the actual working scene, the higher the requirement of the user for the effectiveness of the risk assessment made by the user for the production process corresponding to the target regulatory stage, the greater the value of the length of the formation analysis stage, and a value of the length of the formation analysis stage is provided, and the length of the formation analysis stage is 15 times the length of the regulatory evaluation period. If the stage distribution extension index of the focus distribution combination is greater than the preset stage distribution extension index, it indicates that the focus distribution combination has a more obvious aggravation and diffusion trend in the formation analysis stage, indicating that the equipment lens of the image acquisition device has a dust accumulation feature, and then the equipment is warned for the image acquisition process, and the image acquisition device is processed in time, the value of the preset stage distribution extension index is determined by the user according to the actual working scene, for example, the user can set it according to the production supervision record, the higher the requirement of the user for the effectiveness of the risk assessment made by the user for the production process corresponding to the target regulatory stage, the smaller the value of the preset stage distribution extension index, a preset stage distribution extension index value method is provided, the production supervision record of the equipment warning for the image acquisition process is recorded as the equipment reference record, and the minimum value of the stage distribution extension index in the equipment reference record meeting the requirement of the user for the effectiveness of the risk assessment made by the user for the production process corresponding to the target regulatory stage is recorded as the preset stage distribution extension index.

[0043] So far, the technical scheme of the present application has been described in combination with the preferred embodiments shown in the drawings, but those skilled in the art can easily understand that the protection scope of the present application is obviously not limited to these specific embodiments. Those skilled in the art can make equivalent changes or replacements to the related technical features without departing from the principles of the present application, and the technical schemes after the changes or replacements will fall within the protection scope of the present application.

[0044] The above merely illustrates the preferred embodiments of the present application, and is not used to limit the present application; for those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A knitted wool fabric production system based on artificial intelligence monitoring, characterized by, The method comprises the following steps: a supervision evaluation module is used to detect the monitoring quality parameters of each supervision evaluation period periodically, and determine whether to perform effectiveness analysis on the production supervision image according to the monitoring quality parameters, and determine the production supervision state of the supervision evaluation period according to the reference reflection proportion index and the reference deformation proportion index; an execution analysis module connected with the supervision evaluation module is used to determine whether to perform focus distribution analysis or focus formation analysis on the production supervision image according to the production supervision state; a distribution analysis module connected with the execution analysis module is used to determine the focus distribution evaluation coefficient according to the focus existing in each production supervision image obtained in the target evaluation period, and determine whether to adjust the supervision process parameter according to the reference edge index or adjust the spectral composition according to the focus area color difference index based on the focus distribution evaluation coefficient; a formation analysis module connected with the execution analysis module is used to determine whether to perform equipment early warning on the image acquisition process according to the stage distribution extension index of each focus distribution area.

2. The artificial intelligence monitoring based knitted wool fabric production system as claimed in claim 1 wherein, If the monitoring quality parameter of the target evaluation period is less than or equal to the preset monitoring quality parameter, it is determined to perform effectiveness analysis on the production supervision image; The monitoring quality parameter is determined according to the identification evaluation index of each production supervision image obtained in each supervision evaluation period; The target evaluation period is the supervision evaluation period ending at the current time.

3. The artificial intelligence monitoring based knitted wool fabric production system as claimed in claim 1 wherein, The production supervision state comprises a first preset production supervision state and a second preset production supervision state, wherein, The supervision evaluation period in the first preset production supervision state is the supervision evaluation period in which the reference reflection proportion index is greater than the preset reference reflection proportion index or the reference deformation proportion index is greater than the preset reference deformation proportion index; The supervision evaluation period in the second preset production supervision state is the supervision evaluation period in which the reference reflection proportion index is less than or equal to the preset reference reflection proportion index and the reference deformation proportion index is less than or equal to the preset reference deformation proportion index.

4. The knitted wool fabric production system based on artificial intelligence monitoring according to claim 3 is characterized in that: If the target evaluation period is in the first preset production supervision state, it is determined that the distribution analysis module performs focus distribution analysis on the production supervision image.

5. The knitted wool fabric production system based on artificial intelligence monitoring according to claim 4 is characterized in that: The focus distribution evaluation coefficient is determined according to the reference distribution index and the distribution overlap proportion index; The focus is a reflection pixel point and an unknown pixel point existing in the production supervision image.

6. The artificial intelligence monitoring based knitted woolen fabric production system as claimed in claim 5 wherein, If the focus distribution evaluation coefficient is greater than the preset focus distribution evaluation coefficient, it is determined to adjust the output tension index or the light source input angle based on the reference edge index; When the reference edge index is greater than the preset reference edge index, the output tension index is adjusted by decreasing according to the reference distribution coefficient; When the reference edge index is less than or equal to the preset reference edge index, the light source input angle is adjusted by decreasing according to the distribution overlap proportion index; The supervision process parameter comprises an output tension index and a light source input angle.

7. The artificial intelligence monitoring based knitted woolen fabric production system as claimed in claim 6 wherein, If the focus distribution evaluation coefficient is less than or equal to the preset focus distribution evaluation coefficient, it is determined whether to adjust the spectral composition of the image acquisition light source according to the focus area color difference index. If the focal region color difference index is less than the preset region color difference index, adjustment is made to the spectral composition of the image acquisition light source.

8. The artificial intelligence monitoring based knitted woolen fabric production system as claimed in claim 3, wherein, If the target evaluation period is in a second preset production supervision state, it is determined that the focal formation analysis module performs focal formation analysis on the production supervision image.

9. The artificial intelligence monitoring based knitted woolen fabric production system as claimed in claim 7, wherein, The stage distribution extension index is determined according to the existing focal points in each focal distribution combination in the formation analysis stage; The focal distribution combination is a distribution analysis combination of the existing focal points.

10. The artificial intelligence monitoring based knitted wool fabric production system as claimed in claim 9, wherein, If the stage distribution extension index of the existing focal distribution combination is greater than a preset stage distribution extension index, a device warning is given to the image acquisition process.

Images