Method for processing wound yarn package data, processing device, electronic device, storage medium, and program

Abstract

A method, apparatus, electronic device, storage medium and program for processing wound yarn package data are provided. [Solution] The method includes: when it is determined that the transported wound yarn package has entered the detection area, performing defect detection on the wound yarn package located within the detection area and obtaining a target detection result for the wound yarn package, the target detection result being used to indicate the degree of defect in the wound yarn package; and when it is determined that the target detection result satisfies the predetermined defect requirements, obtaining a target grade for the wound yarn package based on the target detection result of the wound yarn package and a predetermined grade of the wound yarn package.

Description

[Technical field]

[0001] The present disclosure relates to the field of data processing, and more particularly to a method, apparatus, electronic device, storage medium and program for processing wound yarn package data. [Background technology]

[0002] In the chemical fiber industry, operators usually need to perform defect detection on a single ball of yarn package, and then re-evaluate the grade of the yarn package based on the detection results. This detection method relies heavily on human experience and is therefore less efficient, which affects the production and management efficiency of yarn packages. Summary of the Invention [Problem to be solved by the invention]

[0003] The present disclosure provides a method, apparatus, electronic device, storage medium and program for processing wound yarn package data to solve or alleviate one or more technical problems in the prior art. [Means for solving the problem]

[0004] In a first aspect, the present disclosure provides a method for processing wound yarn package data, the method comprising: when it is determined that the transported wound yarn package has entered a detection area, a defect detection is performed on the wound yarn package located within the detection area to obtain a target detection result for the wound yarn package, the target detection result being used to indicate the degree of a defect in the wound yarn package; If it is determined that the target detection result meets the preset defect requirement, obtaining a target grade for the wound yarn package based on the target detection result for the wound yarn package and a preset grade for the wound yarn package.

[0005] In a second aspect, the present disclosure provides an apparatus for processing wound yarn package data, the apparatus comprising: a defect detection unit for performing defect detection on the wound yarn package located within a detection area when it is determined that the transported wound yarn package has entered the detection area, and obtaining a target detection result for the wound yarn package, the target detection result being used to indicate the degree of a defect in the wound yarn package; and a grade determination unit for obtaining a target grade of the wound yarn package based on the target detection result of the wound yarn package and a preset grade of the wound yarn package when it is determined that the target detection result meets the preset defect requirement.

[0006] In a third aspect, the present disclosure provides an electronic device, the device comprising: At least one processor; a memory in communication with the at least one processor; The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, cause the implementation of any one of the methods in the embodiments of the present disclosure.

[0007] In a fourth aspect, there is provided a non-transitory computer-readable storage medium having stored thereon computer instructions for causing a computer to perform any one of the methods in the embodiments of the present disclosure.

[0008] In a fifth aspect, a program is provided, which, when executed by a processor, implements any one of the methods in the embodiments of the present disclosure. Effect of the Invention

[0009] The beneficial effects of the solution provided by the present disclosure include at least the following:

[0010] The beneficial effects of the technical solution provided by the present disclosure at least include the following: the technical solution of the present disclosure first performs automatic defect detection on the appearance of the yarn package, and then accurately evaluates the actual grade of the yarn package based on the actual defect situation of the yarn package (e.g., the target detection result), and then makes a quick adjustment to the grade of the spindle, thus realizing a fully automated processing process from defect detection of the yarn package to grade evaluation of the yarn package, and compared with the traditional manual method, the technical solution of the present disclosure does not need to rely on human experience, and can efficiently detect defects in the yarn package, and can make automatic adjustments to the grade of the yarn package based on the detection result, thus saving a large amount of labor and time costs, and further improving the management efficiency of the yarn package. [Brief description of the drawings]

[0011] [Figure 1] 1 is a schematic flowchart of a method for processing yarn package data according to an embodiment of the present disclosure; [Diagram 2] 2 is a second schematic flowchart of a method for processing winding package data according to an embodiment of the present disclosure. [Diagram 3] FIG. 2 is a schematic diagram illustrating a scenario of image collection of a wound yarn package according to one embodiment of the present disclosure. [Figure 4] FIG. 2 is a schematic diagram illustrating a configuration of an example of a second network layer according to an embodiment of the present disclosure. [Diagram 5] 1 is a schematic diagram illustrating an example of a configuration of a sixth sub-network layer in a second network layer according to an embodiment of the present disclosure. FIG. [Figure 6] FIG. 2 is a schematic diagram illustrating a configuration of a feature weight module according to an embodiment of the present disclosure. [Figure 7] FIG. 1 is a schematic diagram showing the configuration of a multi-view attention module according to one embodiment of the present disclosure. [Figure 8] 11 is a schematic flowchart of a method for processing winding package data according to an embodiment of the present disclosure, part three. [Figure 9]1 is a schematic diagram showing a configuration of a processing device for wound yarn package data according to an embodiment of the present disclosure. [Figure 10] FIG. 2 is a block diagram of an electronic device for implementing a method for processing yarn package data according to an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] It should be understood that the contents described herein are not intended to describe key points or key features of the embodiments of the present disclosure, nor are they used to limit the scope of the present disclosure. Other features of the present disclosure are facilitated by the following specification.

[0013] In the accompanying drawings, unless otherwise specified, the same reference numerals throughout the several accompanying drawings indicate the same or similar components or elements. The accompanying drawings are not necessarily drawn to scale. It should be understood that the drawings depict only some examples provided by the present disclosure and should not be considered as limiting the scope of the present disclosure.

[0014] The present disclosure will now be described in more detail with reference to the accompanying drawings, in which like reference numbers represent the same or similar elements, and in which various aspects of the embodiments are illustrated, and which, unless otherwise noted, are not necessarily drawn to scale.

[0015] Furthermore, in order to better explain the present disclosure, many specific details are described in the following specific embodiments. Those skilled in the art should understand that the present disclosure can be similarly implemented without some details. In some embodiments, methods, means, components, circuits, etc. that are well known to those skilled in the art are not described in detail so that the gist of the present disclosure is clear.

[0016] In the chemical fiber industry, the smallest product unit is a single ball of yarn winding package, and based on the winding ball data stream constructed by a single ball of yarn winding package, transparency of the entire winding package production process for each ball has been achieved.

[0017] The term "wound ball data stream" refers to assigning a code (e.g., two-dimensional code or barcode) to each wound yarn package, and then linking the stages of automatic doffing, wound ball weighing, appearance inspection, automatic packaging line, and wound ball transportation through a physical conveying device, so that the relevant data of each wound yarn package can be immediately and completely collected and transmitted at each stage. However, in the process of appearance inspection of a wound yarn package, it is usually necessary to manually detect defects in the wound yarn package and manually evaluate the actual grade of the wound yarn package based on the detection result, so obviously this detection method is low in efficiency and highly dependent on human experience. Therefore, an automatic processing process is needed to efficiently recognize defects in the wound yarn package and further re-evaluate the grade of the wound yarn package.

[0018] Therefore, the present disclosure provides a method for processing yarn package data to solve the above-mentioned problems.

[0019] 1 is a simplified flowchart of a method for processing yarn package data according to one embodiment of the present disclosure, the method being selectively applied to electronic devices such as personal computers, servers, server clusters, and the like.

[0020] In addition, the method includes at least part of the following contents: As shown in FIG.

[0021] In step S101, when it is determined that the conveyed wound yarn package has entered the detection area, defect detection is performed on the wound yarn package located within the detection area to obtain a target detection result for the wound yarn package.

[0022] Here, the target detection result is used to indicate the degree of defect of the wound yarn package.

[0023] Further, in one embodiment, the main type of yarn in the wound yarn package can include one or more of partially oriented yarns (POY), fully drawn yarns (FDY), drawn textured yarns (DTY) (or low modulus filaments), etc. For example, the types of yarn in the wound yarn package can specifically include polyester partially oriented yarns, polyester fully drawn yarns, polyester drawn yarns, polyester drawn textured yarns, etc.

[0024] Furthermore, in one specific example, defect detection for a wound yarn package can specifically refer to detecting defects such as winding breakage, fuzzing, oil stains, tangles, nips, paper tube damage, thread hanging, and wrong paper tube color.

[0025] In step S102, if it is determined that the target detection result meets the preset defect requirements, a target grade of the wound yarn package is obtained based on the target detection result of the wound yarn package and the preset grade of the wound yarn package.

[0026] In addition, the preset grade of the yarn winding package may be the highest preset grade or may be another grade, for example, in a scenario such as downgrading a yarn winding package, the preset grade of the yarn winding package is set to the highest grade, and in that case, it can be determined whether the preset grade of the yarn winding package needs to be downgraded based on the target detection result of the yarn winding package. In practical applications, the preset grade of the yarn winding package can be set according to the needs of the practical scenario, and the present disclosure does not specifically limit this.

[0027] In this way, the present disclosure can use the target detection result of the yarn package to determine the target grade of the yarn package. Specifically, the present disclosure first performs automatic defect detection on the appearance of the yarn package, and then accurately evaluates the actual grade of the yarn package based on the actual defect situation of the yarn package (e.g., the target detection result), and then quickly adjusts the grade of the yarn package. In this way, a fully automatic processing process from defect detection of the yarn package to grade evaluation of the yarn package is realized. Compared with the traditional manual method, the method of the present disclosure can efficiently detect defects in the yarn package without relying on human experience, and can automatically adjust the grade of the yarn package based on the detection result, thereby saving a large amount of labor and time costs, and further improving the management efficiency of the yarn package.

[0028] In addition, the grade of the yarn package can be automatically adjusted, which can effectively improve the data quality of the wound ball data, better realize the quality control and feedback for the yarn package in the production process, and effectively avoid possible losses for users.

[0029] 2 is a schematic flow chart of a method for processing yarn package data according to an embodiment of the present disclosure. The method can be selectively applied to electronic devices such as personal computers, servers, server clusters, etc. The relevant content of the method shown in FIG. 1 above can also be applied to this embodiment, and the relevant content will not be repeated in this embodiment.

[0030] In addition, the method includes at least part of the following contents: As shown in FIG.

[0031] In step S201, when it is determined that the conveyed wound yarn package has entered the detection area, N images of the wound yarn package are acquired.

[0032] Here, an image in the N images includes at least a partial area of ​​the wound yarn package; for example, an image may include a local area such as the top surface, bottom surface, or annular surface (i.e., side surface) of the wound yarn package.

[0033] Moreover, in one example, N is 3 or greater.

[0034] Further, in one embodiment, images of the yarn package can be collected as follows. Specifically, step S201 can include, when receiving response information to the radio frequency identification corresponding to the yarn package (in this case, it is considered that the yarn package has entered the detection area), using N image collection components to collect images of different angles of the yarn package to obtain N images. For example, as shown in FIG. 3, three image collection components are used to collect images of the yarn package in all directions, that is, image collection component 1 collects images of the top surface of the yarn package, image collection component 2 collects images of the left part of the bottom surface of the yarn package and the left side surface of the yarn package, and image collection component 3 collects images of the right part of the bottom surface of the yarn package and the right side surface of the yarn package. In this way, it is helpful to perform comprehensive defect detection on the yarn package, and lays a foundation for improving the accuracy of defect detection of subsequent yarn packages.

[0035] Here, in one example, the image collection method of the wound thread package may be a method in which, when it is determined that the wound thread package has been transported and entered a detection area, the image collection component generates a control command to instruct a mechanical gripper to grasp and rotate the wound thread package in order to collect images of the wound thread package in all directions, or it may be another image collection method, and the present disclosure is not particularly limited to the image collection method of the wound thread package.

[0036] In step S202, the N images are input into a target detection model to obtain a first detection result corresponding to each image in the N images.

[0037] Here, the target detection model is used to recognize whether there is a defect in a local area of ​​the wound yarn package and obtain a first detection result, the first detection result including at least one of the number of defects, the defect locations, and the defect types.

[0038] In step S203, a target detection result of the wound yarn package is obtained based on a first detection result corresponding to each image in the N images.

[0039] Here, the target detection result is used to indicate the degree of defect in the wound yarn package.

[0040] In step S204, it is determined whether the target detection result satisfies a preset defect requirement, and if so, the process proceeds to step S205, and if not, the process proceeds to step S206.

[0041] In step S205, if it is determined that the target detection result meets the preset defect requirements, a target grade of the wound yarn package is obtained based on the target detection result of the wound yarn package and the preset grade of the wound yarn package.

[0042] In step S206, if it is determined that the target detection result does not satisfy the preset defect requirement, the preset grade of the wound yarn package is set as the target grade of the wound yarn package.

[0043] In this way, the present disclosure can use a model to perform defect detection of a yarn package, and further determine a target grade of the yarn package based on the obtained detection result (e.g., a target detection result), thus allowing the actual grade of the yarn package to be accurately evaluated, thereby realizing automatic adjustment of the grade of the yarn package, and further improving the efficiency of managing the yarn package significantly.In addition, since the present disclosure uses a model to perform defect detection, it can realize efficient recognition of defects in the yarn package, and further improve the accuracy of evaluation of the grade of the yarn package significantly.

[0044] Furthermore, in one example, the target detection model may be a dynamic weight-based wavelet attention neural network, or other models obtained by improving based on Dynamic Wavelet Convolution Networks (DWCNet), and the present disclosure is not particularly limited in this regard.

[0045] Moreover, in one embodiment, the target detection model includes at least a first network layer, a second network layer, and a third network layer.

[0046] Specifically, the first network layer is used to perform feature processing on the input image and obtain a low-level feature map, for example, the first network layer may be a dynamic wavelet convolutional network, where the low-level feature map is used to represent the feature map extracted after removing background noise in the image.

[0047] Furthermore, the second network layer is used for performing feature enhancement processing on at least the key feature information in the low-level feature map to obtain a high-level feature map. For example, in one example, the second network layer can be specifically used for performing feature enhancement processing on the key feature information in the low-level feature map, and accordingly suppressing irrelevant feature information in the low-level feature map, such as suppressing feature information irrelevant to the yarn package, to obtain a high-level feature map.

[0048] Furthermore, the third network layer is used to perform defect recognition based on the high-level feature map to obtain a first detection result.

[0049] In this manner, the present disclosure provides a specific example of a model for rapid defect detection for a wound yarn package, and thus, compared to conventional manual methods, the present disclosure can realize efficient detection of defects in a wound yarn package without requiring human experience, and in particular can detect weak defects in a wound yarn package that are difficult to detect, thereby further improving the accuracy of detection and laying the foundation for subsequent rapid evaluation of the grade of the wound yarn package.

[0050] In an embodiment of the present disclosure, the first network layer may include at least a first subnetwork layer, a second subnetwork layer, a third subnetwork layer, and a fourth subnetwork layer.

[0051] Here, the first sub-network layer is used for performing local feature extraction on the input image to obtain a low-frequency feature map, the second sub-network layer is used for obtaining a target weighting factor of the feature map of the input image, the third sub-network layer is used for performing global feature extraction on the input image to obtain a global feature map (e.g., to obtain a global feature map including low-frequency information and high-frequency information), and the fourth sub-network layer is used for fusing the low-frequency feature map and the global feature map based on the target weighting factor to perform noise filtering and obtain a low-level feature map. That is, the first network layer can combine the extracted local feature information (e.g., low-frequency feature map) and global feature information (e.g., global feature map) to obtain a feature map with higher quality, thus better capturing the feature information of defects, thus improving the accuracy of defect detection, and laying the foundation for accurately evaluating the grade of the wound yarn package thereafter.

[0052] For example, in one example, the target weight coefficient determined by the second sub-network layer is w, the low-frequency feature map obtained by the first sub-network layer is p, the global feature map obtained by the third sub-network layer is q, and the low-level feature map obtained by the fourth sub-network layer is x, and the specific expression is as follows:

[0053]

number

[0054] In this way, the present disclosure can utilize the obtained weighting coefficients to appropriately combine the obtained local feature information and global feature information, and thus adjust the noise present in the feature map, which helps to mitigate the interference of noise on defect detection, and further improve the accuracy of defect detection.

[0055] In another example of the present disclosure, the second network layer includes at least a fifth subnetwork layer, a sixth subnetwork layer, and a seventh subnetwork layer.

[0056] Here, the fifth sub-network layer is used for performing feature extraction on the low-level feature map, and fusing the extracted feature maps to obtain M initial fused feature maps, where M is an integer equal to or greater than 2; the sixth sub-network layer is used for performing key feature extraction on each of the M initial fused feature maps, and performing feature enhancement on the key feature information extracted from each initial fused feature map to obtain M target enhanced feature maps; and the seventh sub-network layer is used for fusing the obtained M target enhanced feature maps to obtain a high-level feature map. In this way, clear and abundant feature information (e.g., a high-level feature map) can be obtained, and the accuracy of defect detection can be further improved, especially for weak defects in the yarn package that are difficult to detect, which significantly improves the accuracy of weak defect detection, and lays the foundation for subsequent accurate yarn package grading.

[0057] For example, the fifth sub-network layer is a feature pyramid network (FPN) layer, the sixth sub-network layer is a feature enhancement network layer, and the seventh sub-network layer is a linear addition (Add) processing layer. As shown in FIG. 4, first, a low-level feature map is input to an FPN layer to obtain M initial fusion feature maps. Then, each of the obtained initial fusion feature maps is input to a feature enhancement network layer, for example, key feature extraction is performed for each initial fusion feature map through the feature enhancement network layer, and feature enhancement processing is performed for the extracted feature map to obtain M target enhancement feature maps. Finally, the obtained M target enhancement feature maps are input to a linear addition processing layer to obtain a high-level feature map. In this way, it is helpful to improve the accuracy of defect and weak defect detection, and provides a basis for accurately evaluating the grade of the subsequent winding package.

[0058] Here, in one example, the i-th target-enhanced feature map among the M target-enhanced feature maps is obtained based on the following method.

[0059] A convolution process is performed on the i-th initial fusion feature map, an i-th weighting factor is obtained according to the result of the convolution process, the i+1-th initial fusion feature map and the i-th initial fusion feature map are fused according to the i-th weighting factor to obtain an i-th target fusion feature map, key features are extracted from the obtained i-th target fusion feature map, feature enhancement processes (e.g., enhancement processes to appropriate degrees for different feature maps) are performed on the extracted feature maps respectively to obtain multiple i-th initial enhanced feature maps, and the multiple i-th initial enhanced feature maps are fused to obtain an i-th target enhanced feature map, where i is an integer between 1 and M-1.

[0060] Furthermore, when i is M, the M+1th initial fusion feature map may be a preset value, in which case the Mth target fusion feature map is obtained, and then the Mth target enhanced feature map is obtained. Alternatively, to obtain the Mth target enhanced feature map, the following may be done:

[0061] Key feature extraction is performed on the Mth initial fusion feature map, and an enhancement process is performed on each of the extracted feature maps to obtain a plurality of Mth initial enhanced feature maps, and the plurality of Mth initial enhanced feature maps are subjected to a fusion process to obtain an Mth target enhanced feature map.

[0062] For example, still take the sixth sub-network layer as an example of a feature enhancement network layer. In this case, the feature enhancement network layer can further specifically include a feature weight module and a multi-view attention module. As shown in FIG. 5, the i-th initial fusion feature map and the i+1-th initial fusion feature map among the M initial fusion feature maps are input into the feature weight module to obtain the i-th target fusion feature map, and the obtained i-th target fusion feature map is input into the multi-view attention module to perform key feature extraction and feature enhancement processing to obtain the i-th target enhanced feature map; and for the M-th initial fusion feature map, the M-th initial fusion feature map is directly used as the M-th target fusion feature map, and the obtained M-th target fusion feature map is input into the multi-view attention module to perform key feature extraction and feature enhancement processing to obtain the M-th target enhanced feature map.

[0063] Furthermore, as shown in FIG. 6, the feature weight module first performs upsampling processing on the (i+1)th initial fusion feature map to obtain the processed (i+1)th initial fusion feature map, performs convolution processing on the i-th initial fusion feature map, and performs Sigmod processing on the result obtained after processing to obtain the i-th weight coefficient (e.g., λ in the figure). i , or 1-λ i ), and then the i-th weight coefficient, e.g., 1-λ iand perform element-by-element multiplication on the i-th initial fused feature map, λ i Then, perform element-by-element multiplication on the i+1th initial fusion feature map processed by the above, and perform element-by-element addition on the obtained result to obtain the ith target fusion feature map. Here, if i is M, the Mth initial fusion feature map does not need to be input to the feature weight module, and can be directly used as the Mth target fusion feature map.

[0064] Furthermore, as shown in FIG. 7, the multi-view attention module first performs a deformable convolution process of the first preset convolution kernel (e.g., 1×1) on the i-th target fusion feature map to extract key features, and obtains a key feature feature map 1; then performs a deformable convolution process of the second preset convolution kernel (e.g., 3×3) on the i-th target fusion feature map to obtain a key feature feature map 2; and performs a deformable convolution process of the third preset convolution kernel (e.g., 5×5) on the i-th target fusion feature map to obtain a key feature feature map 3; then inputs the obtained feature map 1 into the attention submodule A to perform feature enhancement process on the feature map 1 to obtain an i-th initial enhanced feature map 1; similarly, inputs the feature map 2 into the attention submodule B to obtain an i-th initial enhanced feature map 2; inputs the feature map 3 into the attention submodule C to obtain an i-th initial enhanced feature map 3; and finally performs a fusion process (e.g., element addition) of the above three i-th initial enhanced feature maps to obtain an i-th target enhanced feature map.

[0065] In addition to what is shown in FIG. 7, a neural network can be added, for example a wavelet neural network, and the wavelet neural network can be used to directly perform feature extraction on the i-th target fusion feature map, to obtain the i-th supplementary feature map, and then the fusion process can be performed together with the above three i-th initial enhanced feature maps.

[0066] In addition, the structure of the attention submodule (e.g., attention submodule A, or attention submodule B, or attention submodule C) included in the multi-view attention module may be a Squeeze and Excitation (SE) network, or other network structures obtained based on an SE network, and the present disclosure is not limited thereto.

[0067] In this way, the present disclosure can make full use of the sixth sub-network layer to extract key features in the low-level feature map, and then perform feature enhancement processing on the extracted results, thereby directing the network's attention to potential targets (i.e., defects in the yarn package) and helping to better recognize subsequent defects; especially for weak defects that are difficult to detect in the yarn package, key features of weak defects can be obtained more comprehensively, further improving the detection accuracy of weak defects, and laying the foundation for subsequent accurate grade evaluation of the yarn package.

[0068] In a further example of the present disclosure, the third network layer includes an eighth sub-network layer and a ninth sub-network layer.

[0069] Here, the eighth sub-network layer is used to frame candidate regions of false defects in the image based on the high-level feature map, and the ninth sub-network layer is used to recognize whether there is a defect in the candidate region based on the high-level feature map and the frame-selected candidate region. In this way, the defects that may exist in the yarn package can be efficiently recognized, which lays the foundation for the subsequent accurate evaluation and automatic adjustment of the grade of the yarn package.

[0070] 8 is a schematic flow chart of a method for processing yarn package data according to an embodiment of the present disclosure. This method can be selectively applied to electronic devices such as personal computers, servers, server clusters, etc. The relevant contents of the methods shown in Figs. 1 and 2 above can also be applied to this embodiment, and the relevant contents will not be repeated in this embodiment.

[0071] In addition, the method includes at least part of the following contents: As shown in FIG.

[0072] In step S801, if it is determined that the transported wound yarn package has entered the detection area, N images of the wound yarn package are acquired, where an image in the N images includes at least a portion of the wound yarn package, and N is 3 or greater.

[0073] In step S802, the N images are input into a target detection model to obtain a first detection result corresponding to each image in the N images.

[0074] Here, the target detection model is used to recognize whether a defect exists in a local area of ​​the wound yarn package and obtain a first detection result, the first detection result including at least one of the number of defects, the defect positions, and the defect types.

[0075] In step S803, information regarding a defect in the wound yarn package is obtained, and at least one of the following is performed to determine a target detection result representing the degree of the defect in the wound yarn package based on the information regarding the defect in the wound yarn package. Method 1: The number of defects contained in each of the N first detection results is counted to obtain the total number of defects in the wound yarn package. Method 2: Based on the defect positions (e.g., including positions and defect sizes) contained in each of the N first detection results, obtain all defect positions (e.g., obtain the overall defect size) in the wound yarn package. Method 3: Based on the defect type included in each of the N first detection results, a target defect type for the winding package (e.g., this target defect type includes the defect types included in all of the first detection results) is obtained.

[0076] In step S804, it is determined whether the target detection result satisfies the preset defect requirement, and if so, proceed to step S805, otherwise proceed to step S806.

[0077] In step S805, if it is determined that the target detection result meets the preset defect requirements, a target grade of the wound yarn package is obtained based on the target detection result of the wound yarn package and the preset grade of the wound yarn package.

[0078] Further, in one embodiment, the target grade of the yarn package can be obtained by the following method, which can automatically adjust the grade of the yarn package, and further improve the management efficiency of the yarn package, thereby ensuring the data quality of the wound ball data. Specifically, when it is determined that the above target detection result meets the preset defect requirements, obtaining the target grade of the yarn package based on the target detection result of the yarn package and the preset grade of the yarn package (for example, the above step S805) can be specifically performed as follows: If it is determined that at least one of the following is satisfied, the preset grade of the wound yarn package is downgraded to obtain a target grade of the detected wound yarn package.

[0079] Condition 1: The total number of defects in the wound yarn package in the target detection result is greater than a preset threshold value.

[0080] Condition 2: All defect positions of the wound yarn package in the target detection result are within a predetermined defect position range (e.g., the overall defect size is within the predetermined defect position range and exceeds a predetermined area).

[0081] Condition 3: The target defect type of the winding package in the target detection result is within a range of preset defect types (for example, if the target defect types include defect types included in all of the first detection results, at least one defect type is within the range of preset defect types, and then the target defect type can be considered to be within the range of preset defect types).

[0082] In step S806, if it is determined that the target detection result does not satisfy the preset defect requirement, the preset grade of the wound yarn package is set as the target grade of the wound yarn package.

[0083] In this way, the present disclosure can use a model to perform defect detection of a wound yarn package, and determine the grade of the wound yarn package based on the detection result (e.g., a target detection result), thereby automatically adjusting the grade of the wound yarn package, and greatly improving the management efficiency of the wound yarn package, thereby ensuring the data quality of the wound ball data, and effectively avoiding possible losses to users.

[0084] As described above, the technical solution of the present disclosure has several advantages over the prior art, specifically including:

[0085] First, it realizes automatic processing: compared with the traditional manual method, the method disclosed herein can efficiently detect defects in a wound yarn package without relying on human experience, and can automatically adjust the grade of the wound yarn package based on the detection result, thereby saving a large amount of labor and time costs, and further improving the management efficiency of the wound yarn package.

[0086] Second, improve the detection accuracy: Traditional manual detection methods are difficult to detect appearance defects in wound yarn packages and are prone to overlooking, while the solution disclosed herein uses a neural network model that is more suitable for defect detection, and has higher detection accuracy, especially for weak defects that are difficult to detect, thus ensuring the quality of wound yarn packages.

[0087] Third, improve the quality of the wound ball data. In the application process of the wound ball data stream, the solution disclosed herein can further improve the data quality of a wound ball package when detecting appearance defects of a wound ball package, thereby better realizing the quality control and feedback of the wound ball package in the production process, and effectively avoiding possible losses to users.

[0088] The present disclosure further proposes a processing device for wound yarn package data, as shown in FIG. 9, which comprises: a defect detection unit 901 for performing defect detection on the wound yarn package located within a detection area when it is determined that the transported wound yarn package has entered the detection area, and obtaining a target detection result for the wound yarn package, the target detection result being used to indicate the degree of a defect in the wound yarn package; and a grade determination unit 902 for obtaining a target grade of the wound yarn package based on the target detection result of the wound yarn package and a preset grade of the wound yarn package when it is determined that the target detection result meets the preset defect requirements.

[0089] In one embodiment of the present disclosure, the defect detection unit 901 specifically includes: acquiring N images of the wound yarn package, where an image in the N images includes at least a portion of an area of ​​the wound yarn package, where N is 3 or greater; inputting the N images into a target detection model to obtain a first detection result corresponding to each image in the N images, the target detection model being used to recognize whether there is a defect in a local area of ​​the wound yarn package and obtain a first detection result, the first detection result including at least one of a defect number, a defect location, and a defect type; and obtaining a target detection result of the wound yarn package based on a first detection result corresponding to each image in the N images.

[0090] In one embodiment of the present disclosure, the defect detection unit 901 specifically includes: obtaining information about defects in the wound package and determining a target detection result indicative of a degree of the defect in the wound package based on the information about the defects in the wound package; calculating a total number of defects in the yarn package by calculating a number of defects in each of the N first detection results; obtaining all defect positions in the wound yarn package based on defect positions included in each of the N first detection results; and obtaining a target defect type for the wound yarn package based on a defect type included in each of the N first detection results.

[0091] In one embodiment of the present disclosure, the grade determination unit 902 specifically includes: the total number of defects in the yarn package in the target detection result is greater than a preset threshold value; and all defect positions of the wound yarn package in the target detection result are within a preset defect position range; and When it is determined that at least one of the following conditions is satisfied: the target defect type of the wound yarn package in the target detection result is within a range of preset defect types, the target defect type is used to downgrade and adjust the preset grade of the wound yarn package to obtain the target grade of the detected wound yarn package.

[0092] In one embodiment of the present disclosure, the target detection model includes at least a first network layer, a second network layer, and a third network layer; The first network layer is used to perform feature processing on the input image to obtain a low-level feature map, and the low-level feature map is used to represent the feature map extracted after removing background noise in the image; The second network layer is used for performing feature enhancement processing on key feature information in at least the low-level feature map to obtain a high-level feature map; The third network layer is used for performing defect recognition based on the high-level feature map to obtain a first detection result.

[0093] In one embodiment of the present disclosure, the first network layer includes at least a first subnetwork layer, a second subnetwork layer, a third subnetwork layer, and a fourth subnetwork layer; The first sub-network layer is used for performing local feature extraction on the input image to obtain a low-frequency feature map, the second sub-network layer is used for obtaining a target weighting coefficient of the feature map of the input image, the third sub-network layer is used for performing global feature extraction on the input image to obtain a global feature map, and the fourth sub-network layer is used for fusing the low-frequency feature map and the global feature map based on the target weighting coefficient to remove noise and obtain a low-level feature map.

[0094] In one embodiment of the present disclosure, the second network layer includes at least a fifth subnetwork layer, a sixth subnetwork layer, and a seventh subnetwork layer; The fifth sub-network layer is used for performing feature extraction on the low-level feature maps, and fusing the extracted feature maps to obtain M initial fused feature maps, where M is an integer equal to or greater than 2; The sixth sub-network layer is used for extracting key features from each of the M initial fusion feature maps, and performing feature enhancement processing on the key feature information extracted from each of the M initial fusion feature maps to obtain M target enhanced feature maps; The seventh sub-network layer is used to fuse the obtained M target-enhanced feature maps to obtain a high-level feature map.

[0095] In one embodiment of the present disclosure, the i-th target-enhanced feature map of the M target-enhanced feature maps is When i is an integer between 1 and M-1, the method comprises: performing a convolution process on the i-th initial fusion feature map to obtain an i-th weighting factor based on the result of the convolution process; fusing the (i+1)-th initial fusion feature map and the i-th initial fusion feature map based on the i-th weighting factor to obtain an i-th target fusion feature map; performing key feature extraction on the obtained i-th target fusion feature map; performing feature enhancement process on each of the extracted feature maps to obtain a plurality of i-th initial enhanced feature maps; and fusing the plurality of i-th initial enhanced feature maps to obtain the i-th target enhanced feature map; When i is M, the Mth target enhanced feature map is obtained by performing key feature extraction on the Mth initial fusion feature map, performing enhancement processing on each of the extracted feature maps to obtain multiple Mth initial enhanced feature maps, and performing fusion processing on the multiple Mth initial enhanced feature maps to obtain the Mth target enhanced feature map.

[0096] The specific functions and exemplary descriptions of each module and sub-module of the apparatus according to the embodiments of the present disclosure may refer to the relevant descriptions of the corresponding steps in the above-mentioned method embodiments, and will not be repeated here.

[0097] In the technical solution disclosed herein, the acquisition, storage, and application of users' personal information complies with the provisions of relevant laws and regulations and does not violate public order and morals.

[0098] FIG. 10 is a structural block diagram of an electronic device according to an embodiment of the present disclosure. As shown in FIG. 10, the electronic device includes a memory 1010 and a processor 1020, and the memory 1010 stores a computer program executable by the processor 1020. The number of the memory 1010 and the processor 1020 may be one or more. The memory 1010 may store one or more computer programs, which, when executed by the electronic device, cause the electronic device to perform the method provided by the above method embodiment. The electronic device may further include: A communication interface 1030 is used to communicate with an external device and perform data interaction and transmission.

[0099] When the memory 1010, the processor 1020, and the communication interface 1030 are implemented independently, the memory 1010, the processor 1020, and the communication interface 1030 are connected to each other via a bus and can communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be classified into an address bus, a data bus, a control bus, etc. For ease of explanation, only one thick line is shown in FIG. 10, but this does not indicate only one bus or only one type of bus.

[0100] Optionally, in a specific implementation, when the memory 1010, the processor 1020, and the communication interface 1030 are integrated on one chip, the memory 1010, the processor 1020, and the communication interface 1030 can communicate with each other via an internal interface.

[0101] It should be understood that the processor may be a Central Processing Unit (CPU) or may be other general purpose processors, Digital Signal Processing (DSP), Application Specific Integrated Circuits (ASIC), Field Programmable Gate Arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware assemblies, etc. The general purpose processor may be a microprocessor or any conventional processor, etc. Additionally, the processor may be a processor supporting the Advanced RISC Machines (ARM) architecture.

[0102] Additionally, optionally, the memory may include read-only memory and random access memory, and may further include non-volatile random access memory. The memory may be either volatile or non-volatile memory, or may include both volatile and non-volatile memory. Here, non-volatile memory may include Read-Only Memory (ROM), Programmable ROM (PROM), Erasable PROM (EPROM), Electrically EPROM (EEPROM), or flash memory. Volatile memory may include Random Access Memory (RAM), which acts as an external cache. By way of example and not limitation, many forms of RAM are available. For example, static random access memory (Static RAM, SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (Enhanced SDRAM, ESDRAM), synchlink DRAM (Synchlink DRAM, SLDRAM), and direct RAMBUS RAM (DR RAM).

[0103] The above-mentioned embodiments may be implemented in whole or in part in software, hardware, firmware, or any combination thereof. When implemented in software, the whole or in part may be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, a process or function according to the embodiments of the present disclosure is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another, for example, the computer instructions may be transmitted from one website site, computer, server, or data center to another website site, computer, server, or data center via wire (e.g., coaxial cable, optical fiber, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, Bluetooth, microwave, etc.). The computer-readable storage medium may be any available medium that can be accessed by a computer, or a data storage device including a server, a data center, etc. integrated with one or more available media. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a Digital Versatile Disc (DVD)), or a semiconductor medium (e.g., a Solid State Disk (SSD)). Note that the computer-readable storage medium referred to in this disclosure may be a non-volatile storage medium, in other words, a non-transitory storage medium.

[0104] Those skilled in the art can understand that all or part of the steps for realizing the above embodiments may be implemented by hardware, or may be implemented by instructing relevant hardware by a program, and the program may be stored in a computer-readable storage medium, and the storage medium may be a read-only memory, a magnetic disk, an optical disk, etc.

[0105] In the description of the embodiments of the present disclosure, the reference words "one embodiment", "several embodiments", "examples", "specific examples", or "several examples" mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present disclosure. And, the specific features, structures, materials, or characteristics described can be combined in any suitable manner in any one or more embodiments or examples. Furthermore, a person skilled in the art may combine different embodiments or examples described in the present disclosure and features of different embodiments or examples to the extent that they are not mutually inconsistent.

[0106] In the description of the embodiments of the present disclosure, unless otherwise stated, " / " means "or", for example, A / B can mean either A or B. "And / or" in the present disclosure is only for describing the related relationship of related objects, and indicates that three types of relationships may exist, for example, A and / or B can indicate the following three types of situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0107] In the description of the embodiments of the present disclosure, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying a relative importance, nor should they be construed as implying a number of technical features shown. Thus, features defined as "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the embodiments of the present disclosure, "plurality" means two or more, unless otherwise specified.

[0108] The above are merely illustrative examples of the present disclosure, and are not intended to limit the present disclosure. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present disclosure should be included within the scope of protection of the present disclosure.

Claims

1. A method for processing wound yarn package data, comprising the steps of: when it is determined that the transported wound yarn package has entered a detection area, a defect detection is performed on the wound yarn package located within the detection area to obtain a target detection result for the wound yarn package, the target detection result being used to indicate the degree of a defect in the wound yarn package; When it is determined that the target detection result satisfies a preset defect requirement, obtaining a target grade of the wound yarn package based on the target detection result of the wound yarn package and a preset grade of the wound yarn package; Here, performing defect detection on a wound yarn package located within the detection area and obtaining a target detection result for the wound yarn package includes: acquiring N images of the wound yarn package, where the N images include at least a portion of an area of ​​the wound yarn package, and N is 3 or greater; inputting the N images into a target detection model to obtain a first detection result corresponding to each image in the N images, the target detection model being used to recognize whether a defect exists in a local area of ​​the wound yarn package and obtain a first detection result, the first detection result including at least one of a defect number, a defect location, and a defect type; obtaining a target detection result of the wound yarn package based on a first detection result corresponding to each image in the N images; The target detection model includes at least a first network layer, a second network layer, and a third network layer, the first network layer is used for performing feature processing on an input image to obtain a low-level feature map, and the low-level feature map is used for representing a feature map extracted after removing background noise in the image; the second network layer is used for performing feature enhancement processing on at least key feature information in the low-level feature map to obtain a high-level feature map; and the third network layer is used for performing defect recognition based on the high-level feature map to obtain a first detection result; The first network layer includes at least a first sub-network layer, a second sub-network layer, a third sub-network layer, and a fourth sub-network layer, the first sub-network layer is used for performing local feature extraction on the input image to obtain a low-frequency feature map, the second sub-network layer is used for obtaining a target weighting coefficient of the feature map of the input image, the third sub-network layer is used for performing global feature extraction on the input image to obtain a global feature map, and the fourth sub-network layer is used for performing noise removal by fusing the low-frequency feature map and the global feature map based on the target weighting coefficient to obtain a low-level feature map. A method for processing yarn package data.

2. Obtaining a target detection result of the wound yarn package based on a first detection result corresponding to each image in the N number of images, obtaining information about defects in the wound package and determining a target detection result indicative of a degree of the defect in the wound package based on the information about the defects in the wound package; calculating a number of defects included in each of the N first detection results to obtain a total number of defects in the wound yarn package; obtaining all defect positions in the wound yarn package based on defect positions included in each of the N first detection results; obtaining a target defect type for the wound yarn package based on a defect type included in each of the N first detection results. The method for processing yarn package data according to claim 1.

3. obtaining a target grade for the wound yarn package based on the target detection result of the wound yarn package and a preset grade for the wound yarn package when it is determined that the target detection result satisfies a preset defect requirement; the total number of defects in the yarn package in the target detection result is greater than a preset threshold value; and all defect positions of the wound yarn package in the target detection result are within a preset defect position range; and and when it is determined that at least one of the target defect type of the wound yarn package in the target detection result is within a range of preset defect types, downgrading the preset grade of the wound yarn package to obtain a target grade of the detected wound yarn package. The method for processing wound yarn package data according to claim 2.

4. the second network layer includes at least a fifth subnetwork layer, a sixth subnetwork layer, and a seventh subnetwork layer; The fifth sub-network layer is used for performing feature extraction on the low-level feature maps, and fusing the extracted feature maps to obtain M initial fused feature maps, where M is an integer equal to or greater than 2; The sixth sub-network layer is used for extracting key features from each of the M initial fusion feature maps, and performing feature enhancement processing on the key feature information extracted from each of the M initial fusion feature maps to obtain M target enhanced feature maps; The seventh sub-network layer is used to fuse the obtained M target-enhanced feature maps to obtain a high-level feature map; The method for processing yarn package data according to claim 1.

5. The i-th target-enhanced feature map of the M target-enhanced feature maps is When i is an integer between 1 and M-1, the method comprises: performing a convolution process on the i-th initial fusion feature map to obtain an i-th weighting factor based on a result of the convolution process; performing a fusion process on the (i+1)th initial fusion feature map and the i-th initial fusion feature map based on the i-th weighting factor to obtain an i-th target fusion feature map; performing key feature extraction on the obtained i-th target fusion feature map; performing a feature enhancement process on each of the extracted feature maps to obtain a plurality of i-th initial enhanced feature maps; and performing a fusion process on the plurality of i-th initial enhanced feature maps to obtain the i-th target enhanced feature map. When i is M, the Mth initial fusion feature map is subjected to key feature extraction, and the extracted feature maps are subjected to enhancement processing to obtain a plurality of Mth initial enhanced feature maps, and the Mth target enhanced feature map is obtained by performing a fusion processing on the plurality of Mth initial enhanced feature maps. The method for processing wound yarn package data according to claim 4.

6. A processing device for winding package data, comprising: a defect detection unit for performing defect detection on the wound yarn package located within a detection area when it is determined that the transported wound yarn package has entered the detection area, and obtaining a target detection result for the wound yarn package, the target detection result being used to indicate the degree of a defect in the wound yarn package; a grade determination unit for obtaining a target grade of the wound yarn package based on the target detection result of the wound yarn package and a preset grade of the wound yarn package when it is determined that the target detection result meets a preset defect requirement; Here, the defect detection unit specifically includes: acquiring N images of the wound yarn package, where an image in the N images includes at least a portion of an area of ​​the wound yarn package, and N is 3 or greater; inputting the N images into a target detection model to obtain a first detection result corresponding to each image in the N images, the target detection model being used to recognize whether a defect exists in a local area of ​​the wound yarn package and obtain a first detection result, the first detection result including at least one of a defect number, a defect location, and a defect type; obtaining a target detection result of the wound yarn package based on a first detection result corresponding to each image in the N number of images; The target detection model includes at least a first network layer, a second network layer, and a third network layer, the first network layer is used for performing feature processing on an input image to obtain a low-level feature map, and the low-level feature map is used for representing a feature map extracted after removing background noise in the image; the second network layer is used for performing feature enhancement processing on at least key feature information in the low-level feature map to obtain a high-level feature map; and the third network layer is used for performing defect recognition based on the high-level feature map to obtain a first detection result; The first network layer includes at least a first sub-network layer, a second sub-network layer, a third sub-network layer, and a fourth sub-network layer, the first sub-network layer is used for performing local feature extraction on the input image to obtain a low-frequency feature map, the second sub-network layer is used for obtaining a target weighting coefficient of the feature map of the input image, the third sub-network layer is used for performing global feature extraction on the input image to obtain a global feature map, and the fourth sub-network layer is used for performing noise removal by fusing the low-frequency feature map and the global feature map based on the target weighting coefficient to obtain a low-level feature map. A yarn package data processing device.

7. At least one processor; a memory in communication with the at least one processor; The memory stores instructions executable by the at least one processor, the instructions, when executed by the at least one processor, causing the at least one processor to perform a method according to any one of claims 1 to 5. Electronic devices.

8. A non-transitory computer readable storage medium having stored thereon instructions for causing a computer to perform the method of any one of claims 1 to 5.

9. A program for implementing the method according to any one of claims 1 to 5 when executed by a processor in a computer.

Images