Detection method, apparatus, electronic device, storage medium, and program

Abstract

A detection method, a detection apparatus, an electronic device, a storage medium, and a program are provided. The method includes: obtaining a target prompt for instructing a user to search for a target winding package; obtaining a plurality of target images of a target winding package carriage where the target winding package indicated by the target prompt is located; inputting the plurality of target images and the target prompt into a target detection model to obtain a target output image indicating the location where the target winding package is located; and the target detection model identifies the winding package from the input image based on the target winding package indicated by the target prompt. (wherein the input image is one of the plurality of target images) , indicating the location of the target wound yarn package. The target output Image obtain , and includes.

Description

[Technical Field]

[0001] The present disclosure relates to the field of data processing technology, and in particular to a detection method, an apparatus, an electronic device, a storage medium, and a program. [Background technology]

[0002] In the processing of wound yarn packages, particularly in the dyeing process, it is usually necessary to manually search for wound yarn packages from a storage area such as a warehouse in order to process the wound yarn packages with specific processing parameters, for example, dyeing. Summary of the Invention [Problem to be solved by the invention]

[0003] However, this method does not allow for quick and accurate positioning relative to the wound yarn package, resulting in a decrease in processing efficiency. [Means for solving the problem]

[0004] The present disclosure provides a detection method, a detection apparatus, an electronic device, a storage medium, and a program.

[0005] According to a first aspect of the present disclosure, there is provided a detection method applied to a cloud terminal, the method comprising: obtaining a target prompt to instruct the user to locate a target spool package; acquiring a plurality of target images of a target winding package cart on which a target winding package indicated by the target prompt is located, the target winding package cart including two placement areas for placing winding packages, a first image of at least one of the plurality of target images including all of the winding packages placed by the first placement area of ​​the two placement areas, and a second image of at least one of the plurality of target images including all of the winding packages placed by the second placement area of ​​the two placement areas; inputting a plurality of target images and the target prompt into a target detection model to obtain a target output image indicating a location where a target winding package is located, the target detection model identifying the winding package for the input image based on the target winding package indicated by the target prompt; wherein the input image is one of the plurality of target images; Indicate the location of the target yarn package The target output Image obtain , and includes.

[0006] According to a second aspect of the present disclosure, there is provided a detection device applicable to a cloud terminal, the device comprising: an information acquisition unit; and a detection unit; the information acquisition unit is used to acquire a target prompt for instructing the user to search for a target wound thread package; and acquire a plurality of target images of a target wound thread package cart where the target wound thread package indicated by the target prompt is located, the target wound thread package cart including two placement areas for placing wound thread packages, at least one first image of the plurality of target images including all wound thread packages placed by the first placement area of ​​the two placement areas, and at least one second image of the plurality of target images including all wound thread packages placed by the second placement area of ​​the two placement areas; The detection unit inputs a plurality of target images and the target prompt into a target detection model to obtain a target output image indicating a location where a target winding package is located, and the target detection model identifies the winding package for the input image based on the target winding package indicated by the target prompt. (wherein the input image is one of the plurality of target images) , where the target wound yarn package is located The target output Image indicating the position obtain , especially used.

[0007] According to a third aspect of the present disclosure, there is provided an electronic device, the device comprising: at least one processor; a memory communicatively coupled to 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.

[0008] According to a fourth aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform any one of the methods in the embodiments of the present disclosure.

[0009] According to a fifth aspect of the present disclosure, there is provided a program, which, when executed by a processor, implements any one of the methods in the embodiments of the present disclosure.

[0010] Thus, according to the solution of the present disclosure, a target prompt and a target detection model are used to identify a yarn package for each captured target image, and an image (e.g., a target output image) of the location of the target yarn package indicated by the target prompt can be obtained. Thus, compared with the traditional manual search method, the solution of the present disclosure can quickly and accurately determine the specific location of the target yarn package, realize automatic positioning of the yarn package, and allow workers to quickly find the required yarn package based on the acquired location information, thereby saving a lot of labor and time costs and laying the foundation for the subsequent normal operation of the yarn package processing process.

[0011] It should be understood that the contents described herein are not intended to describe key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will be better understood through the following specification. [Brief explanation of the drawings]

[0012] In the accompanying drawings, unless otherwise specified, the same reference numerals throughout the 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 illustrate only some examples provided by the present disclosure and should not be considered as limiting the scope of the present disclosure.

[0013] [Figure 1] 1 is a schematic flowchart of a detection method according to an embodiment of the present disclosure. [Figure 2] 10A-10C are schematic diagrams illustrating a target winding package dolly scenario according to one embodiment of the present disclosure. [Figure 3A] FIG. 2 is a schematic diagram illustrating a model structure of a target detection model according to an embodiment of the present disclosure. [Figure 3B] FIG. 1 is a schematic diagram illustrating the splitting effect of an example split network module according to an embodiment of the present disclosure. [Figure 3C] FIG. 10 is a schematic diagram illustrating the effects of an example information mapping module according to an embodiment of the present disclosure. [Figure 3D] FIG. 1 is a schematic diagram illustrating the effects of an example of an identification module according to an embodiment of the present disclosure. [Figure 4A] FIG. 1 is a block diagram illustrating a split network module according to an embodiment of the present disclosure. [Figure 4B] FIG. 2 is a schematic diagram illustrating segmenting a halftone dot presentation image according to an embodiment of the present disclosure. [Figure 5] FIG. 1 is a block diagram illustrating a pre-feature layer included in a segmentation network module according to an embodiment of the present disclosure. [Figure 6] FIG. 1 is a block diagram illustrating a similarity diagram prior knowledge layer included in a semantics prior knowledge layer according to an embodiment of the present disclosure. [Figure 7] FIG. 1 is a diagram illustrating a configuration of a detection device according to an embodiment of the present disclosure. [Figure 8] FIG. 1 is a block diagram of an electronic device for implementing a detection method according to an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0014] The present disclosure will now be described in more detail with reference to the drawings, in which like reference numerals indicate functionally identical or similar elements, and in which various aspects of the embodiments are shown, but which are not necessarily drawn to scale unless specifically noted.

[0015] Furthermore, numerous specific details are set forth in the following specific embodiments in order to better explain the present disclosure. Those skilled in the art will appreciate that the present disclosure may be similarly practiced and obtained. identification In some instances, methods, procedures, elements, circuits, etc. that are well known to those skilled in the art are not described in detail in order to clarify the scope of the present disclosure.

[0016] The disclosed method proposes a detection method for quickly locating a wound yarn package of specified process parameters.

[0017] 1 is a schematic flow chart of a detection method according to an embodiment of the present disclosure, which is optionally applied to electronic devices such as personal computers, servers, server clusters, and the like.

[0018] Furthermore, the method includes at least part of the following content: As shown in Figure 1, the detection method is applied to the cloud and includes:

[0019] In step S101, a target prompt for instructing to search for a target wound yarn package is obtained.

[0020] In step S102, a plurality of target images of the target wound yarn package carriage where the target wound yarn package indicated by the target prompt is located are acquired.

[0021] Here, the target wound yarn package carriage includes two placement areas for placing wound yarn packages. For example, as shown in Fig. 2, a placement area is provided on each side of the target wound yarn package carriage, and each placement area includes a placement body for placing wound yarn packages. In this case, each placement area can place nine wound yarn packages.

[0022] Furthermore, at least one first image among the plurality of target images includes all winding thread packages placed by the first placement area of ​​the two placement areas (for example, the first image includes all winding thread packages placed by the placement area on one side shown in Figure 2), and at least one second image among the plurality of target images includes all winding thread packages placed by the second placement area of ​​the two placement areas (for example, the second image includes all winding thread packages placed by the placement area on the other side shown in Figure 2).

[0023] In step S103, the plurality of target images and the target prompts are input into a target detection model to obtain a target output image indicating the location of the target wound yarn package.

[0024] Here, the target detection model identifies a winding package for the input image based on the target winding package indicated by the target prompt, and generates an image indicating the location of the target winding package. obtain .

[0025] The target output image can at least show the specific position on the target wound yarn package carriage where the target wound yarn package is located.

[0026] Furthermore, in one example, target presentation information can be output simultaneously with the output of the target output image. The target presentation information indicates the specific location of the target wound yarn package cart where the target wound yarn package is located (e.g., the storage area of ​​the target wound yarn package cart). In this case, the target output image can indicate the specific location of the target wound yarn package on the target wound yarn package cart, and the target presentation information can present the specific location of the target wound yarn package cart. This allows the operator to quickly and accurately locate the required wound yarn package cart based on the target output image and the target presentation information.

[0027] In this way, the solution of the present disclosure uses the target prompt and the target detection model to identify a yarn package for each captured target image, and obtain an image (e.g., a target output image) of the location of the target yarn package indicated by the target prompt. In this way, compared to the traditional manual search method, the solution of the present disclosure can quickly and accurately determine the specific location of the target yarn package, realize automatic positioning of the yarn package, and allow workers to quickly find the required yarn package based on the acquired location information, thereby saving a lot of labor and time costs and laying the foundation for the subsequent normal operation of the yarn package processing process.

[0028] Furthermore, in one specific example, multiple target images can be acquired in the following manner, specifically, acquiring multiple target images of the target wound yarn package cart where the target wound yarn package indicated by the target prompt is located (for example, the above step S102) specifically includes the following:

[0029] In step S102-1, the label information of the target wound yarn package indicated by the target prompt is identification do. In step S102-2, based on the mark information of the target wound yarn package, the mark information of the target wound yarn package carriage on which the target wound yarn package is located is calculated. identification do.

[0030] Here, in an actual scenario, a mapping relationship between a winding yarn package and a winding yarn package cart can be stored in advance, and for example, in this mapping relationship, the label information of the winding yarn package and the label information of the winding yarn package cart are recorded.

[0031] In step S102-3, the storage area where the target wound yarn package cart is located is determined based on the label information of the target wound yarn package cart. identification do. In step S102-4, the plurality of target images are selected from the image data collected by the image collecting device in the storage area where the target wound yarn package cart is located.

[0032] The storage area where the target wound yarn package cart is located is identification In this case, video data stored in a database can be retrieved, the video data being collected by an image collecting device located in the storage area and whose video tag information includes identification information of the target wound yarn package cart, and multiple target images can be selected from the retrieved video data.

[0033] Furthermore, video data can be acquired and stored in the database in the following manner: Specifically, when it is detected that a target wound yarn package cart has entered the start position of a storage area, multiple image acquisition devices located at the start position of the storage area are activated to acquire video data by acquiring video of at least the placement areas on both sides of the target wound yarn package cart that has entered the storage area.

[0034] Furthermore, in order to facilitate video management and data retrieval, the identification information of the target wound yarn package cart that has entered the storage area is automatically identified, and the identification information of the identified target wound yarn package cart is acquired and used as video tag information of the collected video data. Then, the identification information of the target wound yarn package cart is identificationThen, video tag information corresponding to the mark information of the target wound yarn package cart is obtained, and video data is called up based on the video tag information corresponding to the mark information of the target wound yarn package cart, and a plurality of target images are selected from the called up video data.

[0035] In this way, the solution of the present disclosure can use the acquired mark information of the target winding yarn package cart to quickly retrieve video data related to the target winding yarn package cart, and then select and obtain multiple target images from the retrieved video data, thereby laying the foundation for automatically locating the specific location of the target winding yarn package cart indicated by the target prompt from the images.

[0036] In one specific example of the present disclosure, as shown in FIG. 3A, the target detection model includes at least a segmentation network module, an information mapping module, and an identification module.

[0037] Specifically, the segmentation network module identifies an area in the input image where each winding package is located based on a preset winding package prompt, and uses a mask to mask the area in the image where each winding package is located to obtain an initial mask image of the input image, where the input image is one of a plurality of target images.

[0038] For example, as shown in Figure 3B, target image 1 of the plurality of target images is input to a segmentation network module, which identifies the area in target image 1 where each yarn package is located based on a preset yarn package prompt, and uses a mask to mask the area in target image 1 where each yarn package is located, thereby obtaining an initial mask image of target image 1. This lays the foundation for quickly locating the required target yarn package later.

[0039] In one example, the information mapping module is used to perform information mapping processing on the initial mask image based on the target prompt to obtain a target mask image with label information of the yarn winding package corresponding to the input image, where the target mask image with label information of the yarn winding package corresponding to the input image can represent label information of each winding package actually included in the input image. For example, in one example, the information mapping module specifically obtains label information of the yarn winding packages to be placed by each mounting element of the target winding package truck based on label information of the target winding package truck, and maps the label information of the yarn winding packages to be placed by each mounting element of the target winding package truck to each mask plate located at different positions in the initial mask image to obtain a target mask image with label information of the yarn winding package corresponding to the input image, where the label information of the target winding package truck is obtained based on the label information of the target winding package indicated by the target prompt. For example, the information mapping module obtains the label information of the target winding package from a database based on the label information of the target winding package indicated by the target prompt.

[0040] In this example, the winding yarn packages to be placed by the respective placement bodies in the placement area of ​​the target winding yarn package carriage can refer to the winding yarn packages to be placed on the respective placement bodies according to a predetermined placement rule (or order), i.e., the winding yarn packages that should theoretically be placed by the respective placement bodies. Based on this, after obtaining the identification information of the target winding yarn package carriage, the identification information of the winding yarn packages that will theoretically be placed by the respective placement bodies in the target winding yarn package carriage can be obtained, thus laying the foundation for subsequent rapid detection or rapid location of the required winding yarn package.

[0041] Furthermore, the above-mentioned "mapping" can refer to adding the mark information of the yarn package to a mounting body where the yarn package should theoretically be located based on a preset placement rule. Furthermore, since each yarn package in the mask image is masked by a mask plate, the above-mentioned "mapping" can also refer to adding the mark information of the yarn package to a mask plate on a mounting body where the yarn package should theoretically be located based on a preset placement rule.

[0042] For example, continuing to use the initial mask image of target image 1 as an example, as shown in Figure 3C, the information mapping module obtains the label information of the yarn package that is theoretically placed on each of the placement bodies on the target yarn package carriage, and then adds the label information of the yarn package to the mask plate on the placement body where the yarn package should theoretically be located in the initial mask image of target image 1 based on the preset placement rule, thereby obtaining a target mask image with label information of the yarn package corresponding to target image 1. This lays the foundation for improving the efficiency of identifying the required target yarn package in the future.

[0043] In one example, the identification module identifies and marks the target winding package indicated by the target prompt in the target mask image based on the marking information of the winding package (i.e., the marking information of each winding package in the target mask image). For example, as shown in FIG. 3D , the identification module matches the mark 6 of the target winding package indicated by the target prompt with the marking information of each winding package in the target mask image of target image 1, and marks the mask area corresponding to the mark 6 in the target mask image. For example, the mask area corresponding to the mark 6 is highlighted in a target frame to obtain a target output image marking the location of the target winding package.

[0044] In this way, the solution disclosed in the present disclosure uses the target prompt and the target detection model to identify the yarn package for each target image, and obtain the target output image where the target yarn package indicated by the target prompt is located. In this way, the location information of the target yarn package can be quickly and accurately determined and obtained, realizing automatic positioning of the yarn package, further saving a lot of labor and time costs, and ensuring the production efficiency of the subsequent yarn package process.

[0045] Furthermore, in one example, the segmentation network module may be a Segment Anything Model (SAM) based on prior knowledge information, or may be other segmentation models with mask image generation capabilities, and the present disclosure is not limited thereto.

[0046] Furthermore, in one example, as shown in FIG. 4A, the segmentation network module includes at least a prior knowledge feature layer, a halftone dot segmentation layer, and an image segmentation layer.

[0047] Specifically, in one example, the prior knowledge feature layer is used to obtain target prior information based on a preset yarn package prompt and an input image, where the target prior knowledge information can be used to guide the image segmentation layer to identify the yarn package and segment and cut out the area where the yarn package is located, thereby enhancing the identification and segmentation capabilities of the image segmentation layer and generating a mask.

[0048] In another example, the halftone dot division layer can be used to divide a halftone dot presentation image to obtain a plurality of target sub-images with designated halftone dot positions, where the halftone dot positions in the plurality of target sub-images do not overlap, and the halftone dot presentation image is obtained by processing an input image using halftone dots. For example, as shown in FIG. 4B , an input image, such as target image 1 shown in FIG. 3B , is first processed using halftone dots to obtain a halftone dot presentation image corresponding to target image 1. The resulting halftone dot presentation image is then divided, for example, by rows, to obtain a plurality of target sub-images, where the halftone dots in each target sub-image do not overlap. This facilitates batch image processing of each target sub-image, effectively avoids repeated recognition, and lays the foundation for further improving the efficiency of wound yarn package identification.

[0049] In yet another example, the image segmentation layer is used to identify thread packages in each sub-image to be processed based on the target prior knowledge information and segment the area in which each thread package is located, and then mask the area in the sub-image to be processed where each thread package is located using a mask plate to obtain a sub-mask image of each sub-image to be processed, and after obtaining the sub-mask image of each sub-image to be processed, for example, to connect the sub-mask images of each sub-image to be processed based on the sub-mask image of each sub-image to be processed to obtain an initial mask image of the input image.

[0050] As a result, the solution disclosed herein can utilize target prior knowledge information to enhance the identification and segmentation capabilities of the image segmentation layer, and can realize batch image processing based on halftone dot presentation images, thereby improving segmentation efficiency, thereby laying the foundation for automatically locating and obtaining the required wound yarn package and for improving the production efficiency of the subsequent wound yarn package processing process.

[0051] In one embodiment of the present disclosure, the prior knowledge feature layer includes at least a semantic prior knowledge layer and a similarity map prior knowledge layer.

[0052] Here, in one example, the semantic prior knowledge layer is used to obtain semantic prior knowledge features based on at least a winding package feature corresponding to a preset winding package prompt, for example, in one example, a winding package feature corresponding to a preset winding package prompt can be directly used as a semantic prior knowledge feature.

[0053] In another example, the semantic prior knowledge layer may obtain semantic prior knowledge features as follows: Specifically, as shown in FIG. 5, the semantic prior knowledge layer may obtain the semantic prior knowledge features from the yarn package features corresponding to the preset yarn package prompt and the image features of the input image (e.g., the global feature map of the input image). fusion The semantic prior knowledge feature is then used to obtain a feature map (i.e., semantic prior knowledge feature) representing the semantic prior knowledge. For example, the thread package feature corresponding to the preset thread package prompt and the image feature of the input image are multiplied element by element to obtain a feature map for representing the semantic prior knowledge feature.

[0054] In this example, if the dimension of the winding package feature corresponding to the preset winding package prompt does not match the dimension of the image feature of the input image, an upsampling process (e.g., bilinear interpolation process) is performed on the winding package feature corresponding to the preset winding package prompt so that the dimension of the processed winding package feature becomes the same as the dimension of the image feature of the input image, and then both are fusion In this way, the feature information of the obtained semantic prior knowledge features will be enriched, laying the foundation for further enhancing the discrimination and segmentation ability of the image segmentation layer.

[0055] Furthermore, in another example, the similarity map prior knowledge layer is used to estimate the area in the input image where each winding package is located based on the similarity between the winding package features corresponding to a predetermined winding package prompt and the image features of the input image, thereby obtaining a target similarity map.

[0056] It should be noted that the above target prior knowledge information includes the semantic prior knowledge features and the target similarity map.

[0057] Furthermore, in one example, the similarity map prior knowledge layer may generate the target similarity map as follows: identification Specifically, the similarity map prior knowledge layer is specifically used to estimate the area in the input image where each winding yarn package exists based on the similarity between the obtained semantic prior knowledge features and the image features of the input image, and obtain the target similarity map, as shown in Figures 5 and 6, for example: performing an aggregation process on the obtained semantic prior knowledge features to obtain aggregated semantic prior knowledge features, and then, for example, adding (or averaging, etc.) pixel values ​​for each column of a feature vector representing the semantic prior knowledge features to obtain an aggregated feature vector; Obtaining multiple sub-feature vectors of the global feature map of the input image, for example, by dividing the feature vector representing the global feature map, for example by rows, to obtain multiple sub-feature vectors, where the dimension of each obtained sub-feature vector is the same as the dimension of the feature vector after the aggregation process, which is useful for calculating the similarity between the two; and The method is used to obtain a similarity between each sub-feature vector in the plurality of sub-feature vectors and the feature vector after the aggregation process, and to obtain a target similarity map based on the similarity.

[0058] This allows for precise identification of areas in the input image where the yarn packages are located, which then forms the basis for accurately identifying and separating the yarn packages in the image and obtaining a mask image.

[0059] Furthermore, in one example, the prior knowledge feature layer can include a feature enhancement layer, for example, to perform feature enhancement on the target similarity map, the obtained target similarity map can be input into the feature enhancement layer to obtain a feature-enhanced target similarity map, in which case the image segmentation layer can specifically perform classification and segmentation based on the semantic prior knowledge features and the feature-enhanced target similarity map, thereby further improving the accuracy of image classification and segmentation.

[0060] Alternatively, in another example, the prior knowledge feature layer can include a labeling layer. For example, as shown in FIG. 5, the obtained target similarity map (or the feature-enhanced target similarity map) is input into the labeling layer, and a labeling process is performed on the input target similarity map to obtain a label feature map. In this case, the target prior knowledge information specifically includes the semantic prior knowledge features and the label feature map.

[0061] Here, in the label feature map (for example, using "0" and "1" as labeling), if the value of a certain area is 1, it indicates that the area is a positive area, i.e., that a yarn package or a part of a yarn package is present; otherwise, it indicates that the area is a negative area, which helps the image segmentation layer to ignore the negative areas and focus on segmenting the positive areas, thereby further enhancing the ability to identify and segment the yarn package, which in turn helps to more accurately identify and separate the yarn package in the image, and can effectively improve the efficiency of identification and segmentation.

[0062] It should be noted that the above-mentioned prior knowledge feature layer may include a feature enhancement layer, or a labeling layer, or a feature enhancement layer and a labeling layer, or may include other processing layers for improving image recognition and segmentation capabilities. In actual applications, this can be set according to actual needs, and the present disclosure is not specifically limited thereto.

[0063] In this way, the solution disclosed herein utilizes the semantic prior knowledge features and the target similarity map in the target prior knowledge information, allowing the image segmentation layer to focus more on identifying and segmenting the wound yarn packages in the image, thereby improving the image identification and segmentation capabilities of the image segmentation layer, and thereby effectively improving the identification accuracy and identification efficiency.

[0064] The solution of the present disclosure also provides a detection device applied to cloud, as shown in FIG. 7, comprising: an information acquiring unit 701; and a detecting unit 702; the information acquisition unit 701 is used for acquiring a target prompt for instructing the user to search for a target wound thread package; and acquiring a plurality of target images of a target wound thread package cart where the target wound thread package indicated by the target prompt is located, the target wound thread package cart including two placement areas for placing wound thread packages, at least one first image of the plurality of target images including all wound thread packages placed by the first placement area of ​​the two placement areas, and at least one second image of the plurality of target images including all wound thread packages placed by the second placement area of ​​the two placement areas; The detection unit 702 inputs a plurality of target images and the target prompt into a target detection model to obtain a target output image indicating a location where a target winding package is located, and the target detection model identifies a winding package for the input image based on the target winding package indicated by the target prompt, and outputs an image indicating the location where the target winding package is located. obtain , especially used.

[0065] In one specific example of the solution of the present disclosure, the information acquiring unit specifically comprises: The target yarn package indicated by the target prompt is identified by the tag information. identification To do, Based on the mark information of the target wound yarn package, mark information of the target wound yarn package carriage where the target wound yarn package is located is obtained. identificationTo do, Based on the sign information of the target wound yarn package cart, a storage area where the target wound yarn package cart is located is determined. identification To do, and selecting the plurality of target images from image data collected by an image collecting device in a storage area where the target wound yarn package cart is located.

[0066] In one embodiment of the solution of the present disclosure, the target detection model includes at least a segmentation network module, an information mapping module, and an identification module; The segmentation network module is used to identify an area in the input image where each winding package is located based on a preset winding package prompt, and use a mask to mask the area in the image where each winding package is located to obtain an initial mask image of the input image, where the input image is one of a plurality of target images; the information mapping module is used to perform information mapping processing on the initial mask image according to the target prompt to obtain a target mask image with label information of the yarn package corresponding to the input image, wherein the target mask image with label information of the yarn package corresponding to the input image can represent label information of each yarn package actually included in the input image; The identification module is used to identify and mark the target winding package indicated by the target prompt in the target mask image based on the marking information of the winding package, and obtain a target output image.

[0067] In one embodiment of the solution of the present disclosure, the information mapping module specifically comprises: based on the mark information of the target wound yarn package carriage, obtain mark information of the wound yarn package to be placed by each placement body of the target wound yarn package carriage, and map the mark information of the wound yarn package to be placed by each placement body of the target wound yarn package carriage to each mask plate located at different positions in the initial mask image to obtain a target mask image with mark information of the wound yarn package corresponding to the input image; Here, the identification information of the target wound yarn package carriage is obtained based on the identification information of the target wound yarn package indicated by the target prompt.

[0068] In one embodiment of the solution of the present disclosure, the segmentation network module includes at least a prior knowledge feature layer, a halftone dot segmentation layer and an image segmentation layer; The prior knowledge feature layer is used to obtain target prior knowledge information based on a preset winding package prompt and an input image, which is one of a plurality of target images; the halftone dot division layer is used to divide a halftone dot presentation image obtained by processing an input image using halftone dots, to obtain a plurality of processing target sub-images having designated halftone dot positions, the halftone dot positions in different processing target sub-images not overlapping; The image segmentation layer is used to identify the yarn packages in each sub-image to be processed based on the target prior knowledge information, mask the areas in the sub-image to be processed where each yarn package is located using a mask plate to obtain sub-mask images of each sub-image to be processed, and obtain an initial mask image of the input image based on the sub-mask images of each sub-image to be processed.

[0069] In one embodiment of the solution of the present disclosure, the prior knowledge feature layer includes at least a semantics prior knowledge layer and a similarity map prior knowledge layer; The semantic prior knowledge layer is used to obtain semantic prior knowledge features based on at least a winding package feature corresponding to a preset winding package prompt; The similarity map prior knowledge layer is used to estimate an area in the input image where each winding package is located based on a similarity between a winding package feature corresponding to a preset winding package prompt and an image feature of the input image, thereby obtaining a target similarity map; The target prior knowledge information includes the semantic prior knowledge features and the target similarity map.

[0070] In one embodiment of the solution of the present disclosure, the semantic prior knowledge layer specifically includes: a winding package characteristic corresponding to the preset winding package prompt, and an image characteristic of the input image. Combined with This is used to obtain semantic prior knowledge features.

[0071] In one embodiment of the solution of the present disclosure, the similarity map prior knowledge layer specifically includes: Based on the similarity between the obtained semantic prior knowledge features and the image features of the input image, the area in the input image where each wound yarn package exists is estimated, and used to obtain a target similarity map.

[0072] For the specific functions and exemplary descriptions of each module of the apparatus according to the embodiments of the present disclosure, please refer to the relevant descriptions of the corresponding steps in the above-mentioned method embodiments, and they will not be repeated here.

[0073] In the technical solution of the present disclosure, the acquisition, storage, and application of users' personal information comply with the provisions of relevant laws and regulations and do not violate public order and morals.

[0074] FIG. 8 is a structural block diagram of an electronic device according to an embodiment of the present disclosure. As shown in FIG. 8, the electronic device includes a memory 810 and a processor 820, and the memory 810 stores a computer program executable by the processor 820. The number of memories 810 and processors 820 may be one or more. The memory 810 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 embodiments. The electronic device may further include: a communication interface 830 for communicating with external devices and for data interaction and transmission;

[0075] When the memory 810, the processor 820, and the communication interface 830 are implemented independently, the memory 810, the processor 820, and the communication interface 830 are connected to each other via a bus to enable communication between them. The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The bus may be classified into an address bus, a data bus, a control bus, and the like. For ease of explanation, only one bold line is shown in FIG. 8, but this does not represent only one bus or only one type of bus.

[0076] Optionally, in a specific implementation, when the memory 810, the processor 820, and the communication interface 830 are integrated on one chip, the memory 810, the processor 820, and the communication interface 830 can communicate with each other via an internal interface.

[0077] 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. The processor may be a processor supporting the Advanced RISC Machines (ARM) architecture.

[0078] Furthermore, 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 memory or non-volatile memory, or may include both volatile and non-volatile memory. Here, non-volatile memory includes ROM (Read-Only Memory), PROM (Programmable ROM), EPROM (Erasable ROM), etc. Volatile memory can include electrically volatile PROM (Programmable ROM), electrically EEPROM (Electrically EPROM), or flash memory. Volatile memory can include random access memory (RAM), which acts as an external cache. By way of example and not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchlink DRAM (SLDRAM), and direct RAMBUS RAM (DR RAM).

[0079] The above-described embodiments may be implemented, in whole or in part, in software, hardware, firmware, or any combination thereof. When implemented in software, they may be implemented, in whole or in part, in the form of a computer program product. A 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 an embodiment 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.

[0080] 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.

[0081] In describing embodiments of the present disclosure, the references "one embodiment," "some embodiments," "examples," "specific examples," or "some 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. Furthermore, the described specific features, structures, materials, or characteristics may 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 and features of different embodiments or examples described in the present disclosure to the extent that they are not inconsistent with each other.

[0082] In the description of the embodiments of the present disclosure, unless otherwise specified, " / " means "or," for example, A / B can mean either A or B. In the present disclosure, "and / or" merely describes 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 situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0083] In describing the embodiments of the present disclosure, the terms "first" and "second" are used for descriptive purposes only and should not be interpreted as indicating or implying relative importance, nor should they be interpreted as implying the number of technical features shown. Thus, features defined as "first" and "second" may explicitly or implicitly include one or more of such features. In describing the embodiments of the present disclosure, "plurality" means two or more, unless otherwise specified.

[0084] The above are merely illustrative examples of the present disclosure, and do not limit the present disclosure. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present disclosure should be included in the scope of protection of the present disclosure.

Claims

1. A detection method applied to a cloud terminal, comprising: obtaining a target prompt to instruct the user to locate a target spool package; acquiring a plurality of target images of a target winding package cart on which a target winding package indicated by the target prompt is located, the target winding package cart including two placement areas for placing winding packages, at least one first image among the plurality of target images including all winding packages placed by the first placement area of ​​the two placement areas, and at least one second image among the plurality of target images including all winding packages placed by the second placement area of ​​the two placement areas; inputting a plurality of target images and the target prompt into a target detection model to obtain a target output image indicating a location where a target winding package is located, wherein the target detection model identifies a winding package for the input image based on the target winding package indicated by the target prompt, wherein the input image is one of the plurality of target images, and obtaining the target output image indicating a location where the target winding package is located. Detection method.

2. acquiring a plurality of target images of a target winding package carriage on which a target winding package indicated by the target prompt is located; Identifying tag information for the target wound yarn package indicated by the target prompt; Identifying the mark information of a target wound yarn package carriage on which the target wound yarn package is located based on the mark information of the target wound yarn package; Identifying a storage area where the target wound yarn package cart is located based on the label information of the target wound yarn package cart; and selecting the plurality of target images from image data collected by an image collecting device in a storage area where the target wound yarn package cart is located. The detection method according to claim 1 .

3. The target detection model includes at least a segmentation network module, an information mapping module, and an identification module; The segmentation network module is used to identify an area in the input image where each winding package is located based on a preset winding package prompt, and to mask the area in the image where each winding package is located using a masking plate to obtain an initial mask image of the input image; the information mapping module is used to perform information mapping processing on the initial mask image according to the target prompt to obtain a target mask image with label information of the yarn package corresponding to the input image, wherein the target mask image with label information of the yarn package corresponding to the input image can represent label information of each yarn package actually included in the input image; the identification module is used to identify and mark the target winding package indicated by the target prompt in the target mask image according to the marking information of the winding package, and obtain a target output image; The detection method according to claim 1 .

4. The information mapping module specifically: based on the mark information of the target wound yarn package carriage, obtain mark information of the wound yarn package to be placed by each placement body of the target wound yarn package carriage, and map the mark information of the wound yarn package to be placed by each placement body of the target wound yarn package carriage to each mask plate located at different positions in the initial mask image to obtain a target mask image with mark information of the wound yarn package corresponding to the input image; Here, the identification information of the target wound yarn package carriage is obtained based on identification information of the target wound yarn package indicated by the target prompt. The detection method according to claim 3 .

5. The segmentation network module includes at least a prior knowledge feature layer, a halftone dot segmentation layer, and an image segmentation layer; The prior knowledge feature layer is used to obtain target prior knowledge information based on a preset winding package prompt and an input image, which is one of a plurality of target images; the halftone dot division layer is used to divide a halftone dot presentation image obtained by processing an input image using halftone dots, to obtain a plurality of processing target sub-images having designated halftone dot positions, the halftone dot positions in different processing target sub-images not overlapping; the image segmentation layer is used to identify a yarn package in each of the target sub-images based on the target prior knowledge information, mask an area in the target sub-image where each yarn package is located using a masking plate to obtain a sub-mask image of each of the target sub-images, and obtain an initial mask image of the input image based on the sub-mask image of each of the target sub-images; The detection method according to claim 3 .

6. the prior knowledge feature layer includes at least a semantics prior knowledge layer and a similarity map prior knowledge layer; The semantic prior knowledge layer is used to obtain semantic prior knowledge features based on at least a winding package feature corresponding to a preset winding package prompt; The similarity map prior knowledge layer is used to estimate an area in the input image where each winding package is located based on a similarity between a winding package feature corresponding to a preset winding package prompt and an image feature of the input image, thereby obtaining a target similarity map; the target prior knowledge information includes the semantic prior knowledge features and the target similarity map; The detection method according to claim 5 .

7. Specifically, the semantics prior knowledge layer: The yarn package features corresponding to the preset yarn package prompt are used to fuse with the image features of the input image to obtain semantic prior knowledge features. The detection method according to claim 6.

8. Specifically, the similarity map prior knowledge layer includes: Based on the similarity between the obtained semantic prior knowledge features and the image features of the input image, the regions in the input image where each wound yarn package exists are estimated to obtain a target similarity map. The detection method according to claim 6.

9. A detection device applied to a cloud terminal, comprising: an information acquisition unit; and a detection unit; the information acquisition unit is used to acquire a target prompt for instructing the user to search for a target wound thread package; and acquire a plurality of target images of a target winding thread package cart where the target winding thread package indicated by the target prompt is located, the target winding thread package cart including two placement areas for placing winding thread packages, at least one first image among the plurality of target images including all winding thread packages placed by the first placement area of ​​the two placement areas, and at least one second image among the plurality of target images including all winding thread packages placed by the second placement area of ​​the two placement areas; the detection unit is used to input a plurality of target images and the target prompt into a target detection model to obtain a target output image indicating a location where a target winding package is located, the target detection model identifying a winding package for the input image based on the target winding package indicated by the target prompt, wherein the input image is one of the plurality of target images, and the target output image indicating a location where the target winding package is located is obtained. Detection device.

10. at least one processor; a memory communicatively coupled to 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 the method of any one of claims 1 to 8. Electronic devices.

11. A non-transitory computer readable storage medium for storing instructions that cause a computer to perform the method of any one of claims 1 to 8.

12. A program for implementing the method of any one of claims 1 to 8 when executed by a processor in a computer.

Images