Logistics packaging recycling management method, device, equipment and readable storage medium

Abstract

The present invention provides a logistics package recycling management method, device, equipment and readable storage medium, the method comprising: acquiring an image of each logistics package within a preset time period; extracting characteristic information on the logistics package according to the image of the logistics package to obtain characteristic information; dividing all the logistics package images into multiple groups according to the characteristic information, adjusting the logistics package images in each group so that the number of logistics package images in each group is consistent, obtaining a grouping result, and classifying and recycling all the logistics packages according to the grouping result. The method of the present invention can complete the grouping of all the logistics packages according to the characteristics of each logistics package, which is conducive to the subsequent recycling of the logistics packages.

Description

Technical Field

[0001] The present invention relates to the field of logistics packaging technology, and in particular to a logistics packaging recycling management method, device, equipment and readable storage medium. Background Art

[0002] At present, people are paying more and more attention to environmental protection issues. As the transportation industry continues to develop, more and more logistics packaging is generated. How to deal with and recycle these logistics packaging is very important. Summary of the invention

[0003] The purpose of the present invention is to provide a logistics packaging recycling management method, device, equipment and readable storage medium to improve the above-mentioned problems.

[0004] In order to achieve the above objectives, the present application provides the following technical solutions:

[0005] In a first aspect, an embodiment of the present application provides a logistics packaging recycling management method, the method comprising:

[0006] Acquire images of each logistics package within a preset time period;

[0007] Performing image enhancement processing on the image of the logistics package to obtain a logistics package processed image; acquiring multiple historical logistics package images, performing saliency detection on the historical logistics package images using a visual attention mechanism, detecting salient objects on the historical logistics package images, and segmenting the salient objects from the historical logistics package images to obtain a first image; extracting a first feature of the first image, and performing cluster analysis on all of the first images according to the first feature to obtain a clustering result, screening the first images according to the clustering result and calculating the annotation information corresponding to the remaining first images; training a preset first model using the remaining first images and the corresponding annotation information to obtain a feature extraction model;

[0008] Based on the logistics packaging processing image and the feature extraction model, feature information corresponding to the logistics packaging processing image is obtained; according to the feature information, all the logistics packaging images are divided into multiple groups, the logistics packaging images in each group are adjusted to make the number of logistics packaging images in each group consistent, to obtain a grouping result, and all the logistics packaging are classified and recycled according to the grouping result.

[0009] Optionally, extracting a first feature of the first image, performing cluster analysis on all the first images according to the first feature to obtain a clustering result, screening the first images according to the clustering result and calculating annotation information corresponding to the remaining first images, includes:

[0010] Using a directional gradient histogram feature extraction model to extract first features from the first image; using a K-MEDOIDS algorithm based on all the extracted first features to perform cluster analysis on all the first images to obtain clustering results;

[0011] In each of the clustering results, a preset number of first images are selected according to the distance from the cluster center for grouping, so as to obtain a first set corresponding to each clustering result; each first set is analyzed, wherein the number of first images containing special information in each of the first sets is obtained, and when the number is greater than a preset value, the first set is retained, and when the number is less than or equal to the preset value, the first set is deleted, and each retained first set is recorded as a second set, wherein the special information is text information or pattern information; labeling information for each of the second sets is obtained, and the labeling information is used as the labeling information of all first images in the clustering result corresponding to the second set, so as to obtain the first image and the corresponding labeling information.

[0012] Optionally, the preset first model is trained using the remaining first images and the corresponding annotation information to obtain a feature extraction model, including:

[0013] In the remaining first images, the first model is trained using the first image and the corresponding annotation information to obtain a second model, and the first image is input into the second model to obtain the first annotation information corresponding to the first image, wherein the first model is a model trained by the first training set, the function of the first model is to identify the number of people in the picture, and the number of samples in the first training set is greater than the number of first images with annotation information; the convolutional neural network is trained for multiple rounds using the first image and the corresponding annotation information and the first annotation information to obtain a feature extraction model, wherein in the first N rounds of training, the value of the first loss function between the first annotation information of the first image and the second annotation information corresponding to the first image output by the convolutional neural network is used as the target of training the convolutional neural network, the first loss function is a logarithmic loss function, and N is a positive integer greater than 2; in the remaining rounds of training, the value of the second loss function between the annotation information of the first image and the second annotation information corresponding to the first image output by the convolutional neural network is used as the target of training the convolutional neural network, and the second loss function is a cross entropy loss function.

[0014] Optionally, all the images of the logistics packages are divided into multiple groups according to the characteristic information, the images of the logistics packages in each group are adjusted so that the number of the images of the logistics packages in each group is consistent, a grouping result is obtained, and all the logistics packages are classified and recycled according to the grouping result, including:

[0015] The characteristic information corresponding to all the images of the logistics packages are clustered to obtain a plurality of clusters; all the clusters are sorted, wherein the distance between each cluster and the characteristic information corresponding to each logistics package image is calculated, and all the distances are summed up to obtain the total distance corresponding to each cluster, and the larger the total distance corresponding to the cluster, the higher the ranking; after sorting, each cluster is processed in turn, wherein it is determined whether the number of characteristic information corresponding to the images of the logistics packages contained in such cluster is less than a preset value, if so, the characteristic information corresponding to the images of the logistics packages closest to such cluster is moved into such cluster, and when the number of characteristic information corresponding to the images of the logistics packages contained in such cluster is the preset value, the moving in is stopped; at the same time, it is determined whether the number of characteristic information corresponding to the images of the logistics packages contained in such cluster is greater than the preset value, if so, the characteristic information corresponding to the image of the logistics package farthest from the centroid in such cluster is moved out of such cluster, and when the number of characteristic information corresponding to the images of the logistics packages contained in such cluster is the preset value, the moving out is stopped, wherein the total number of images of the logistics packages can be evenly divided.

[0016] In a second aspect, an embodiment of the present application provides a logistics packaging recycling management device, including:

[0017] An acquisition module, used to acquire an image of each logistics package within a preset time period;

[0018] The extraction module is used to perform image enhancement processing on the image of the logistics package to obtain a logistics package processed image; obtain multiple historical logistics package images, use a visual attention mechanism to perform saliency detection on the historical logistics package images, detect salient objects on the historical logistics package images, and segment the salient objects from the historical logistics package images to obtain a first image; extract a first feature of the first image, and perform cluster analysis on all the first images according to the first feature to obtain a clustering result, screen the first images according to the clustering result and calculate the annotation information corresponding to the remaining first images; use the remaining first images and the corresponding annotation information to train a preset first model to obtain a feature extraction model;

[0019] A grouping module is used to obtain feature information corresponding to the logistics packaging processing image based on the logistics packaging processing image and the feature extraction model; divide all the logistics packaging images into multiple groups according to the feature information, adjust the logistics packaging images in each group so that the number of logistics packaging images in each group is consistent, obtain grouping results, and classify and recycle all the logistics packaging according to the grouping results.

[0020] In a third aspect, an embodiment of the present application provides a logistics packaging recycling management device, the device comprising a memory and a processor. The memory is used to store a computer program; the processor is used to implement the steps of the above-mentioned logistics packaging recycling management method when executing the computer program.

[0021] In a fourth aspect, an embodiment of the present application provides a readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned logistics packaging recycling management method are implemented.

[0022] The beneficial effects of the present invention are:

[0023] 1. The present invention obtains an image of a logistics package and multiple images of historical logistics packages, trains a feature extraction model based on the images of the multiple historical logistics packages, obtains feature information corresponding to the image of the logistics package based on the feature extraction model, and classifies the image of the logistics package into multiple groups based on the feature information. After grouping, considering that all logistics packages may need to be grouped for transportation and processing in the future, it is best to keep the number of each group consistent. Therefore, grouping is finally performed and the quantity is adjusted to make the number of each group consistent. The method in the present invention can complete the grouping of all logistics packages according to the characteristics of each logistics package, which is conducive to the subsequent recycling and processing.

[0024] 2. In the present invention, considering that some of the extracted first images may not be very relevant to the classification of logistics packaging, it is not very meaningful to use these first images to train the feature extraction model, and it will also increase the training time, which is not conducive to improving the training efficiency. Therefore, this application analyzes each clustering result and screens all the first images according to the analysis results; at the same time, when performing information annotation, information annotation is not performed on each first image, but on each clustering result. In this way, the annotation work can be completed quickly, which saves a lot of annotation time for the staff.

[0025] 3. During training, considering the efficiency of training, a relatively mature first model is used for preliminary training, wherein the first model is a pre-trained model with a mature network structure and network parameters; in this way, the training efficiency in the subsequent training process can be improved.

[0026] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or be understood by implementing the embodiments of the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0028] Figure 1 It is a schematic diagram of the flow chart of the logistics packaging recycling management method described in an embodiment of the present invention;

[0029] Figure 2 It is a structural schematic diagram of the logistics packaging recycling management device described in an embodiment of the present invention;

[0030] Figure 3 It is a schematic diagram of the structure of the logistics packaging recycling management equipment described in an embodiment of the present invention. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0032] It should be noted that similar reference numerals or letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings. At the same time, in the description of the present invention, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0033] Example 1

[0034] like Figure 1 As shown, this embodiment provides a logistics packaging recycling management method, which includes step S1, step S2 and step S3.

[0035] Step S1, acquiring an image of each logistics package within a preset time period;

[0036] In this step, the preset time period can be from the past month to the present, or from the past week to the present, and can be customized according to the needs of the user; the image of each logistics package can be acquired by the image acquisition device;

[0037] Step S2, performing image enhancement processing on the image of the logistics package to obtain a logistics package processed image; acquiring multiple historical logistics package images, performing saliency detection on the historical logistics package images using a visual attention mechanism, detecting salient objects on the historical logistics package images, and segmenting the salient objects from the historical logistics package images to obtain a first image; extracting a first feature of the first image, and performing cluster analysis on all of the first images according to the first feature to obtain a clustering result, screening the first images according to the clustering result and calculating the annotation information corresponding to the remaining first images; training a preset first model using the remaining first images and the corresponding annotation information to obtain a feature extraction model;

[0038] In this step, considering that the logistics packaging may be damaged in the previous transportation, the quality of the captured image may not be very good. Therefore, in this step, image enhancement processing is performed first, and then feature information is extracted;

[0039] In this step, since the logistics packaging is to be recycled, considering that there may be some special information on the logistics packaging, such as some texts and some patterns, information about the logistics packaging can be extracted from these special information, so this step extracts these special information to obtain feature information;

[0040] In this step, considering that there is special information on general logistics packaging, which is quite different from the surrounding color or structure, the visual attention mechanism is used to perform saliency detection on the image of the historical logistics packaging, so that the special information can be quickly located, which is convenient for the next step of feature information extraction; at the same time, when segmenting, the minimum rectangle can be determined according to the coordinate information of the salient object, and the area formed by the minimum rectangle can be segmented to obtain the first image;

[0041] In this step, the first feature of the first image is extracted, and cluster analysis is performed on all the first images according to the first feature to obtain a clustering result. The first images are screened according to the clustering result and the annotation information corresponding to the remaining first images is calculated. The specific implementation steps include step S21 and step S22;

[0042] Step S21, using a directional gradient histogram feature extraction model to extract first features from the first image; based on all the extracted first features, using a K-MEDOIDS algorithm to perform cluster analysis on all the first images to obtain a clustering result;

[0043] Step S22. In each of the clustering results, a preset number of first images are selected according to the distance from the cluster center for grouping to obtain a first set corresponding to each clustering result; each first set is analyzed, wherein the number of first images containing special information in each of the first sets is obtained, and when the number is greater than a preset value, the first set is retained, and when the number is less than or equal to the preset value, the first set is deleted, and each retained first set is recorded as a second set, wherein the special information is text information or pattern information; labeling information for each of the second sets is obtained, and the labeling information is used as the labeling information of all first images in the clustering result corresponding to the second set to obtain the first image and the corresponding labeling information.

[0044] In this step, considering that in the process of saliency detection, some objects extracted from logistics images may not be related to the classification of logistics packaging, then using these first images to train the model is not very meaningful, and it will also increase the training time, which is not conducive to improving efficiency. Therefore, this application analyzes each clustering result; In addition, in this step, the preset number and preset value can be customized;

[0045] In this step, the annotation information of each second set is one annotation information, and the staff can perform a comprehensive analysis based on the first image corresponding to each second set to obtain one annotation information, wherein the annotation information can be understood as, for example, packaging for transporting dangerous goods, packaging for transporting fragile goods, foam packaging, etc.; at the same time, after annotation, the annotation information is directly used as the annotation information of all the first images in the corresponding clustering results, which can save a lot of annotation time for the staff, which is conducive to improving the efficiency of training;

[0046] In this step, the preset first model is trained using the remaining first images and the corresponding annotation information to obtain the feature extraction model. The specific implementation steps include step S23;

[0047] Step S23: In the remaining first images, the first model is trained using the first image and the corresponding annotation information to obtain a second model, and the first image is input into the second model to obtain the first annotation information corresponding to the first image, wherein the first model is a model trained by the first training set, the function of the first model is to identify the number of people in the picture, and the number of samples in the first training set is greater than the number of first images with annotation information; the convolutional neural network is trained for multiple rounds using the first image and the corresponding annotation information and the first annotation information to obtain a feature extraction model, wherein in the first N rounds of training, the value of the first loss function between the first annotation information of the first image and the second annotation information corresponding to the first image output by the convolutional neural network is used as the target of training the convolutional neural network, the first loss function is a logarithmic loss function, and N is a positive integer greater than 2; in the remaining rounds of training, the value of the second loss function between the annotation information of the first image and the second annotation information corresponding to the first image output by the convolutional neural network is used as the target of training the convolutional neural network, and the second loss function is a cross entropy loss function.

[0048] In this step, the first model is a pre-trained model, in this way the first model has a mature network structure and network parameters, based on this situation, the training efficiency in the subsequent training process can be improved;

[0049] Step S3: based on the logistics packaging processing image and the feature extraction model, obtain feature information corresponding to the logistics packaging processing image; divide all logistics packaging images into multiple groups according to the feature information, adjust the logistics packaging images in each group so that the number of logistics packaging images in each group is consistent, obtain grouping results, and classify and recycle all logistics packaging according to the grouping results.

[0050] In this step, the logistics packaging processing image is input into the feature extraction model to obtain the feature information.

[0051] In this step, considering that all logistics packages may be recycled later, such as group transportation and other recycling processes, then considering the efficiency of transportation, the logistics packages transported by each vehicle are preferably the same, which can improve the efficiency of transportation. Therefore, this step adjusts the logistics package images in each group according to the clustering results, and after the adjustment, the number of logistics packages corresponding to each group is consistent, which is convenient for transportation. The specific implementation steps include step S31;

[0052] Step S31: clustering the feature information corresponding to all the images of logistics packages to obtain multiple clusters; sorting all the clusters, wherein the distance between each cluster and the feature information corresponding to each logistics package image is calculated, and all the distances are summed up to obtain the total distance corresponding to each cluster, and the larger the total distance corresponding to the cluster, the higher the ranking; after sorting, processing each cluster in turn, wherein it is determined whether the number of feature information corresponding to the images of logistics packages contained in such cluster is less than a preset value, if so, the feature information corresponding to the image of the logistics package closest to such cluster is moved into such cluster, and when the number of feature information corresponding to the image of the logistics package contained in such cluster is the preset value, the moving in is stopped; at the same time, it is determined whether the number of feature information corresponding to the images of logistics packages contained in such cluster is greater than the preset value, if so, the feature information corresponding to the image of the logistics package farthest from the centroid in such cluster is moved out of such cluster, and when the number of feature information corresponding to the images of logistics packages contained in such cluster is the preset value, the moving out is stopped, wherein the total number of images of logistics packages can be evenly divided.

[0053] In this embodiment, an image of a logistics package and multiple images of historical logistics packages are first obtained, a feature extraction model is trained based on the images of the multiple historical logistics packages, feature information corresponding to the image of the logistics package is obtained based on the feature extraction model, and the image of the logistics package is classified into multiple groups based on the feature information. After grouping, considering that all logistics packages may need to be grouped for transportation and processing in the future, the number of each group is preferably kept consistent. Therefore, grouping is finally performed and the number is adjusted to make the number of each group consistent. The method in the present invention can complete the grouping of all logistics packages according to the characteristics of each logistics package, which is conducive to subsequent recycling and processing.

[0054] Example 2

[0055] like Figure 2 As shown, this embodiment provides a logistics packaging recycling management device, which includes an acquisition module 1, an extraction module 2 and a grouping module 3.

[0056] An acquisition module 1 is used to acquire an image of each logistics package within a preset time period;

[0057] Extraction module 2 is used to perform image enhancement processing on the image of the logistics package to obtain a logistics package processed image; obtain multiple historical logistics package images, use a visual attention mechanism to perform saliency detection on the historical logistics package images, detect salient objects on the historical logistics package images, and segment the salient objects from the historical logistics package images to obtain a first image; extract a first feature of the first image, and perform cluster analysis on all the first images according to the first feature to obtain a clustering result, screen the first images according to the clustering result and calculate the annotation information corresponding to the remaining first images; use the remaining first images and the corresponding annotation information to train a preset first model to obtain a feature extraction model;

[0058] The grouping module 3 is used to obtain the feature information corresponding to the logistics packaging processing image based on the logistics packaging processing image and the feature extraction model; divide all the logistics packaging images into multiple groups according to the feature information, adjust the logistics packaging images in each group so that the number of logistics packaging images in each group is consistent, obtain the grouping result, and classify and recycle all the logistics packaging according to the grouping result.

[0059] It should be noted that, regarding the device in the above embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment of the method, and will not be elaborated here.

[0060] Example 3

[0061] Corresponding to the above method embodiment, the embodiment of the present disclosure also provides a logistics packaging recycling management device. The logistics packaging recycling management device described below and the logistics packaging recycling management method described above can be referenced to each other.

[0062] Figure 3 FIG. 3 is a block diagram of a logistics packaging recycling management device 300 according to an exemplary embodiment. Figure 3 As shown, the logistics package recycling management device 300 may include: a processor 301, a memory 302. The logistics package recycling management device 300 may also include one or more of a multimedia component 303, an I / O interface 304, and a communication component 305.

[0063] The processor 301 is used to control the overall operation of the logistics packaging recycling management device 300 to complete all or part of the steps in the above-mentioned logistics packaging recycling management method. The memory 302 is used to store various types of data to support the operation of the logistics packaging recycling management device 300. These data may include, for example, instructions for any application or method operating on the logistics packaging recycling management device 300, and application-related data, such as contact data, sent and received messages, pictures, audio, video, etc. The memory 302 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, disk or optical disk. The multimedia component 303 may include a screen and an audio component. The screen may be, for example, a touch screen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signal may be further stored in the memory 302 or sent via the communication component 305. The audio component also includes at least one speaker for outputting audio signals. The I / O interface 304 provides an interface between the processor 301 and other interface modules, and the above-mentioned other interface modules can be keyboards, mice, buttons, etc. These buttons can be virtual buttons or physical buttons. The communication component 305 is used for wired or wireless communication between the logistics packaging recycling management device 300 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G or 4G, or a combination of one or more of them, so the corresponding communication component 305 can include: Wi-Fi module, Bluetooth module, NFC module.

[0064] In an exemplary embodiment, the logistics packaging recycling management device 300 can be implemented by one or more application specific integrated circuits (Application Specific Integrated Circuit, referred to as ASIC), digital signal processors (Digital Signal Processor, referred to as DSP), digital signal processing devices (Digital Signal Processing Device, referred to as DSPD), programmable logic devices (Programmable Logic Device, referred to as PLD), field programmable gate arrays (Field Programmable Gate Array, referred to as FPGA), controllers, microcontrollers, microprocessors or other electronic components to execute the above-mentioned logistics packaging recycling management method.

[0065] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided, and when the program instructions are executed by a processor, the steps of the above-mentioned logistics packaging recycling management method are implemented. For example, the computer-readable storage medium can be the above-mentioned memory 302 including program instructions, and the above-mentioned program instructions can be executed by the processor 301 of the logistics packaging recycling management device 300 to complete the above-mentioned logistics packaging recycling management method.

[0066] Example 4

[0067] Corresponding to the above method embodiment, the embodiment of the present disclosure further provides a readable storage medium. The readable storage medium described below and the logistics packaging recycling management method described above can be referenced to each other.

[0068] A readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the logistics packaging recycling management method of the above method embodiment.

[0069] The readable storage medium may specifically be a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, or other readable storage medium that can store program codes.

[0070] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. Logistics packaging recycling management method, characterized in that: include: Acquire images of each logistics package within a preset time period; Performing image enhancement processing on the image of the logistics package to obtain a logistics package processed image; Acquire multiple images of historical logistics packages, perform saliency detection on the images of the historical logistics packages using a visual attention mechanism, detect salient objects on the images of the historical logistics packages, and segment the salient objects from the images of the historical logistics packages to obtain a first image; Extracting a first feature of the first image, performing cluster analysis on all the first images according to the first feature to obtain a clustering result, screening the first images according to the clustering result and calculating annotation information corresponding to the remaining first images; Using the remaining first images and corresponding annotation information to train the preset first model to obtain a feature extraction model; Based on the logistics package processing image and the feature extraction model, feature information corresponding to the logistics package processing image is obtained; all the logistics package images are divided into multiple groups according to the feature information, the logistics package images in each group are adjusted so that the number of logistics package images in each group is consistent, a grouping result is obtained, and all the logistics packages are classified and recycled according to the grouping result; The method further comprises: dividing all the images of the logistics packages into a plurality of groups according to the characteristic information, adjusting the images of the logistics packages in each group so that the number of the images of the logistics packages in each group is consistent, obtaining a grouping result, and classifying and recycling all the logistics packages according to the grouping result, including: The characteristic information corresponding to all the images of the logistics packages are clustered to obtain a plurality of clusters; all the clusters are sorted, wherein the distance between each cluster and the characteristic information corresponding to each logistics package image is calculated, and all the distances are summed up to obtain the total distance corresponding to each cluster, and the larger the total distance corresponding to the cluster, the higher the ranking; after sorting, each cluster is processed in turn, wherein it is determined whether the number of characteristic information corresponding to the images of the logistics packages contained in such cluster is less than a preset value, if so, the characteristic information corresponding to the images of the logistics packages closest to such cluster is moved into such cluster, and when the number of characteristic information corresponding to the images of the logistics packages contained in such cluster is the preset value, the moving in is stopped; at the same time, it is determined whether the number of characteristic information corresponding to the images of the logistics packages contained in such cluster is greater than the preset value, if so, the characteristic information corresponding to the image of the logistics package farthest from the centroid in such cluster is moved out of such cluster, and when the number of characteristic information corresponding to the images of the logistics packages contained in such cluster is the preset value, the moving out is stopped, wherein the total number of images of the logistics packages can be evenly divided.

2. The logistics packaging recycling management method according to claim 1 is characterized in that: Extracting a first feature of the first image, performing cluster analysis on all the first images according to the first feature to obtain a clustering result, screening the first images according to the clustering result and calculating annotation information corresponding to the remaining first images, including: Using a directional gradient histogram feature extraction model to extract first features from the first image; using a K-MEDOIDS algorithm based on all the extracted first features to perform cluster analysis on all the first images to obtain clustering results; In each of the clustering results, a preset number of first images are selected according to the distance from the cluster center for grouping, so as to obtain a first set corresponding to each clustering result; each first set is analyzed, wherein the number of first images containing special information in each of the first sets is obtained, and when the number is greater than a preset value, the first set is retained, and when the number is less than or equal to the preset value, the first set is deleted, and each retained first set is recorded as a second set, wherein the special information is text information or pattern information; labeling information for each of the second sets is obtained, and the labeling information is used as the labeling information of all first images in the clustering result corresponding to the second set, so as to obtain the first image and the corresponding labeling information.

3. The logistics packaging recycling management method according to claim 1 is characterized in that: The preset first model is trained using the remaining first images and the corresponding annotation information to obtain a feature extraction model, including: In the remaining first images, the first model is trained using the first image and the corresponding annotation information to obtain a second model, and the first image is input into the second model to obtain the first annotation information corresponding to the first image, wherein the first model is a model trained by the first training set, the function of the first model is to identify the number of people in the picture, and the number of samples in the first training set is greater than the number of first images with annotation information; the convolutional neural network is trained for multiple rounds using the first image and the corresponding annotation information and the first annotation information to obtain a feature extraction model, wherein in the first N rounds of training, the value of the first loss function between the first annotation information of the first image and the second annotation information corresponding to the first image output by the convolutional neural network is used as the target of training the convolutional neural network, the first loss function is a logarithmic loss function, and N is a positive integer greater than 2; in the remaining rounds of training, the value of the second loss function between the annotation information of the first image and the second annotation information corresponding to the first image output by the convolutional neural network is used as the target of training the convolutional neural network, and the second loss function is a cross entropy loss function.

4. Logistics packaging recycling management device, characterized in that: include: An acquisition module, used to acquire an image of each logistics package within a preset time period; An extraction module, used for performing image enhancement processing on the image of the logistics package to obtain a logistics package processed image; Acquire multiple images of historical logistics packages, perform saliency detection on the images of the historical logistics packages using a visual attention mechanism, detect salient objects on the images of the historical logistics packages, and segment the salient objects from the images of the historical logistics packages to obtain a first image; Extracting a first feature of the first image, performing cluster analysis on all the first images according to the first feature to obtain a clustering result, screening the first images according to the clustering result and calculating annotation information corresponding to the remaining first images; Using the remaining first images and corresponding annotation information to train the preset first model to obtain a feature extraction model; A grouping module is used to obtain feature information corresponding to the logistics package processing image based on the logistics package processing image and the feature extraction model; divide all the logistics package images into multiple groups according to the feature information, adjust the logistics package images in each group so that the number of logistics package images in each group is consistent, obtain a grouping result, and classify and recycle all the logistics packages according to the grouping result; The method further comprises: dividing all the images of the logistics packages into a plurality of groups according to the characteristic information, adjusting the images of the logistics packages in each group so that the number of the images of the logistics packages in each group is consistent, obtaining a grouping result, and classifying and recycling all the logistics packages according to the grouping result, including: The characteristic information corresponding to all the images of the logistics packages are clustered to obtain a plurality of clusters; all the clusters are sorted, wherein the distance between each cluster and the characteristic information corresponding to each logistics package image is calculated, and all the distances are summed up to obtain the total distance corresponding to each cluster, and the larger the total distance corresponding to the cluster, the higher the ranking; after sorting, each cluster is processed in turn, wherein it is determined whether the number of characteristic information corresponding to the images of the logistics packages contained in such cluster is less than a preset value, if so, the characteristic information corresponding to the images of the logistics packages closest to such cluster is moved into such cluster, and when the number of characteristic information corresponding to the images of the logistics packages contained in such cluster is the preset value, the moving in is stopped; at the same time, it is determined whether the number of characteristic information corresponding to the images of the logistics packages contained in such cluster is greater than the preset value, if so, the characteristic information corresponding to the image of the logistics package farthest from the centroid in such cluster is moved out of such cluster, and when the number of characteristic information corresponding to the images of the logistics packages contained in such cluster is the preset value, the moving out is stopped, wherein the total number of images of the logistics packages can be evenly divided.

5. Logistics packaging recycling management equipment, characterized in that: include: Memory for storing computer programs; A processor is used to implement the steps of the logistics packaging recycling management method as described in any one of claims 1 to 3 when executing the computer program.

6. A readable storage medium, characterized in that: The readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the logistics packaging recycling management method as described in any one of claims 1 to 3 are implemented.

Images