Comprehensive safety supervision system and method for postal industry

By applying video and image data annotation, convolution feature extraction network, LSTM and SVM technologies in the postal industry security supervision system, the problem of insufficient data collection and risk prediction capabilities of existing systems has been solved, and the in-depth development of the express delivery industry has been achieved Monitoring and risk warning have improved the accuracy and efficiency of safety supervision.

CN118918503AActive Publication Date: 2024-11-08BEIJING GAOCHENG TECH DEV CO LTD
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202410817088.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-06-24
Publication Date
2024-11-08
Estimated Expiration
2044-06-24

AI Technical Summary

Technical Problem

The existing postal industry safety supervision system is insufficient in data collection and risk prediction capabilities, and it is difficult to monitor streets and end networks in depth, resulting in insufficient risk prediction capabilities and unable to effectively ensure the safe and healthy development of the express delivery industry.

Method used

A comprehensive safety supervision system for the postal industry was designed, including video security patrol supervision module, receipt and inspection system supervision module, pass-through security system supervision module and business pressure risk warning module, using VOTT, LabelMe, convolution and pooling characteristics Extract network, LSTM units, SVM classification algorithm and other technologies, build target detection and classification models, realize the annotation and identification of video and picture data, generate intelligent security alarm information, and provide early warning rule management functions.

Benefits of technology

By improving the recognition accuracy of video and image data and data retrieval efficiency, enhancing the recognition ability of scenes and packaged images, improving the accuracy of security alarms and the reliability of early warnings, real-time monitoring and early warning of outlet operation status, helping Respond to safety risks in the express delivery industry in a timely manner.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118918503A_ABST
    Figure CN118918503A_ABST
Patent Text Reader

Abstract

The invention relates to a comprehensive safety supervision system and method for the postal industry, in particular to the field of information monitoring, and the method comprises the steps: collecting video data of related scenes of the postal industry, constructing a feature extraction network structure based on convolution and pooling, combining a target detection network of a residual block, and outputting a final scene category recognition result. The method comprises the following steps: collecting picture data of postal industry related scenes, selecting an LSTM unit to build a recurrent neural network model, outputting a picture data prediction classification result with category labels, collecting package image data of machine security check related scenes, separating the package image data of different labels by using an SVM classification algorithm, and effectively distinguishing the image data of different labels. The identification accuracy of the package image data is improved, the label type of the package image data is judged through the positive and negative class relation of the prediction result of the package image data, the management function of the early warning rule is provided, the management function of adding, modifying and deleting of multiple types of business rules is supported, and the operation efficiency and the express service quality are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of information monitoring, and more specifically, to a comprehensive security monitoring system and method for the postal industry. Background Art

[0002] With the rapid development of e-commerce, express delivery has gradually become an indispensable part of daily life. While the express delivery industry has made great contributions to the development of the national economy, many chaos have also emerged. However, the current safety supervision of the express delivery industry is weak, the supervision method is single, and the information support is insufficient. It cannot adapt to the rapid development of the express delivery industry and cannot effectively ensure the safe and healthy development of the express delivery industry.

[0003] The postal industry has a decentralized production process and generates data at a high frequency, making it a typical data-intensive industry. At present, the monitoring of market operation by information support can reach the city and county level, which can meet the relatively macro-level regulatory needs. However, it is difficult to penetrate deeper networks such as streets, outlets and terminals. The lack of in-depth data collection capabilities will inevitably lead to insufficient or missing risk prediction capabilities. Lacking the necessary risk prediction capabilities, support for safety supervision basically remains at the level of status quo and historical presentation, and the support is insufficient. Under certain conditions, it may even become a disguised burden.

[0004] Therefore, there is an urgent need for a comprehensive security supervision system and method for the postal industry. Summary of the invention

[0005] The present invention aims at the technical problems existing in the prior art and provides a comprehensive security supervision system and method for the postal industry to solve the problems raised in the above background technology.

[0006] The technical solution of the present invention to solve the above technical problems is as follows: A comprehensive safety supervision system for postal industry, comprising:

[0007] In a preferred embodiment, it includes a video security patrol supervision module, a mail inspection system supervision module, a machine security inspection system supervision module and a business pressure risk warning module;

[0008] Video security inspection and supervision module: Use cameras to collect video data of postal-related scenes and use VOTT annotation tools to annotate regions, build a feature extraction network structure based on convolution and pooling, and combine it with a target detection network with residual blocks;

[0009] Supervision module of the postal inspection system: connect to the image database of postal-related scenes to collect image data of postal-related scenes, use the Label Me annotation tool to select and specify category labels, select LSTM units to build a recurrent neural network model, and build an autoencoder model;

[0010] Machine security inspection system supervision module: Use cameras around the security inspection equipment to collect package image data of machine security inspection related scenes, use the SVM classification algorithm to separate the package image data with different labels, and automatically generate intelligent security inspection alarm information;

[0011] Business pressure risk warning module: provides warning rule management functions and supports the addition, modification, and deletion of multiple types of business rules. It uses historical data to verify the degree of rule matching and submits formal rules as the basis for automatic warning according to the rules.

[0012] In a preferred embodiment, the video security patrol supervision module is divided into a video acquisition and annotation subunit and a video scene recognition subunit, which specifically includes the following contents:

[0013] Video acquisition and annotation subunit: Use the camera to collect video data of postal industry-related scenes and transmit it back to the server, store it in the form of files, use the VOTT annotation tool to select rectangular annotation shapes to annotate the video data area, and add corresponding task labels to each annotated area.

[0014] Video scene recognition subunit: Construct a feature extraction network structure based on convolution and pooling and combine it with a target detection network with residual blocks. Input video data of postal-related scenes, add convolution layers to extract basic features of video data through multiple convolution kernels, and combine multiple convolution layers and jump connections to form a residual unit. After each residual unit, ReLU is used as the activation function layer to introduce nonlinearity and use convolution kernels to perform convolution operations on video data. The specific formula of the convolution operation is:

[0015]

[0016] Among them, A(i,j) represents an element of the basic feature of the video data, I(m,n) represents the basic feature pixel value of the video data, K(im,jn) represents the basic feature weight value of the video data, m and n represent the row and column indexes inside the convolution kernel respectively, i and j represent the row and column indexes of the basic features of the video data respectively, through the classification head global average pooling layer and the convolution layer The basic features of the video data are downsampled, the first convolution and pooling operations are repeated, and the high-level features of the video data are flattened into a one-dimensional vector through the fully connected layer of the classification head and the regression head, and input into the fully connected layer. The high-level features of the video data are mapped to different scene categories of the video data through the fully connected layer through the weight matrix, and the final scene category recognition result is output.

[0017] In a preferred embodiment, the receiving and inspection system supervision module is divided into a picture collection and annotation subunit and a picture scene recognition subunit, which specifically includes the following contents:

[0018] Image collection and annotation subunit: connect to the image database of postal industry-related scenes to collect image data of postal industry-related scenes and transmit it back to the server, store it in the form of files, and use the LabelMe annotation tool to select each object that appears in the image data of postal industry-related scenes and assign category labels.

[0019] Image scene recognition subunit: Select the LSTM unit to build a recurrent neural network model, input the image data with category labels and the hidden state of the previous timestamp according to the current timestamp, and calculate the activation value of the input layer through the Sigmoid activation function. The specific formula is:

[0020] T l =S(w l *[I t-1 ,x t ]+b l )

[0021] Among them, T l represents the activation value of the input layer, S() represents the Sigmoid activation function, and w l represents the weight matrix, I t-1 represents the hidden state of the previous timestamp, x t Indicates the image data with category label input at the current timestamp, b l Represents the bias term. The output layer selects the fully connected layer and inputs the image data with category labels at the current timestamp, the hidden state at the previous timestamp, and the cell state of the image data with category labels at the current timestamp. The activation value of the output layer is calculated by the Sigmoid activation function. The specific formula is:

[0022] T o =S(w o *[I t-1 ,x t ]+b o )

[0023] Among them, T o represents the activation value of the forgetting layer, S() represents the Sigmoid activation function, and w o represents the weight matrix, I t-1 represents the hidden state of the previous timestamp, x t Indicates the image data with category label input at the current timestamp, b o Represents the bias term, outputs the predicted classification results of the image data with category labels through the fully connected layer, builds the autoencoder model, inputs the image data with category labels to predict the classification results and maps them to the latent space, and supervises the autoencoder training by predicting the classification results of the image data with category labels.

[0024] In a preferred embodiment, the machine security inspection system supervision module is divided into a security inspection collection and labeling subunit and a security inspection scene recognition subunit, which specifically includes the following contents:

[0025] Security inspection collection and labeling subunit: Use cameras around the security inspection machine to collect packaged image data of scenes related to machine security inspection and transmit them back to the server, store them in the form of files, use the LabelMe labeling tool to select the packaged image data of scenes related to machine security inspection and assign different labels, and use the color histogram to extract the features of the packaged image data.

[0026] Security inspection scene recognition subunit: Use the SVM classification algorithm to separate the package image data of different labels and use the kernel function to map the features of the package image data of different labels to the high-dimensional space to construct the optimal hyperplane, so as to better separate the package image data of different labels in the high-dimensional space and maximize the projection point of the separated package image data on the hyperplane. The basic formula of the SVM is:

[0027] f(x)=sign(w*x+b)

[0028] Among them, f(x) represents the prediction result of the package image data, w represents the weight vector, which is used to indicate the direction of the hyperplane, x represents the feature vector of the package image data with different labels, and b represents the bias term. The label type of the package image data is judged by the positive and negative class relationship of f(x), and the positive and negative signals are fed back to the intelligent security inspection alarm information database to automatically generate intelligent security inspection alarm information. The abnormal express parcels found during the security inspection are recorded and reported to the postal management department, and then reported to the relevant departments according to the situation.

[0029] In a preferred embodiment, the business pressure risk warning module provides a warning rule management function and supports the management functions of adding, modifying and deleting multiple types of business rules. After the authorized user completes the rule setting, the historical data is used to verify the degree of rule matching and submit the formal rules as the basis for automatic warning according to the rules. The warning management function is continuously improved according to the actual situation. The historical data analysis is used to calculate the data change trajectory of the historical business and the number of couriers of each outlet, and the characteristics of the historical business changes and the number of couriers in the daily operation of the outlets are depicted. The real-time comparison is made with the current status of the outlets, and the current operating status of each outlet is marked. Timely warnings are issued for outlets that are about to be overwhelmed, existing, and backlogged. The classification and grading of warehouse explosion, existing, and backlog detection are automatically generated and combined with manual assistance. According to different event types, the event type is manually set in advance and the judgment rules are set.

[0030] In a preferred embodiment, the method specifically includes the following steps:

[0031] S101. Collect video data of postal-related scenes and use VOTT annotation tools to annotate regions, build a feature extraction network structure based on convolution and pooling and combine it with a target detection network of residual blocks, and output the final scene category recognition results;

[0032] S102. Collect image data of postal industry-related scenes, use Label Me annotation tool to select and specify category labels, select LSTM unit to build recurrent neural network model, output image data prediction and classification results with category labels, and build autoencoder model to integrate image data prediction and classification results;

[0033] S103. Collecting package image data of scenes related to machine security inspection, extracting features of package image data using color histogram, separating package image data with different labels using SVM classification algorithm, and determining the label type of package image data by the positive and negative class relationship of the prediction results of package image data;

[0034] S104. Provides management functions for early warning rules and supports the addition, modification, and deletion of multiple types of business rules. It uses historical data to verify the degree of rule matching and submits formal rules as the basis for automatic early warning according to the rules.

[0035] The beneficial effects of the present invention are as follows: regional labeling of collected video data of postal-related scenes is performed through the VOTT labeling tool, which helps to build a target detection network, ensures that the video data is clear and reduces blur and distortion problems, improves recognition accuracy, extracts basic features of video data through multiple convolution kernels, and effectively improves the recognition ability of scene categories; labels the collected image data through the LabelMe labeling tool, which helps to build a labeled data set required for training the model and improves model accuracy; selects LSTM units to build a recurrent neural network model, and uses the autoencoder model to integrate image data to predict classification results, which can effectively process time-series-related image data and extract advanced features, which helps to improve data retrieval efficiency and management; extracts package images through color histograms The characteristics of the data help to capture the color distribution information of the image. The SVM classification algorithm is used to classify the package image data, which can effectively distinguish image data with different labels and improve the recognition accuracy of the package image data. The positive and negative class relationships of the prediction results are used to determine the label type of the package image data. The SVM classification algorithm is combined for security inspection anomaly detection to generate intelligent security inspection alarm information to help security personnel quickly identify abnormal packages. By using historical data to verify the degree of rule matching, it is possible to more accurately judge situations that meet the requirements of the rules, improve the accuracy and reliability of early warnings, and automatically issue early warnings according to preset rules to achieve real-time monitoring and early warning of the operating status of outlets, help to promptly respond to problems such as impending warehouse explosions, existing and backlogs, and improve operational efficiency and express service quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 is a flow chart of the method of the present invention;

[0037] Figure 2 This is a system structure block diagram of the present invention. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0039] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, "plurality" means two or more, unless otherwise clearly and specifically defined.

[0040] In the description of the present application, the term "for example" is used to mean "used as an example, illustration or description". Any embodiment described as "for example" in the present application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any technician in the field to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes will not be elaborated in detail to avoid unnecessary details to obscure the description of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in the present application.

[0041] Example 1

[0042] This embodiment provides Figure 1 A comprehensive postal safety supervision method is shown, which specifically includes the following steps:

[0043] S101. Collect video data of postal-related scenes and use VOTT annotation tools to annotate regions, build a feature extraction network structure based on convolution and pooling and combine it with a target detection network of residual blocks, and output the final scene category recognition results;

[0044] S102. Collect image data of postal industry-related scenes, use Label Me annotation tool to select and specify category labels, select LSTM unit to build recurrent neural network model, output image data prediction and classification results with category labels, and build autoencoder model to integrate image data prediction and classification results;

[0045] S103. Collecting package image data of scenes related to machine security inspection, extracting features of package image data using color histogram, separating package image data with different labels using SVM classification algorithm, and determining the label type of package image data by the positive and negative class relationship of the prediction results of package image data;

[0046] S104. Provides management functions for early warning rules and supports the addition, modification, and deletion of multiple types of business rules. It uses historical data to verify the degree of rule matching and submits formal rules as the basis for automatic early warning according to the rules.

[0047] Example 2

[0048] This embodiment provides Figure 2 A comprehensive safety supervision system for the postal industry is shown, specifically including: a video security patrol supervision module, a collection and delivery inspection system supervision module, a machine security inspection system supervision module and a business pressure risk warning module;

[0049] Video security inspection and supervision module: Use cameras to collect video data of postal-related scenes and use VOTT annotation tools to annotate regions, build a feature extraction network structure based on convolution and pooling, and combine it with a target detection network with residual blocks;

[0050] Supervision module of the postal inspection system: connect to the image database of postal-related scenes to collect image data of postal-related scenes, use the Label Me annotation tool to select and specify category labels, select LSTM units to build a recurrent neural network model, and build an autoencoder model;

[0051] Machine security inspection system supervision module: Use cameras around the security inspection equipment to collect package image data of machine security inspection related scenes, use the SVM classification algorithm to separate the package image data with different labels, and automatically generate intelligent security inspection alarm information;

[0052] Business pressure risk warning module: provides warning rule management functions and supports the addition, modification, and deletion of multiple types of business rules. It uses historical data to verify the degree of rule matching and submits formal rules as the basis for automatic warning according to the rules.

[0053] S101. Collect video data of postal-related scenes and use VOTT annotation tools to annotate regions, build a feature extraction network structure based on convolution and pooling and combine it with a target detection network of residual blocks, and output the final scene category recognition results;

[0054] Furthermore, the video data of postal industry-related scenes are collected by cameras and transmitted back to the server and stored in the form of files. The VOTT annotation tool is used to select rectangular annotation shapes to perform regional annotation on the video data, and corresponding task labels are added to each annotated area, including target detection, target tracking, and target recognition, so as to quickly and accurately complete the regional annotation task of video data.

[0055] Furthermore, the camera resolution is set to ≥ 1080P to ensure that the video data has less blur, screen distortion and distortion problems, and the camera shooting angle is fixed so that the scene in the video data is clear. In the 1080P video data, the resolution of the human head of the company's internal personnel in the picture is ensured to be ≥ 40×40, and the overlapping and blocked part of the human head in the picture does not exceed 20%. The lighting range of the camera scene is guaranteed to be between 70l ux and 600l ux. If the light intensity is insufficient, white light is used for fill lighting.

[0056] Furthermore, a target detection network based on a feature extraction network structure of convolution and pooling combined with a residual block is constructed, video data of postal industry-related scenes is input, a convolution layer is added to extract basic features of the video data through multiple convolution kernels, multiple convolution layers and jump connections are combined into a residual unit, wherein the residual unit is combined into a residual block, and the gradient vanishing problem in the network structure is effectively solved through jump connections. After each residual unit, ReLU is used as an activation function layer to introduce nonlinearity and convolution kernels are used to perform convolution operations on the video data, so as to increase the stability and nonlinearity of the network structure. The specific formula of the convolution operation is:

[0057]

[0058] Among them, A(i,j) represents an element of the basic feature of the video data, I(m,n) represents the basic feature pixel value of the video data, K(im,jn) represents the basic feature weight value of the video data, m and n represent the row and column indexes inside the convolution kernel respectively, i and j represent the row and column indexes of the basic features of the video data respectively, through the classification head global average pooling layer and the convolution layer The basic features of the video data are downsampled to reduce the size of the basic features of the video data and retain important feature information. The first convolution and pooling operations are repeated to extract high-level features of the video data. The high-level features of the video data are flattened into a one-dimensional vector through the fully connected layer of the classification head and the regression head, and are input into the fully connected layer. The fully connected layer is used to map the high-level features of the video data to different scene categories of the video data through the weight matrix, and the final scene category recognition result is output.

[0059] Furthermore, the scene category recognition results are divided into personnel detection, mail detection, operation detection, and behavior detection, among which personnel detection includes personnel leaving their posts detection, face recognition, person-document matching, and hair and clothing detection; mail detection includes mail and express stacking status detection, mail and express landing detection, mail and express damage detection, postal mailbox unpacking frequency detection, and unpacking inspection detection; operation detection includes security inspection machine operation compliance detection, operation out-of-bounds detection, and conveyor belt cargo drop detection; and behavior detection includes violent sorting behavior detection and conveyor belt operation violation detection.

[0060] S102. Collect image data of postal industry-related scenes, use Label Me annotation tool to select and specify category labels, select LSTM unit to build recurrent neural network model, output image data prediction and classification results with category labels, and build autoencoder model to integrate image data prediction and classification results;

[0061] Furthermore, the image database of postal industry-related scenes is connected to collect image data of postal industry-related scenes and transmit it back to the server, which is stored in the form of files. The Label Me annotation tool is used to select each object appearing in the image data of postal industry-related scenes and assign category labels, including clothing, documents, shoes, food gift boxes, fruits, accessories, skin care products, tea, data, mobile phones, computers, alcohol, medicines, meat, and unopened boxes.

[0062] Furthermore, the entire data is divided into multiple data folders according to the category labels of the image data. Each data folder corresponds to a category label. The number of image data in the data folder of each category label is greater than 50, and the total amount of image data is limited to 100,000. The target number of category labels in a single image data cannot exceed 1,000. The image data size is limited to 14M, and the aspect ratio of the image data is within 3:1, with the longest side less than 4096px and the shortest side greater than 30px.

[0063] Furthermore, the LSTM unit is selected to build a recurrent neural network model. The image data with the category label and the hidden state of the previous timestamp are input according to the current timestamp. The activation value of the input layer is calculated by the Sigmoid activation function. The specific formula is:

[0064] T l =S(w l *[I t-1 ,x t ]+b l )

[0065] Among them, T l represents the activation value of the input layer, S() represents the Sigmoid activation function, and w l represents the weight matrix, I t-1 represents the hidden state of the previous timestamp, x t Indicates the image data with category label input at the current timestamp, b l Represents the bias term. The output layer selects the fully connected layer and inputs the image data with category labels at the current timestamp, the hidden state at the previous timestamp, and the cell state of the image data with category labels at the current timestamp. The activation value of the output layer is calculated by the Sigmoid activation function. The specific formula is:

[0066] T o =S(w o *[I t-1 ,x t ]+b o )

[0067] Among them, T o represents the activation value of the forgetting layer, S() represents the Sigmoid activation function, and w o represents the weight matrix, I t-1 represents the hidden state of the previous timestamp, x t Indicates the image data with category label input at the current timestamp, b oRepresents the bias term, outputs the predicted classification results of image data with category labels through the fully connected layer, builds an autoencoder model to input the predicted classification results of image data with category labels and maps them to the latent space, supervises the autoencoder training through the predicted classification results of image data with category labels, predicts the classification results for image data without category labels, obtains the reconstruction loss through the difference measurement between the reconstructed output of the decoder and the image data without category labels, and integrates the predicted classification results of image data with category labels and the predicted classification results of image data without category labels to help train the recurrent neural network model and the autoencoder model.

[0068] S103. Collecting package image data of scenes related to machine security inspection, extracting features of package image data using color histogram, separating package image data with different labels using SVM classification algorithm, and determining the label type of package image data by the positive and negative class relationship of the prediction results of package image data;

[0069] Furthermore, cameras are used around the security inspection machine to collect packaged image data of scenes related to machine security inspection and transmit them back to the server and store them in the form of files. The LabelMe annotation tool is used to select the packaged image data of scenes related to machine security inspection and assign different labels, and the color histogram is used to extract the features of the packaged image data.

[0070] Furthermore, the SVM classification algorithm is used to separate the package image data of different labels and the kernel function is used to map the features of the package image data of different labels to a high-dimensional space to construct an optimal hyperplane, so as to better separate the package image data of different labels in the high-dimensional space and maximize the projection point of the separated package image data on the hyperplane. The basic formula of the SVM is:

[0071] f(x)=sign(w*x+b)

[0072] Among them, f(x) represents the prediction result of the package image data, w represents the weight vector, which is used to indicate the direction of the hyperplane, x represents the feature vector of the package image data with different labels, b represents the bias term, which is used to adjust the distance between the hyperplane and the origin, and the label type of the package image data is judged by the positive and negative class relationship of f(x). When f(x)<0, the prediction is judged to be positive, and when f(x)>0, the prediction is judged to be negative. The positive and negative signals are fed back to the intelligent security inspection alarm information database, and the intelligent security inspection alarm information is automatically generated. The abnormal express parcels found during the security inspection are recorded and reported to the postal management department, and reported to the relevant departments according to the situation. The abnormal security inspection alarm information includes: the time and place of the problem, the type of problem, the waybill number, the package picture, and the waybill picture.

[0073] S104. Provide management functions for early warning rules and support the management functions of adding, modifying and deleting multiple types of business rules, use historical data to verify the matching degree of rules, and submit formal rules as the basis for automatic early warning according to the rules;

[0074] Furthermore, it provides the management function of early warning rules and supports the management functions of adding, modifying and deleting multiple types of business rules. After the authorized user completes the rule setting, he can use historical data to verify the degree of rule matching and submit the formal rules as the basis for automatic early warning according to the rules. The early warning management function is continuously improved according to the actual situation. Through historical data analysis, the data change trajectory of the historical business and the number of couriers of each outlet is calculated, and the characteristics of the historical business changes and the number of couriers in the daily operation of the outlets are depicted. They are compared with the current status of the outlets in real time, and the current operating status of each outlet is marked. Timely warnings are issued for outlets that are about to be overwhelmed, existing, and backlogged. The classification and grading of warehouse overflow, existing, and backlog detection are automatically generated and combined with manual assistance. According to different event types, the event type is manually set in advance and the judgment rules are set.

[0075] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0076] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0077] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0078] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0079] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0080] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0081] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A comprehensive safety supervision system for the postal industry, characterized in that: Specifically include: It includes video security inspection and supervision module, mail and mail inspection and supervision module, machine security inspection and supervision module, and business pressure risk warning module; Video security inspection and supervision module: Use cameras to collect video data of postal-related scenes and use VOTT annotation tools to annotate regions, build a feature extraction network structure based on convolution and pooling, and combine it with a target detection network with residual blocks; Supervision module of the mail inspection system: connect to the image database of postal industry-related scenes to collect image data of postal industry-related scenes, use the LabelMe annotation tool to select and specify category labels, select LSTM units to build a recurrent neural network model, and build an autoencoder model; Machine security inspection system supervision module: Use cameras around the security inspection equipment to collect package image data of machine security inspection related scenes, use the SVM classification algorithm to separate the package image data with different labels, and automatically generate intelligent security inspection alarm information; Business pressure risk warning module: provides warning rule management functions and supports the addition, modification and deletion of multiple types of business rules. It uses historical data to verify the degree of rule matching and submits formal rules as the basis for automatic warning according to the rules.

2. A postal industry comprehensive safety supervision system according to claim 1, characterized in that: The video security inspection and supervision module uses a camera to collect video data of scenes related to the postal industry, uses the VOTT annotation tool to perform regional annotation on the video data and add corresponding task labels, constructs a feature extraction network structure based on convolution and pooling and combines a target detection network with a residual block, inputs video data of scenes related to the postal industry, adds a convolution layer to extract basic features of the video data through multiple convolution kernels, uses ReLU as an activation function layer to introduce nonlinearity and uses convolution kernels to perform convolution operations on the video data, and downsamples the basic features of the video data of the convolution layer, repeatedly iterates the first convolution and pooling operations, flattens the high-level features of the video data into a one-dimensional vector through the fully connected layers of the classification head and the regression head, and inputs them into the fully connected layer, uses the fully connected layer to map the high-level features of the video data to different scene categories of the video data through a weight matrix, and outputs the final scene category recognition result.

3. A postal industry comprehensive safety supervision system according to claim 2, characterized in that: The specific formula of the convolution operation is: Among them, A(i,j) represents an element of the basic feature of the video data, I(m,n) represents the basic feature pixel value of the video data, K(im,jn) represents the basic feature weight value of the video data, m and n represent the row and column indexes inside the convolution kernel respectively, and i and j represent the row and column indexes of the basic feature of the video data respectively.

4. A postal industry comprehensive safety supervision system according to claim 1, characterized in that: The collection and inspection system supervision module is connected to the image database of postal industry-related scenes to collect image data of postal industry-related scenes, uses the LabelMe annotation tool to select each object appearing in the image data of postal industry-related scenes and assign category labels, selects the LSTM unit to build a recurrent neural network model, inputs the image data with category labels and the hidden state of the previous timestamp according to the current timestamp, calculates the activation value of the input layer through the Sigmoid activation function, selects the fully connected layer as the output layer, and inputs the image data with category labels according to the current timestamp, the hidden state of the previous timestamp, and the cell state of the image data with category labels at the current timestamp, calculates the activation value of the output layer through the Sigmoid activation function, outputs the image data with category labels through the fully connected layer to predict the classification results, constructs an autoencoder model to input the image data with category labels to predict the classification results and maps them to the latent space, and supervises the autoencoder training by predicting the classification results through the image data with category labels.

5. A postal industry comprehensive safety supervision system according to claim 4, characterized in that: The specific formula for calculating the activation value of the input layer is: T l =S(w l *[I t-1 ,x t ]+b l ) Among them, T l represents the activation value of the input layer, S() represents the Sigmoid activation function, and w l represents the weight matrix, I t-1 represents the hidden state of the previous timestamp, x t Indicates the image data with category label input at the current timestamp, b l Represents the bias term.

6. A postal industry comprehensive safety supervision system according to claim 4, characterized in that: The specific formula for calculating the activation value of the output layer is: T o =S(w o *[I t-1 ,x t ]+b o ) Among them, T o represents the activation value of the forgetting layer, S() represents the Sigmoid activation function, and w o represents the weight matrix, I t-1 represents the hidden state of the previous timestamp, x t Indicates the image data with category label input at the current timestamp, b o Represents the bias term.

7. A postal industry comprehensive safety supervision system according to claim 1, characterized in that: The machine security inspection system supervision module uses cameras around the security inspection machine equipment to collect packaged image data of machine security inspection related scenes, uses the LabelMe annotation tool to select the packaged image data of machine security inspection related scenes and assign different labels, uses the SVM classification algorithm to separate the packaged image data with different labels and uses the kernel function to map the features of the packaged image data with different labels to a high-dimensional space to construct an optimal hyperplane, so as to better separate the packaged image data with different labels in the high-dimensional space and maximize the projection point of the separated packaged image data on the hyperplane.

8. A postal industry comprehensive safety supervision system according to claim 7, characterized in that: The basic formula form of SVM is: f(x)=sign(w*x+b) Where f(x) represents the prediction result of the packet image data, w represents the weight vector used to represent the direction of the hyperplane, x represents the feature vector of the packet image data with different labels, and b represents the bias term.

9. A postal industry comprehensive safety supervision system according to claim 1, characterized in that: The business pressure risk warning module provides the management function of warning rules and supports the management functions of adding, modifying and deleting multiple types of business rules. It continuously improves the warning management function according to the actual situation, calculates the data change trajectory of the historical business and the number of couriers of each outlet through historical data analysis, and describes the characteristics of the historical business changes and the number of couriers of the daily operation of the outlets, compares them with the current status of the outlets in real time, marks the current operation status of each outlet, automatically generates and combines manual assistance to comprehensively set the classification and grading of warehouse explosion, existing and backlog detection, and manually sets the event type in advance according to the different event types, and sets the judgment rules.

10. A postal industry comprehensive safety supervision method is applied to a postal industry comprehensive safety supervision system as claimed in any one of claims 1 to 9, characterized in that: The specific steps include: S101. Collect video data of postal-related scenes and use VOTT annotation tools to annotate regions, build a feature extraction network structure based on convolution and pooling and combine it with a target detection network of residual blocks, and output the final scene category recognition results; S102. Collect image data of postal industry-related scenes, use LabelMe annotation tool to select and specify category labels, select LSTM unit to build recurrent neural network model, output image data prediction and classification results with category labels, and build autoencoder model to integrate image data prediction and classification results; S103. Collecting package image data of scenes related to machine security inspection, extracting features of package image data using color histogram, separating package image data with different labels using SVM classification algorithm, and determining the label type of package image data by the positive and negative class relationship of the prediction results of package image data; S104. Provides management functions for early warning rules and supports the addition, modification, and deletion of multiple types of business rules. It uses historical data to verify the degree of rule matching and submits formal rules as the basis for automatic early warning according to the rules.

Citation Information

Patent Citations

  • Urban rail transit panoramic monitoring video fault detection method based on depth learning

    CN106709511A

  • Scene recognition method based on convolutional neural network and recursive automatic encoder model

    CN110188827A

  • Streetscape image segmentation method and system based on bidirectional attention network

    CN114677507A

  • Postal standard mailing supervision method based on deep learning

    CN115984750A

  • Power artificial-intelligence visual-analysis system on basis of multi-core heterogeneous parallel computing

    WO2019109771A1