Video monitoring product pushing method, device, equipment and computer program product

Abstract

The application provides a video monitoring product pushing method, device, equipment and computer program product. The method comprises: determining a potential video monitoring demand of a target family based on a family portrait attribute of the target family; determining at least one target atomic capability suitable for the potential video monitoring demand from an atomic capability library of a video monitoring service of a computing power network; configuring a target video monitoring product run by the computing power network based on a container image created in advance for the at least one target atomic capability; and pushing the target video monitoring product to the target family.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, and particularly relates to a video monitoring product pushing method and device, equipment and computer program product. BACKGROUND

[0002] Computing power network is a new type of information infrastructure which allocates and flexibly schedules computing resources, storage resources and network resources according to business needs among cloud, network and edge.

[0003] With the continuous improvement of people's living standards, it has become a future development trend to carry out intelligent video monitoring services for families on the computing power network. How to customize video monitoring products of the computing power network for families is a technical problem to be solved by the present application. SUMMARY

[0004] The present application aims to provide a video monitoring product pushing method, device, equipment and computer program product, which can customize video monitoring products of the computing power network for families.

[0005] In order to achieve the above-mentioned purpose, the embodiments of the present application are implemented as follows:

[0006] In a first aspect, a video monitoring product pushing method is provided, comprising:

[0007] determining potential video monitoring needs of a target family based on family portrait attributes of the target family;

[0008] determining at least one target atomic capability suitable for the potential video monitoring needs from an atomic capability library of video monitoring services of a computing power network; the atomic capability refers to a basic independent capability;

[0009] configuring a target video monitoring product run by the computing power network based on a container image created in advance for the at least one target atomic capability;

[0010] pushing the target video monitoring product to the target family.

[0011] In a second aspect, a video monitoring product pushing device is provided, comprising:

[0012] a demand matching module configured to determine potential video monitoring needs of a target family based on family portrait attributes of the target family;

[0013] a capability matching module configured to determine at least one target atomic capability suitable for the potential video monitoring needs from an atomic capability library of video monitoring services of a computing power network;

[0014] A product configuration module configures a target video monitoring product run by the computing power network based on a container image created in advance for the at least one target atomic capability.

[0015] A product pushing module pushes the target video monitoring product to the target family.

[0016] In a third aspect, an electronic device is provided, including a processor, and a memory configured to store computer executable instructions that, when executed, cause the processor to perform the method of the first aspect.

[0017] In a fourth aspect, a computer program product is provided, including a computer readable storage medium storing a computer program, the computer program being operable to cause a computer to perform the method of the first aspect.

[0018] The embodiments of the present application decouple the video monitoring service of the computing power network in advance to construct an atomic capability library, and create corresponding container images for the atomic capabilities in the atomic capability library. When a video monitoring product needs to be customized for a target family, the potential video monitoring demand of the target family is analyzed based on the family portrait attributes of the target family, and then at least one target atomic capability that matches the potential video monitoring demand is queried from the atomic capability library to configure a target video monitoring product run by the computing power network and suitable for the target family. In related applications, since the atomic capability container images can be reused, the computing power network can customize more accurate and more personalized video monitoring products for different users through the atomic capability library to meet the video monitoring demands of different users. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments described in the embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0020] Figure 1 The first flowchart of the pushing method of the video monitoring product of the embodiments of the present application.

[0021] Figure 2 The second flowchart of the pushing method of the video monitoring product of the embodiments of the present application.

[0022] Figure 3 The third flowchart of the pushing method of the video monitoring product of the embodiments of the present application.

[0023] Figure 4 is a first schematic diagram of an LSTM algorithm structure.

[0024] Figure 5 is a second schematic diagram of an LSTM algorithm structure.

[0025] Figure 6 is a third schematic diagram of an LSTM algorithm structure.

[0026] Figure 7 is a fourth schematic diagram of an LSTM algorithm structure.

[0027] Figure 8 is a schematic diagram of a family portrait attribute.

[0028] Figure 9 is a fourth flowchart of a pushing method of a video monitoring product according to an embodiment of the present application.

[0029] Figure 10 is a fifth flowchart of a pushing method of a video monitoring product according to an embodiment of the present application.

[0030] Figure 11 is a structural schematic diagram of a pushing device of a video monitoring product according to an embodiment of the present application.

[0031] Figure 12 is a structural schematic diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0032] In order to enable persons skilled in the art to better understand the technical solutions in the present specification, the technical solutions in the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present specification, but not all the embodiments. Based on the embodiments in the present specification, all other embodiments obtained by persons skilled in the art without creative labor should fall within the protection scope of the present specification.

[0033] An embodiment of the present application provides a pushing method of a video monitoring product. Figure 1 is a flowchart of the pushing method, comprising:

[0034] S102, determining a potential video monitoring demand of the target family based on a family portrait attribute of the target family.

[0035] In the embodiment, the family portrait attribute of the target family can be obtained by fusing user portrait attributes of each family member in the target family.

[0036] The user portrait attribute can be, but is not limited to, at least one of basic information, a social network, a living habit, a technical preference, an economic condition, and an interest, and is not specifically limited herein.

[0037] In an implementation manner, weights of the family members of the target family can be determined based on interaction information between the family members of the target family; and then, the user portrait attributes of the family members of the target family are fused by weighting according to the corresponding weights, to obtain the family portrait attribute of the target family.

[0038] On the basis described above, the embodiment can use a deep learning technology to realize the prediction of the user portrait attribute. That is, a deep learning model is trained based on user data of a sample family member and a user portrait attribute label (ground truth) corresponding to the sample family member. In the training process, after the user data of the sample family member is input into the deep learning model, the user portrait attribute of the sample family member is predicted by the deep learning model, to serve as a training result. Then, a loss between the training result and the user portrait attribute label (ground truth) is calculated, and the model parameters of the deep learning model are adjusted for the purpose of reducing the loss. Through iteration of the model parameters, the user portrait attribute of the sample family member predicted by the deep learning model finally coincides with the ground truth, so that the deep learning model has the prediction capability of the user portrait attribute. After the deep learning model is trained, the user data of the family members in the target family can be input into the deep learning model to predict the user portrait attributes of the family members in the target family; and then, the user portrait attributes of the family members in the target family are fused to obtain the family portrait attribute of the target family.

[0039] It should be noted that the deep learning model of the embodiment is used to realize the prediction based on a classification function, and any classification model can be applied to the scheme of the embodiment.

[0040] In addition, to improve the prediction effect of the deep learning model, in the training stage, the user data of the candidate sample family members that have been prepared can be clustered based on a Gaussian Mixture Model (GMM) clustering algorithm to obtain a plurality of clustering clusters. Then, it is determined whether there is a common user portrait attribute in each clustering cluster, and in the case of existence, the common user portrait attribute is taken as the user portrait attribute label of the candidate sample family member corresponding to the clustering cluster, and a sample family member is selected from the candidate sample family members with the user portrait attribute label. Through the clustering of the GMM, effective user portrait attributes can be mined to train the deep learning model.

[0041] The GMM clustering principle is used to determine at least two Gaussian distributions corresponding to the user data of the candidate sample family member, and each Gaussian distribution represents a cluster.

[0042] S104, determining at least one target atomic capability suitable for the potential video monitoring requirement from the atomic capability library of the video monitoring service of the computing power network.

[0043] The embodiment decouples the functions of the video monitoring service of the computing power network to obtain a plurality of basic independent functions, i.e., atomic capabilities. Then, the atomic capability library is created for the atomic capabilities for users to match. For example, after determining the potential video monitoring requirement of the target family, at least one target atomic capability can be matched from the atomic capability library according to the potential video monitoring requirement of the target family.

[0044] S106, based on the container image created in advance for the at least one target atomic capability, a target video monitoring product running on the computing power network is configured.

[0045] The embodiment creates a corresponding container file for each atomic capability in the atomic capability library, and adds the running configuration (such as library file, configuration file, environment variable, network setting, and data volume mapping storage path) and running script of each atomic capability to the corresponding container file to create a container image of each atomic capability.

[0046] It should be understood that the container image can be reused. The embodiment can flexibly select the container image of the atomic capability according to the standard design rules of the video monitoring product to assemble different video monitoring products, thereby providing customized video monitoring products for different users.

[0047] S108, pushing the target video monitoring product to the target family.

[0048] Specifically, after pushing the target video monitoring product to the target family, the embodiment can also provide a demonstration of the target video monitoring product for the target family. The target family can provide opinions in the demonstration and adjust the target video monitoring product according to the opinions.

[0049] As an exemplary introduction, a mapping model of the computing power network can be created based on a system modeling language (SysML), and the mapping model includes logical relationships between atomic capabilities and resources of the computing power network. Then, based on the mapping model, the resource requirement of at least one target atomic capability and the resource supply amount of the computing power network are taken as input parameters, and the resource configuration strategy of the computing power network for the target video monitoring product is solved according to the target optimization direction, to obtain a target resource configuration strategy. Finally, the target resource configuration strategy is provided to the target family, and the target video monitoring product is demonstrated to the target family based on the computing power network according to the target resource configuration strategy.

[0050] The target optimization direction can be to minimize resource overhead or maximize target atomic capability performance. For example, a resource configuration strategy is first determined according to the minimization of resource overhead, and the target video monitoring product is demonstrated to the target family according to the resource configuration strategy. If the target family is satisfied, the target video monitoring product is finally landed using the minimization of resource overhead; if the target family is not satisfied, a resource configuration strategy is determined again to maximize the performance of the target atomic capability, so as to demonstrate the target video monitoring product to the target family. In this way, the resources of the computing power network can be solved only on the premise of meeting the use demand of the target family.

[0051] In summary, the method of the embodiment decouples the video monitoring service of the computing power network in advance to construct an atomic capability library, and creates a corresponding container image for the atomic capability in the atomic capability library. When a video monitoring product needs to be customized for a target family, the potential video monitoring demand of the target family is analyzed based on the family portrait attributes of the target family, and at least one target atomic capability that matches the potential video monitoring demand is queried from the atomic capability library to configure a target video monitoring product suitable for the target family run by the computing power network. In related applications, since the atomic capability container image can be reused after being containerized, the computing power network can customize more accurate and personalized video monitoring products for different users through the atomic capability library to meet the video monitoring demands of different users.

[0052] The specific application of the method of the embodiment will be introduced below.

[0053] The computing power network of the method of the embodiment is the ABCDNETS (Artificial intelligence, Blockchain, Cloud Computing, Big Data, Network, Edge Computing, Terminal, Security) computing power network. The deep learning model used is the Long Short-Term Memory (LSTM) model. Reference Figure 2 As shown in the figure, the process mainly includes the following 9 parts:

[0054] 1) User portrait construction:

[0055] Collect the corresponding family user information through various channels, use GMM and LSTM to construct the family user portrait, and match the appropriate family label;

[0056] 2) Potential demand prediction:

[0057] In-depth analysis of the family user portrait and its label, and timely prediction of the potential video monitoring demand of the family and its members;

[0058] 3) Product demand matching:

[0059] According to the potential business demand of the family, automatically match the corresponding ABCDNETS atomic ability (such as computing power, network, AI, etc.) in the video monitoring product atomic ability library;

[0060] 4) Personalized product design:

[0061] Match the ABCDNETS atomic ability according to the standard product design rules to generate a new specification of video monitoring product;

[0062] 5) Product scheme evaluation:

[0063] Evaluate the new specification of video monitoring product, including technical feasibility, cost-benefit analysis, etc.;

[0064] 6) Promotion strategy formulation:

[0065] Promote the evaluated new specification of video monitoring product to the relevant family users in time;

[0066] 7) Product scheme implementation:

[0067] According to the mapping relationship between the video monitoring product atomic ability and the related algorithm network resources, combined with various strategies (such as cost priority, performance priority), generate a specific resource implementation scheme;

[0068] 8) Product demonstration and trial:

[0069] Provide online product demonstrations for home users to remotely understand product operation and effects, and also provide free trials of products so that they can experience the advantages of the products;

[0070] 9) User feedback and product iteration:

[0071] Establish a user feedback mechanism to collect user experience and improvement suggestions for the product, and continuously iterate product functions based on user feedback to improve user satisfaction.

[0072] The construction process of the home user portrait is shown in Figure 3 Collect multi-source data on video monitoring needs of home users through interest communities, home network devices, web browsing records, social media platforms, and other means, and perform data backup, cleaning, verification, and integration encryption to ensure data quality and security, and build a multi-level and comprehensive data processing and protection system. Then use GMM algorithm to perform in-depth feature analysis on the integrated home user data, output the final clustering result, and use LSTM to analyze the data and match the appropriate home label based on the user clustering result to construct the home portrait. When new data is collected, it is also processed according to the standard process, and then the GMM algorithm is used to cluster user data, the LSTM analyzes the clustering result and matches the appropriate label, and updates to the previously constructed home portrait.

[0073] The GMM clustering algorithm first preprocesses the data so that each dimension of the data is in the same order of magnitude. This makes the data more stable and reduces the relative deviation between data. Second, a Gaussian model is established, assuming that the data is composed of several Gaussian distributions. Each Gaussian distribution is a parameter corresponding to a probability distribution function. Therefore, the Gaussian model can be used to describe the distribution of data. Then the parameters, mean and variance of each Gaussian distribution are determined to calculate the probability distribution of each data point. After that, the maximum likelihood method is used to determine which Gaussian distribution each data point belongs to. Finally, the EM algorithm is used to solve the parameters of the GMM. In the EM algorithm, first initialize the model parameters, then calculate the response of all data points, then update the parameters of each Gaussian distribution and calculate the likelihood function of the model, and finally determine whether the parameters of the model have converged. If converged, end the iteration. Otherwise, continue to calculate the response of all data points and repeat the subsequent steps until the parameters of the model converge

[0074] Gaussian Mixture Model (GMM) gives the probability of each sample point belonging to each class, rather than determining that it completely belongs to one class, which is called soft clustering. When the sample data X is one-dimensional data, the Gaussian distribution follows the probability density function below:

[0075]

[0076] where u is the data mean (expectation), and σ is the data standard deviation. In GMM, it is assumed that the data is composed of several Gaussian distributions. The probability density function of Gaussian distribution is:

[0077]

[0078] where u is the data mean (expectation), Σ is the covariance, and n is the data dimension.

[0079] The GMM algorithm assumes that the data points are generated by one or more Gaussian distributions, and estimates the parameters of the Gaussian distribution of each cluster by the method of maximum likelihood estimation. The Gaussian mixture model can be regarded as a model composed of K single Gaussian models, and the K sub-models are the hidden variables of the mixture model. The core idea is to assume that the data can be generated from multiple Gaussian distributions, and each individual sub-model is a standard Gaussian model, whose mean u j and variance Σ j are parameters to be estimated. In addition, each sub-model also has a parameter π j , which can be understood as a weight or a probability of generating data. The goal of GMM is to maximize the likelihood function of the data:

[0080]

[0081] The Expectation Maximization (EM) algorithm is used to solve the parameters of GMM. The flow of EM algorithm is as follows: 1, initialize the parameters of the model; 2, calculate the responsibility of all data points; 3, update the parameters of each Gaussian distribution; 4, calculate the likelihood function of the model; 5, judge whether the parameters of the model have converged. If it converges, end the iteration. Otherwise, return to step 2.

[0082] The following formula is used to calculate the responsibility of each data point:

[0083]

[0084] The following formula is used to update the parameters of each Gaussian distribution:

[0085]

[0086]

[0087]

[0088] Through continuous iteration, the data can be clustered using the GMM clustering algorithm. After data clustering, the data is parsed through the LSTM algorithm, and the LSTM algorithm structure is as shown in Figure 4 The LSTM algorithm is a special RNN (Recurrent Neural Network), which can learn long-term dependencies and can solve the problems of gradient disappearance and gradient explosion in long sequence training process, and can have better performance in longer sequences. Unlike RNN, LSTM can remove or add information of "cell state" through "gate" structure, realize the retention of important content and the removal of unimportant content, and output a probability value between 0 and 1 through the Sigmoid layer, which describes how much quantity can pass, 0 means "no task variable is allowed to pass", and 1 means "all variables are allowed to pass". Through the forget gate, the input gate and the output gate control the memory state of the previous information, the input information and the output information, so as to ensure that the network can better learn long-distance dependencies.

[0089] The structure of the LSTM forget gate is as shown in Figure 5 It is responsible for deciding how much of the previous unit state is retained in the current unit state, that is, deciding what information to discard from the cell state. The gate reads h t-1 and x t , and then outputs a number f t between 0 and 1 after passing through the sigmoid layer. f t-1 is multiplied point by point to each number in the cell state C t . The value of f t is 0, which means complete discard, and 1 means complete retention.

[0090] f t =σ(W f ·[h t-1 ,x t ]+b f )

[0091] The structure of the LSTM input gate is as shown in Figure 6 It is responsible for deciding how much of the current input is retained in the current unit state, which includes two parts. The first part is the sigmoid layer, which decides what value to update, and the second part is the tanh layer, which updates the information that needs to be updated to the cell state. The tanh layer creates a new cell state value vector which will be added to the state. Then update the old cell state C t-1 to C t , and the old state to f tMultiplication, discard the information that needs to be discarded, plus Update the cell state.

[0092] The LSTM output gate structure is shown in Figure 7 , which is responsible for determining how much of the current unit state is output at the current time. The sigmoid layer determines which part of the cell state will be output. The cell state is processed through tanh to get a value between -1 and 1, and it is multiplied by the output of the sigmoid gate to finally output only the part that determines the output. The structure is shown in the following figure:

[0093] i t = σ(W i · [h t-1 , x t ] + b i )

[0094]

[0095] By clustering family user data using GMM and then using LSTM to parse the data, each family member is matched with the corresponding attribute label, and each member is given a personal weight based on their interaction data, forming a complete family portrait label system. This lays the foundation for the precise marketing of personalized video monitoring products. The final family portrait effect is shown in Figure 8 , which should at least include six dimensions of feature information of family members (basic information, social network, living habits, technology preference, economic status, and interest), and each member is given a corresponding personal weight based on their interaction information, showing a more realistic family situation.

[0096] In addition to building a precise family portrait label system, to support the rapid construction of personalized solutions for video monitoring products, video monitoring products need to be packaged according to the ABCDNETS (network, cloud, data, intelligence, security, edge, end, chain) dimensions, such as Figure 9The main implementation steps are: 1. Analyze the functional requirements of the fusion monitoring product, identify the key function points in each technical field, and determine the atomic capabilities that need to be encapsulated; 2. Use containerization technology (such as Docker) to package each atomic capability and its dependencies, and encapsulate ABCDNETS atomic capabilities as independent and reusable services; 3. Deploy the containerized atomic capabilities to the cloud platform or edge node, and register the service address and metadata of the atomic capabilities in the service registry center; 4. Deploy the atomic capabilities to the appropriate runtime environment, use container orchestration tools such as Kubernetes to manage the deployment and expansion of atomic capabilities, configure load balancing and service discovery mechanisms, and ensure the high availability of atomic capabilities; 5. Initiate a call request to the atomic capabilities through the API gateway, and the API gateway routes the request to the corresponding atomic capability service according to the service discovery information, and composes the overall function of the fusion monitoring product; 6. Use workflow engine (such as Apache Airflow) to orchestrate the calling sequence of atomic capabilities, design workflow templates, dynamically generate workflow instances according to different business scenarios, and combine multiple atomic capabilities into complex workflows through orchestration tools; 7. Use monitoring tools (such as Prometheus) to collect performance metrics, monitor the performance of ABCDNETS atomic capabilities, analyze performance bottlenecks, and optimize the service implementation of atomic capabilities; 8. Build a video monitoring product atomic capability library, and the atomic capabilities in the library can be combined in a task-based manner according to customer requirements to form new specifications of video monitoring products.Among them, the most important is the second step of Docker packaging atomic ability: 1) Determine the containerization target: clearly identify the specific services or functions that need to be containerized in each technical field (such as AI), list all dependencies required for atomic ability running, including library files, configuration files, environment variables, etc.; 2) Create Dockerfile: write a Dockerfile for each atomic ability as a containerization configuration script, choose a suitable base image such as the official Python, Node.js image; 3) Install dependencies and add code: use the instructions in the Dockerfile to install the required dependencies, such as: scikit-learn, express, copy the source code of the atomic ability into the Docker container; 4) Configuration and environment settings: set the necessary environment variables in the Dockerfile, copy the configuration file into the container or pass it through the environment variable; 5) Network and storage configuration: configure the network settings of the container according to the network requirements of the atomic ability, if you need to persist data, use data volume mapping to store the path; 6) Define the startup command: define the command that runs when the container starts in the Dockerfile; 7) Build container image: use the Docker build command to build the container image according to the Dockerfile; 8) Image testing: run the container using the built image and test to ensure everything is working properly; 9) Push the image to the image repository: push the locally built image to Docker Hub or other container image repositories.Through this set of processes, a video monitoring product atomic capability library can be built. In terms of AI capabilities, the video monitoring product can provide AI video analysis capabilities such as face recognition, license plate recognition, and abnormal behavior detection. Combined with AI video analysis results, the corresponding alarm system can be triggered. In terms of chain capabilities (B), the video monitoring product can provide blockchain technology to record monitoring events, ensure that the records are tamper-proof, and automatically execute tasks related to monitoring, such as automatic alarm and automatic notification. In terms of cloud computing (C), the video monitoring product can provide remote access services, allowing users to remotely view monitoring videos through the cloud platform. It can also provide cloud storage solutions to support long-term storage and historical playback of videos. In terms of data capabilities (D), the video monitoring product supports user behavior analysis by collecting and analyzing user viewing habits to optimize the use experience of video monitoring and conduct big data analysis on video content to identify abnormal behavior. In terms of network capabilities (N), the video monitoring product can transmit data at high speed, support fast and stable transmission of high-definition video streams, and provide wired and wireless connection options. In terms of edge computing (E), the video monitoring product supports video analysis on cameras or local devices, reducing dependence on the cloud and supporting preliminary processing and storage of videos locally. In terms of terminal capabilities (T), the video monitoring product provides multi-terminal compatible services, supporting video monitoring on multiple devices including mobile devices and PC devices, ensuring compatibility with cameras of different brands and models. In terms of security capabilities (S), the video monitoring product supports data encryption to protect privacy through encryption technology during transmission and storage. It also provides security authentication services, and the product needs to pass security authentication such as anti-hacking and anti-cracking.

[0097] With the support of family portraits and product atomic capability library, the personalized video monitoring product scheme construction process is as follows Figure 10The proposed method is based on the concept of ABCDNETS, which is a set of atomic capabilities that can be combined to form a video surveillance product. The first step is to determine the potential needs based on the family portrait tags and perform task-based combination of ABCDNETS atomic capabilities to build a new specification of video surveillance products. Then, according to the mapping relationship between video surveillance product atomic capabilities and cross-ABCDNETS resource pool, relevant resources are scheduled to quickly complete the development of demonstration schemes and the deployment of related resources, which can give users a direct product experience and improve their experience and satisfaction. The process also supports adjusting product schemes based on customer feedback, responding to customer needs in a timely manner, and generating resource deployment schemes that meet customer's actual needs. The most important part of this process is the rapid design of resource schemes, which includes the following steps: 1. Resource assessment and capability analysis: assess the current state of computing resources, including computing power, storage capacity, network bandwidth, etc. At the same time, analyze the performance requirements and dependency relationships of atomic capabilities; 2. Design mapping model: use system modeling language (such as SysML) to create a mapping model that defines the logical relationship between atomic capabilities and resources. The model should include the type, quantity, performance indicators, and cost of resources; 3. Multi-objective optimization: build an optimization model that takes the resource requirements of atomic capabilities and the supply of computing resources as variables, and apply multi-objective genetic algorithms to search for the optimal solution; 4. Cloud computing services: adjust resource allocation based on the load of atomic capabilities, such as starting or terminating virtual machine instances, using the automatic scaling function of cloud services; 5. Security integration: add a security scanning step to the mapping algorithm, use tools such as OWASPZAP to perform security checks on configured resources, and adjust the mapping scheme based on the results; 6. Automation scripts: write Python scripts to call cloud service APIs and automatically configure or adjust resources based on the mapping results, such as allocating IP addresses, setting security rules, etc.; 7. Cost-benefit analysis model: build a cost model that takes into account resource costs, operating costs, expected returns, etc., and use financial analysis methods to calculate and compare the net present value (NPV) or internal rate of return (IRR) of different mapping schemes; 8. Disaster recovery and backup: develop data backup and disaster recovery strategies to improve the reliability and fault tolerance of resource schemes.The capability-resource mapping model built by SysML is indispensable in the rapid construction of resource solutions. The model construction process is as follows: 1) Create a system overview diagram: use the system overview diagram (SOD) of SysML to represent the high-level structure of the system, including the main atomic capabilities and resources, and show the composition and main interactions of the system; 2) Construct a block definition diagram: use the block definition diagram (BDD) of SysML to define the blocks in the system (such as AI services, databases, network connections, etc.) and their attributes and operations; 3) Describe the relationship between blocks: use the internal block diagram (IBD) of SysML to show the internal structure of the block, as well as the interface and dependency relationship between blocks, to show the connection and interaction between different blocks; 4) Model system behavior: use the sequence diagram (Sequence Diagram) of SysML to describe the sequence of events and the interaction between blocks, and the state diagram (State Diagram) to describe the lifecycle and state changes of the block, to describe the dynamic behavior of the system, including the activation of atomic capabilities, the allocation and release of resources, etc.; 5) Define requirements and constraints: use the requirement diagram (Requirement Diagram) of SysML to represent the requirements of the system, including performance indicators, safety standards, etc., and use the constraint diagram (Constraint Diagram) to represent the design constraints; 6) Verification and simulation: use simulation tools to simulate the behavior of the system according to the SysML model, check whether there are problems or conflicts, and verify the correctness and feasibility of the model; 7) Iteration and optimization: according to the simulation results and the feedback of stakeholders, adjust and optimize the model to better meet the requirements. Through the above process, the corresponding resource solution can be generated according to the actual needs of the customer, supporting the scheduling of ABCDNETS resources with the strategies of cost priority, performance priority, etc., so that the video monitoring product solution has diversity and flexibility, can be personalized adjusted according to the tag characteristics of different family groups, and maximizes the satisfaction of the actual video monitoring needs of the family. Not only can it realize the timely response to the video monitoring needs of the family, but also can ensure the efficiency and individualization of the whole service process, and ensure that the family can enjoy the efficient and safe one-stop video monitoring experience.

[0098] In summary, the method of the embodiment has the following characteristics:

[0099] 1) Dynamic family portrait tag matching system: Fusion of multi-dimensional family user information, use GMM and LSTM to build dynamic user portrait, and match appropriate attribute tags for each family member and the whole family, support to generate corresponding recommendation scheme according to family tag information;

[0100] 2) Video monitoring product atomic capability library: encapsulate the ABCDNETS atomic capabilities of the video monitoring product using Docker technology, build a cross-domain basic capability library, and the capabilities in the library can be combined in a task-based manner according to customer needs to form a new specification product, and support for atomic capability upgrade iteration;

[0101] 3) Product scheme rapid generation: according to customer real-time feedback, the actual product scheme can be quickly generated and the corresponding ABCDNETS resources can be matched according to the capability-resource mapping model constructed by SysML, supporting the generation of different specifications of recommended selection schemes according to customer needs to meet the diversified needs of customers;

[0102] 4) Product capability rapid demonstration: according to the actual needs of customers, the corresponding product capabilities (such as AI capabilities, alarm capabilities, etc.) can be quickly demonstrated and deployed, allowing customers to have a clear understanding of the required products and improving customer experience and satisfaction.

[0103] Corresponding to Figure 11 The method shown, another embodiment of the present application provides a video monitoring product pushing device. Figure 11 is a structural schematic diagram of the pushing device 1100, which includes:

[0104] The demand matching module 1110 determines the potential video monitoring demand of the target family based on the family portrait attributes of the target family.

[0105] The capability matching module 1120 determines at least one target atomic capability suitable for the potential video monitoring demand from the atomic capability library of the video monitoring service of the computing power network; the atomic capability refers to a basic independent capability.

[0106] The product configuration module 1130 configures a target video monitoring product run by the computing power network based on a container image created in advance for the at least one target atomic capability.

[0107] The product pushing module 1140 pushes the target video monitoring product to the target family.

[0108] The device of the embodiment decouples the video monitoring service of the computing power network in advance to construct an atomic capability library, and creates a corresponding container image for an atomic capability in the atomic capability library. When a video monitoring product needs to be customized for a target family, the potential video monitoring demand of the target family is analyzed based on the family portrait attribute of the target family, and at least one target atomic capability that matches the potential video monitoring demand is queried from the atomic capability library to configure a target video monitoring product suitable for the target family run by the computing power network. In related applications, since the atomic capability container image can be reused, the computing power network can customize more accurate and more personalized video monitoring products for different users through the atomic capability library to meet the video monitoring demands of different users.

[0109] Optionally, before determining the potential video monitoring demand of the target family based on the family portrait attribute of the target family, the demand matching module 1110 further performs: inputting the user data of each family member in the target family into a deep learning model to predict the user portrait attribute of each family member in the target family; wherein the deep learning model is trained based on the user data of a sample family member and a user portrait attribute label corresponding to the sample family member; and fusing the user portrait attribute of each family member in the target family to obtain the family portrait attribute of the target family.

[0110] Optionally, before inputting the user data of each family member in the target family into a deep learning model to predict the user portrait attribute of each family member in the target family, the demand matching module 1110 further performs: clustering the user data of candidate sample family members based on a Gaussian mixture model clustering algorithm to obtain a plurality of clustering clusters; determining whether there is a common user portrait attribute in each clustering cluster, and if so, taking the common user portrait attribute as the user portrait attribute label of the candidate sample family member corresponding to the clustering cluster; and selecting the sample family member from the candidate sample family members with the user portrait attribute label.

[0111] Optionally, the demand matching module 1110 clusters the user data of the candidate sample family members based on the Gaussian mixture model clustering algorithm, including: determining at least two Gaussian distributions corresponding to the user data of the candidate sample family members, each Gaussian distribution representing a clustering cluster; and iteratively determining the parameters of each Gaussian distribution based on an expectation maximization algorithm to determine the clustering cluster to which the user data of the candidate sample family members belongs.

[0112] Optionally, the demand matching module 1110 fuses the user portrait attributes of each family member in the target family to obtain the family portrait attributes of the target family, including: determining the weight of each family member of the target family based on the interaction information between the family members of the target family; and weighting and fusing the user portrait attributes of each family member of the target family according to the corresponding weights to obtain the family portrait attributes of the target family.

[0113] Optionally, after the product pushing module 1140 configures the at least one target atomic capability to run the target video monitoring product based on the computing power resources of the computing power network according to the monitoring product design rule, the product pushing module 1140 further performs: creating a mapping model of the computing power network based on a system modeling language, the mapping model containing the logical relationship between the atomic capabilities and resources of the computing power network; based on the mapping model, taking the resource demand of the at least one target atomic capability and the resource supply amount of the computing power network as input parameters, solving the resource configuration strategy of the computing power network for the target video monitoring product according to a target optimization direction to obtain a target resource configuration strategy; the target optimization direction is to minimize resource overhead or maximize target atomic capability performance; providing the target resource configuration strategy to the target family, and providing a demonstration of the target video monitoring product to the target family according to the target resource configuration strategy based on the computing power network.

[0114] Optionally, the apparatus of the embodiment further includes:

[0115] The container configuration module, before configuring the target video monitoring product running on the computing power network based on the container image created in advance for the at least one target atomic capability, performs: functionally decoupling the video monitoring service of the computing power network to obtain a plurality of atomic capabilities; creating a corresponding container file for each atomic capability; and adding the running configuration and running script on which each atomic capability depends to the corresponding container file to create a container image of each atomic capability.

[0116] It should be noted that the video monitoring product pushing method of the embodiment can be used as Figure 1 the execution subject of the method shown in the method, so as to realize the steps and functions in the method shown in the method, which will not be repeated here. Figure 1

[0117] Figure 12 is a structural schematic diagram of an electronic device provided by an embodiment of the present application. Please refer to Figure 12 ​At the hardware level, the electronic device includes a processor, and optionally further includes an internal bus, a network interface, and a memory. The memory can include a memory such as a random-access memory (RAM), and can further include a non-volatile memory such as at least one disk memory. Of course, the electronic device can further include other hardware required by a business.

[0118] The processor, the network interface, and the memory can be connected to each other through the internal bus, which can be an industry standard architecture (ISA) bus, a peripheral component interconnect (PCI) bus, or an extended industry standard architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, and a control bus, etc. For ease of representation, Figure 12 Only one bidirectional arrow is used in the figure to represent the bus, but it does not mean that there is only one bus or only one type of bus.

[0119] The memory is used to store a program. Specifically, the program can include program code including computer operation instructions. The memory can include a memory and a non-volatile memory, and provide instructions and data to the processor.

[0120] The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs, and forms the push device of the video monitoring product at the logical level. The processor executes the program stored in the memory, and is specifically used to perform the following operations:

[0121] Determine the potential video monitoring demand of the target family based on the family portrait attributes of the target family.

[0122] From the atomic capability library of the video monitoring service of the computing power network, determine at least one target atomic capability suitable for the potential video monitoring demand.

[0123] Based on the container image created in advance for the at least one target atomic capability, configure a target video monitoring product run by the computing power network.

[0124] Push the target video monitoring product to the target family.

[0125] The above as described in the present application Figure 2The method disclosed by the embodiment shown can be applied to a processor or implemented by the processor. The processor can be an integrated circuit chip with processing capability of signals. In the implementation process, each step of the above method can be completed by integrated logic circuits of hardware in the processor or instructions in the form of software. The above processor can be a general processor, including a central processing unit (CPU), a network processor (NP), etc.; or a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The disclosed methods, steps and logic block diagrams in one or more embodiments of the present application can be implemented or executed. The general processor can be a microprocessor or the processor can also be any conventional processor. The steps of the method disclosed in combination with one or more embodiments of the present application can be directly embodied as a hardware code processor for execution, or a combination of hardware and software modules in the code processor for execution. The software module can be located in a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register or other mature storage medium in the art. The storage medium is located in the memory, and the processor reads the information in the memory and combines the hardware to complete the steps of the above method.

[0126] Of course, in addition to the software implementation, the electronic device of the present application does not exclude other implementation manners, such as logic devices or a combination of software and hardware, etc. That is, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0127] The embodiment of the present application further proposes a computer program product, which includes a computer readable storage medium storing a computer program, and the computer program is operable to make a computer execute the following operations:

[0128] Based on the household portrait attribute of the target family, the potential video monitoring demand of the target family is determined.

[0129] From the atomic ability library of the video monitoring service of the computing power network, at least one target atomic ability suitable for the potential video monitoring demand is determined.

[0130] Based on the container image created in advance for the at least one target atomic ability, a target video monitoring product running by the computing power network is configured.

[0131] The target video monitoring product is pushed to the target family.

[0132] In conclusion, the above only describes the preferred embodiments of the present application, and is not intended to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of the present application shall be included in the protection scope of one or more embodiments of the present application.

[0133] The system, device, and module illustrated in the above embodiments can be specifically implemented by a computer chip or entity, or by a product with certain functions. A typical implementation device is a computer. Specifically, the computer may, for example, be a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0134] The computer readable medium includes permanent and non-permanent, removable and non-removable media, which can be implemented by any method or technology to store information. The information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read only memory (ROM), electrically erasable programmable read only memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic tape disk storage or other magnetic storage device, or any other non-transmission medium that can be used to store information accessible by a computing device. According to the definition herein, computer readable medium does not include transitory media such as modulated data signals and carriers.

[0135] It should also be noted that the term "comprising" or "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that processes, methods, articles or devices including a series of elements not only include those elements, but also include other elements not explicitly listed or inherent to such processes, methods, articles or devices. Without more limitations, the element defined by the statement "including a" does not exclude the presence of another identical element in the process, method, article or device including the element.

[0136] The various embodiments in this application are described in a progressive manner, and the same or similar parts among the various embodiments can be referred to each other. Each embodiment focuses on the difference from other embodiments. In particular, the system embodiments are described in a relatively simple manner, and the relevant parts can be referred to the description of the method embodiments.

Claims

1. A method of pushing a video surveillance product, characterized by, The method comprises: determining potential video monitoring needs of a target family based on family portrait attributes of the target family; determining at least one target atomic capability suitable for the potential video monitoring needs from an atomic capability library of video monitoring services of a computing power network; the atomic capability refers to a basic independent capability; configuring a target video monitoring product run by the computing power network based on a container image previously created for the at least one target atomic capability; pushing the target video monitoring product to the target family; After the at least one target atomic capability is configured into the target video monitoring product run by the computing power resources of the computing power network according to monitoring product design rules, the method further comprises: creating a mapping model of the computing power network based on a system modeling language, the mapping model containing logical relationships between atomic capabilities and resources of the computing power network; solving a resource configuration strategy of the computing power network for the target video monitoring product according to a target optimization direction based on the mapping model, resource requirements of the at least one target atomic capability, and resource supply of the computing power network as input parameters, to obtain a target resource configuration strategy; the target optimization direction is to minimize resource overhead or maximize performance of the target atomic capability; providing the target resource configuration strategy to the target family, and providing a demonstration of the target video monitoring product to the target family based on the computing power network according to the target resource configuration strategy.

2. The method of claim 1, wherein, before determining potential video monitoring needs of a target family based on family portrait attributes of the target family, the method comprises: inputting user data of each family member in the target family into a deep learning model to predict user portrait attributes of each family member in the target family; wherein the deep learning model is trained based on user data of sample family members and user portrait attribute labels corresponding to the sample family members; fusing user portrait attributes of each family member in the target family to obtain family portrait attributes of the target family.

3. The method of claim 2, wherein, before inputting user data of each family member in the target family into a deep learning model to predict user portrait attributes of each family member in the target family, the method further comprises: clustering user data of candidate sample family members based on a Gaussian mixture model clustering algorithm to obtain multiple clustering clusters; determining whether each clustering cluster has common user portrait attributes, and if so, taking the common user portrait attributes as user portrait attribute labels of the candidate sample family members corresponding to the clustering cluster; selecting the sample family members from the candidate sample family members with user portrait attribute labels.

4. The method of claim 3, wherein, clustering user data of candidate sample family members based on a Gaussian mixture model clustering algorithm comprises: determining at least two Gaussian distributions corresponding to the user data of the candidate sample family members, each Gaussian distribution representing a clustering cluster. ​ ​ ​ Parameters of each Gaussian distribution are iterated based on an expectation maximization algorithm to determine a cluster to which user data of the candidate sample family member belongs.

5. The method of claim 2, wherein, fusing user portrait attributes of each family member in the target family to obtain family portrait attributes of the target family, including: determining weights of each family member of the target family based on interaction information between family members of the target family; weighting and fusing user portrait attributes of each family member of the target family according to corresponding weights to obtain family portrait attributes of the target family.

6. The method of any one of claims 1 to 5, wherein, before configuring a target video monitoring product run by the computing power network based on a container image created in advance for the at least one target atomic capability, the method further includes: functionally decoupling video monitoring services of the computing power network to obtain a plurality of atomic capabilities; creating a corresponding container file for each atomic capability; adding a running configuration and a running script on which each atomic capability depends to the corresponding container file to create a container image of each atomic capability.

7. A push device for video surveillance products, characterized in that including: a demand matching module that determines potential video monitoring demands of a target family based on family portrait attributes of the target family; an ability matching module that determines at least one target atomic capability suitable for the potential video monitoring demands from an atomic capability library of video monitoring services of the computing power network; the atomic capability refers to a basic independent capability; a product configuration module that configures a target video monitoring product run by the computing power network based on a container image created in advance for the at least one target atomic capability; a product pushing module that pushes the target video monitoring product to the target family; after configuring the at least one target atomic capability into a target video monitoring product run based on computing power resources of the computing power network according to monitoring product design rules, creating a mapping model of the computing power network based on a system modeling language, the mapping model containing a logical relationship between atomic capabilities and resources of the computing power network; based on the mapping model, taking resource demand of the at least one target atomic capability and resource supply amount of the computing power network as input parameters, solving a resource configuration strategy of the computing power network for the target video monitoring product according to a target optimization direction to obtain a target resource configuration strategy; the target optimization direction is to minimize resource overhead or maximize performance of the target atomic capability; providing the target resource configuration strategy to the target family and providing a demonstration of the target video monitoring product to the target family based on the computing power network according to the target resource configuration strategy.

8. An electronic device, comprising: a processor; and a memory arranged to store computer-executable instructions that, when executed, cause the processor to perform the method of any one of claims 1 to 6. a processor; and a memory arranged to store computer-executable instructions that, when executed, cause the processor to perform the method of any one of claims 1 to 6.

9. A computer program product comprising a computer readable storage medium storing a computer program operable to cause a computer to perform the method of any one of claims 1 to 6.

Images