A crowdsourcing mapping system and method based on cloud-edge

Abstract

The present invention discloses a cloud-edge-based crowdsourcing surveying and mapping system and method, wherein the crowdsourcing surveying and mapping system includes a cloud, an edge and terminal users, and the crowdsourcing surveying and mapping method is implemented based on the crowdsourcing surveying and mapping system; the technical solution of the present invention performs multi-layer dimensionality reduction processing on geographic information data, disassembles complex large-volume surveying and mapping tasks into lightweight task packages with low difficulty and low threshold, and uses random transformation matrix encryption technology to ensure the security and privacy protection of data during transmission and processing; the technical solution of the present invention effectively solves the two major bottlenecks of information security and professionalism commonly faced in the current surveying and mapping production model, and on the basis of ensuring information security, makes full use of social resources to realize the crowdsourcing completion of large-scale complex and professional work, providing a set of efficient and secure collaborative solutions for the surveying and mapping geographic information industry, and constructs a new flexible production model for surveying and mapping.

Description

Technical Field

[0001] The present invention belongs to the technical field of surveying and mapping architecture and surveying and mapping methods, and specifically relates to a crowdsourcing surveying and mapping system and method based on cloud edge. Background Art

[0002] With the continuous expansion of surveying and mapping geographic information applications, its service areas cover many fields such as urban planning, environmental monitoring and resource management, which has led to an explosive growth in the demand for various types of geographic information data products. At present, the traditional geographic information data collection method mainly relies on the centralized operation of professional technicians and equipment in a specific physical environment. This model not only requires the investment of a large number of technicians and professional equipment, but also has cumbersome and inefficient management processes, making it difficult to quickly respond to large-scale and large-volume data product needs.

[0003] In addition, when it comes to tasks involving multiple processes and scenarios, the traditional model is often limited by established technical processes, making it difficult to flexibly expand and unable to adapt to diverse surveying and mapping data service tasks. As a result, crowdsourcing surveying and mapping has emerged. Crowdsourcing surveying and mapping refers to the use of crowdsourcing models to assign surveying and mapping tasks to a large number of dispersed individuals or groups through a network platform, who use their own equipment and time to complete data collection, processing and analysis. Crowdsourcing surveying and mapping combines the Internet and mobile device technology to efficiently, cheaply and widely acquire and process geographic information data.

[0004] At present, crowdsourcing surveying and mapping has made some progress in theoretical research. For example, Tang Luliang proposed a theoretical research framework system for crowdsourcing surveying and mapping of high-temporal and spatial precision road information based on GNSS trajectory big data, and Wu Xinyu proposed a large-scale topographic map update solution based on crowdsourcing surveying and mapping mode. At the industry application level, the foreign "Map Wikipedia" Open Street Map successfully built a global editable map through the re-reporting model, and it is mainly used in the fields of navigation electronic maps and high-precision map production in China, such as AutoNavi's development of "AutoNavi Gold Rush" to collect and update navigation electronic maps.

[0005] These studies and applications are mainly concentrated in specific scenarios such as electronic map data collection and update, and have application limitations. At the same time, they are restricted by the bottlenecks of quality control and data confidentiality. At present, they mainly handle single, low-precision tasks, but are difficult to handle complex, high-precision data tasks. Summary of the invention

[0006] In order to overcome the shortcomings of the prior art, the purpose of the present invention is to provide a cloud-edge-based crowdsourcing mapping system and method that can handle complex and high-precision data tasks.

[0007] The purpose of the present invention is achieved through the following technical solutions:

[0008] A cloud-edge-based crowdsourcing mapping system, characterized by comprising a cloud, an edge, and a terminal user;

[0009] The cloud includes a geographic information data dimension reduction processing module, a data encryption and security management module, an intelligent scheduling module, a real-time monitoring module, and a data integration and quality control module;

[0010] The edge end includes a task allocation and feedback module, a dynamic scheduling management module, a quality and safety management module, and an up-down feedback coordination module; the edge end is responsible for receiving task packages distributed from the cloud, and performing specific task allocation, dynamic scheduling and quality management of the task packages;

[0011] The terminal user includes a data download and acquisition module, a tool embedding and optimization module, a quality control and data return module, and a real-time feedback and support module; the terminal user is responsible for the specific execution of the task package.

[0012] A cloud-edge-based crowdsourcing mapping method, using the above-mentioned cloud-edge-based crowdsourcing mapping system, includes the following steps:

[0013] S10: In the cloud, the geographic information data dimensionality reduction processing module analyzes the original task and extracts key processes and key elements, builds a decomposition model, and performs dimensionality reduction processing on the geographic information data, thereby reducing the dimensionality of data tasks with large data volumes and high professional thresholds into lightweight task packages that are easy to use and easy to handle; the data encryption and security management module encrypts the task packages using random transformation matrix encryption technology; the intelligent scheduling module distributes each task package after dimensionality reduction processing and data encryption to the edge end with matching processing capabilities through resource evaluation algorithms and intelligent scheduling algorithms;

[0014] S20: After receiving the task package, the task allocation and return module at the edge end performs task analysis on the received task package, establishes an adaptive resource allocation module, automatically calculates the matching degree of the corresponding terminal user, and then allocates the task package to the terminal user with the appropriate matching degree for effective execution; and returns the terminal data that has been executed and passed the quality control to the cloud;

[0015] S30: The data download and collection module of the terminal user receives the specific task package and task analysis assigned by the edge end, and then collects geographic information data according to relevant requirements (such as change spot extraction), selects or optimizes the corresponding tool through the tool embedding and optimization module, and uses the relevant tool to execute the corresponding task package. The quality control and data return module performs automatic quality inspection on the collected information data according to the quality control model provided by the edge end, and returns the data that passes the quality inspection to the edge end; at the same time, the real-time feedback and support module is used to feedback problems to the edge end in real time and obtain technical support;

[0016] S40: The up-down feedback coordination module at the edge collects the execution status and feedback information of the terminal users in real time, so that the edge can dynamically schedule matching resources and continuously iterate and optimize the adaptive resource allocation model. At the same time, it regularly reports the task execution data and available resource data to the cloud, so that the cloud can dynamically schedule other matching edge terminals according to actual needs.

[0017] S50: The real-time monitoring module in the cloud monitors the execution process of the task package in real time and makes necessary interventions based on the data reported by the upper and lower feedback collaboration modules; at the same time, the data integration and quality control module integrates the data sent back by the task package and performs overall quality control on the integrated data to ensure that the geographic information data products meet the predetermined quality standards.

[0018] The present invention has the following beneficial effects:

[0019] 1. Compared with conventional surveying and mapping schemes in the prior art, the crowdsourcing surveying and mapping system based on the cloud-edge terminal of the present invention can use the geographic information data dimensionality reduction processing module to realize dimensionality reduction processing of relevant data in the original task (various geographic information data, such as image data, terrain data, natural geography-related thematic data, attribute data and metadata, etc.), so as to reduce the data tasks with large data volume and high professional threshold into low-dimensional, easy-to-process and lightweight data task packages, and then use the intelligent scheduling module to distribute each task package to the edge terminal. During the distribution process, the data encryption and security management module independently and randomly encrypts each task package to ensure the security of each task package during transmission and processing. Therefore, it can realize efficient processing of large-volume, high-difficulty and high-complexity data tasks, and at the same time solve the security problems of geographic information during transmission and processing through the unique matching random encryption matrix technology.

[0020] 2. The crowdsourcing surveying and mapping method of the present invention is implemented through steps S10 to S50. In S10, the method first analyzes and performs multi-layer dimensionality reduction processing on the original task through the geographic information data dimensionality reduction processing module to disassemble and reduce the complex large-scale surveying and mapping task package into a lightweight task package with low difficulty and low threshold, and then completes the task execution of the task package through steps S20 to S40, and finally integrates the data of the returned task package through S50, and performs overall quality control on the integrated data to ensure that the geographic information data product meets the predetermined quality standards. This crowdsourcing surveying and mapping method based on the cloud-edge terminal realizes the crowdsourcing completion of large-scale complex professional geographic information work, provides a set of efficient and secure collaborative solutions for the surveying and mapping geographic information industry, and builds a new flexible production model for surveying and mapping. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1This is the overall architecture of the cloud-edge-based crowdsourcing mapping of the present invention.

[0022] Figure 2 This is a flow chart of the cloud-edge-based crowdsourcing mapping method of the present invention.

[0023] Figure 3 This is the main process of dimension reduction of geographic information data in the present invention.

[0024] Figure 4 This is a schematic diagram of the outer expansion ring topology fidelity editing technology of the present invention.

[0025] Figure 5 This is a flow chart of the "one package one key" random transformation matrix encryption technology of the present invention. DETAILED DESCRIPTION

[0026] In the following, only some exemplary embodiments are briefly described. As those skilled in the art will appreciate, the described embodiments may be modified in various ways without departing from the spirit or scope of the present application. Therefore, the drawings and descriptions are considered to be exemplary and non-restrictive in nature.

[0027] The present invention provides a cloud-edge-based crowdsourcing mapping system and method, which performs multi-layer dimensionality reduction processing on geographic information data to break down complex, large-scale mapping tasks into lightweight task packages with low difficulty and low threshold, and uses random transformation matrix encryption technology to ensure the security and privacy protection of data during transmission and processing. The technical solution of the present invention can effectively solve the two major bottlenecks of information security and professionalism commonly faced in the current mapping production model. On the basis of ensuring information security, it makes full use of social resources to realize crowdsourcing tasks of large-scale, complex and professional work, providing a set of efficient and secure collaborative solutions for the mapping and geographic information industry, and building a new flexible production model for mapping.

[0028] The technical solution of the present invention will be described in detail below through the crowdsourcing surveying and mapping system of Example 1 and the crowdsourcing surveying and mapping method of Example 2:

[0029] Example 1

[0030] A crowdsourcing mapping system based on cloud-edge terminals, such as Figure 1 As shown in the figure, it includes the cloud, edge, and end users. The cloud-edge-end structure is an important distributed computing model architecture in the current computing system. It emphasizes the reasonable allocation of data processing, storage, and application services between the cloud, edge, and terminal devices. The three levels work together to complete complex tasks. This paper organically combines the cloud-edge-end architecture with crowdsourcing mapping, and designs a crowdsourcing mapping service solution that is dynamically responsive, resource-efficient, and quality-secure.

[0031] As the overall data center and core management and dispatching center of the crowdsourcing surveying and mapping system, the cloud makes full use of powerful processing capabilities and vast storage space. The cloud includes geographic information data dimensionality reduction processing modules, data encryption and security management modules, intelligent scheduling modules, real-time monitoring modules, data integration and quality control modules, so that the cloud has geographic information data dimensionality reduction capabilities, data encryption capabilities, security management capabilities, intelligent scheduling capabilities, real-time monitoring capabilities, data integration capabilities and quality control capabilities.

[0032] The following is a detailed introduction to each module of the cloud:

[0033] The geographic information data dimensionality reduction processing module is responsible for automatically analyzing and extracting key processes and key elements according to the original task content, technical standards, data requirements and application scenarios, using artificial intelligence, topological analysis, and multi-scale analysis technology, simplifying data structure and attribute information, and optimizing spatial and temporal data expression. It gradually simplifies large-volume, high-dimensional, and multi-attribute geographic information data tasks into lightweight, low-dimensional, and easy-to-process lightweight task packages; in this way, it greatly optimizes the cloud task crowdsourcing process, and mainly includes data volume dimensionality reduction, spatial dimensionality reduction, semantic dimensionality reduction, and temporal dimensionality reduction on the basis of ensuring data consistency and integrity during task package distribution and return.

[0034] The data encryption and security management module is responsible for data encryption and decryption, authority management and access control of the task package. By designing random transformation matrix encryption technology (also known as "one package one key" random transformation matrix encryption technology, "one package one key" means one task package corresponds to a unique key), the encryption and decryption mechanism of random transformation matrix encryption technology is used to ensure the security and confidentiality of the geographic location data of the task package. At the same time, by building a permission management system, role management and reasonable allocation of permissions for various types of users are provided, and by integrating third-party identity authentication services, the security of data flow and storage is ensured.

[0035] The intelligent scheduling module is responsible for intelligently allocating task packages, that is, it is responsible for distributing and intelligently scheduling the task packages that have undergone dimensionality reduction processing and data encryption according to the resource situation at the edge. Specifically, by designing resource evaluation algorithms and intelligent scheduling algorithms, the resource evaluation algorithm can evaluate the advantages and limitations of available resources, and distribute task packages to the best processing edge nodes based on the characteristics of the task packages and the resource evaluation results, so as to achieve the effective load of the edge nodes and the priority management of the task packages. According to the real-time execution monitoring situation, tasks can be dynamically reallocated or task priorities can be adjusted. At the same time, the execution information of the edge team is regularly integrated to update the intelligent scheduling algorithm and improve the next round of task decomposition and intelligent allocation strategies.

[0036] The real-time monitoring module is responsible for monitoring the task execution process and making necessary interventions. It tracks and monitors the task packages assigned by the intelligent scheduling module in real time, including the task execution progress and resource usage of the edge team, and feeds back the monitored risks and problems to the intelligent scheduling module in real time, dynamically scheduling matching resources for corresponding remediation or optimizing task execution.

[0037] The data integration and quality control module is responsible for decrypting and integrating the data of each returned task package, and performing overall quality control on the integrated data products to ensure complete geographic information data products that meet the predetermined quality standards.

[0038] As the middle management layer and the upper and lower coordination layer, the edge is responsible for receiving the task packages distributed from the cloud, and performing specific task allocation, dynamic scheduling and quality management of the task packages. Among them, the edge includes task allocation and feedback module, dynamic scheduling management module, quality and safety management module, and upper and lower feedback coordination module; so that the edge has task package reception and allocation, dynamic scheduling management task package, safety management task package, and upper and lower feedback coordination.

[0039] The following is a detailed introduction to each module on the edge:

[0040] The task allocation and feedback module is responsible for receiving the complete task package from the cloud, performing task analysis, understanding the specific requirements, priorities and resource requirements of the task, and automatically calculating the matching degree of subordinate terminal users by establishing an adaptive resource allocation model based on terminal user capabilities, historical execution data, task requirements and other indicators, and further allocating the task package to the appropriate terminal user for effective execution. The task allocation and feedback module will also transmit the terminal feedback data with quality control back to the cloud.

[0041] The dynamic scheduling management module is responsible for real-time monitoring and recording of the execution progress of terminal users, timely warning of discovered progress risks, real-time calculation of terminal resources that match the evaluation capabilities, and dynamic scheduling of available matching resources as needed, including reallocation of task packages or increasing investment in matching terminals to ensure that tasks are completed on time.

[0042] The quality and safety management module builds a standardized quality control model based on the technical quality requirements of the task packages sent from the cloud, implements strict and unified quality management on subordinate terminal users, and ensures that the data returned by the terminals meets the technical quality standards; the quality and safety management module also establishes a role-based access control and security authentication mechanism for subordinate terminal users to effectively prevent unauthorized data access and further strengthen data security protection.

[0043] The upper and lower feedback coordination module is used for the edge to regularly report task execution data and available resource data to the cloud, so that the cloud can dynamically schedule other matching edge teams according to actual needs; the edge collects the execution status and feedback information of terminal users in real time, including task execution progress, work difficulties and other issues, so as to facilitate the dynamic scheduling and matching of resources at the edge, and continuously iterate and optimize the adaptive resource allocation model. At the same time, a timestamp, version control and data synchronization mechanism are established in the overall architecture to ensure the consistency of data on the cloud, edge and end system architecture.

[0044] As the front line of crowdsourcing surveying and mapping data production, the end user is responsible for the specific execution of the task package. The end user includes data download and acquisition module, tool embedding and optimization module, quality control and data return module, and real-time feedback and support module.

[0045] The following is a detailed introduction to each module for end users:

[0046] The data download and collection module is used to receive specific task packages and task analysis assigned by the edge end, and collect geographic information data according to relevant requirements (such as change map extraction).

[0047] The tool embedding and optimization module is used to provide embeddable, installation-free, and easy-to-operate collection tools, which greatly reduces the specialized threshold and work difficulty, effectively improves the operating efficiency of end users, and ensures the quality of production data.

[0048] The quality control and data feedback module performs automated quality checks on the collected data through the quality control model provided by the edge, and automatically transmits the passed data back to the edge according to the specified rules.

[0049] The real-time feedback and support module provides real-time feedback and technical support to help end users solve problems encountered during the production process.

[0050] To summarize, the cloud-edge-based crowdsourcing mapping system of the present invention sets up a cloud, an edge, and terminal users. The cloud mainly realizes its corresponding functions through various modules, and uses the geographic information data dimensionality reduction processing module to realize dimensionality reduction processing of relevant data in the original task (various geographic information data, such as image data, terrain data, natural geography-related thematic data, attribute data, and metadata, etc.), so as to reduce the dimension of data tasks with large data volume and high professional threshold into low-dimensional, easy-to-process, and lightweight data task packages, and then use the intelligent scheduling module to distribute multiple task packages to the edge. During the distribution process, the data encryption and security management module independently and randomly encrypts each task package to ensure the security of each task package during transmission and processing; the real-time monitoring module monitors the execution process of the task package in real time according to the data reported by the upper and lower feedback collaboration modules, and performs necessary intervention; the data integration and quality control module integrates the data returned by the task package, and performs overall quality control on the integrated data to ensure that the geographic information data product meets the predetermined quality standards.

[0051] The edge end implements its corresponding functions by setting up various modules, uses the task allocation and feedback module to receive the encrypted task package processed by the cloud after dimensionality reduction, and then distributes the task package to the terminal users with matching capabilities, so as to use the terminal users to execute or process the tasks of the task package. In this process, the dynamic scheduling management module monitors the progress of the terminal users in executing or processing the task package in real time, and calculates the capabilities of each terminal user at the same time, so as to continuously optimize the allocation of task packages. In addition, the edge end, as the middle architecture of the entire crowdsourcing mapping system, also plays a connecting role, which is mainly realized through the upper and lower feedback collaboration module. The upper and lower feedback collaboration module is used to collect feedback information from terminal users in real time, so that the edge end can dynamically schedule more matching resources, and continuously iterate and optimize the adaptive resource allocation model. The upper and lower feedback collaboration module is also used to regularly report task execution data and available resource data to the cloud, so that the cloud can dynamically schedule other matching edge ends (edge ​​teams) according to actual needs.

[0052] The terminal user implements the corresponding functions by setting up each module, uses the data download and collection module to receive the specific task package assigned by the edge, and then collects the relevant data or information in the task package; the tool embedding and optimization module can embed the required tools as needed, such as data editing tools, and support the optimization and upgrade of related tools to keep the terminal user's processing tools in the best state at all times; the quality control and data feedback module can be used to perform automated quality checks on the task package data based on the quality control module provided by the edge, and the processing progress or status of the task package can be fed back to the edge in real time for feedback; the real-time feedback and support module can provide problem feedback and technical support to help terminal users solve problems encountered in the production process.

[0053] Compared with conventional surveying and mapping schemes in the prior art, the crowdsourcing surveying and mapping system based on the cloud-edge terminal of the present invention can utilize the geographic information data dimensionality reduction processing module to realize dimensionality reduction processing of relevant data in the original task (various geographic information data, such as image data, terrain data, natural geography-related thematic data, attribute data and metadata, etc.), so as to reduce the data tasks with large data volume and high professional threshold into low-dimensional, easy-to-process and lightweight data task packages, and then utilize the intelligent scheduling module to distribute each task package to the edge terminal. During the distribution process, the data encryption and security management module independently and randomly encrypts each task package to ensure the security of each task package during transmission and processing. Therefore, it can realize efficient processing of large-volume, high-difficulty and high-complexity data tasks, and at the same time solve the security problems of geographic information during transmission and processing through the unique matching random encryption matrix technology.

[0054] Example 2

[0055] A crowdsourcing mapping method based on cloud-edge, such as Figure 1 and Figure 2 As shown, the cloud-edge-end based crowdsourcing mapping system of Example 1 includes the following steps:

[0056] S10: In the cloud, the geographic information data dimensionality reduction processing module analyzes the original task and extracts key processes and key elements, builds a decomposition model, and performs dimensionality reduction processing on the geographic information data, thereby reducing the dimensionality of data tasks with large data volumes and high professional thresholds into lightweight task packages that are easy to use and easy to handle; the data encryption and security management mode encrypts the task package through the "one package, one key" random transformation matrix encryption technology; the intelligent scheduling module distributes each task package after dimensionality reduction processing and data encryption to the edge end with matching processing capabilities through resource evaluation algorithms and intelligent scheduling algorithms.

[0057] S20: After receiving the task package, the task allocation and feedback module at the edge end performs task analysis on the received task package, automatically calculates the matching degree of the corresponding terminal user through the established adaptive resource allocation module, and then allocates the task package to the terminal user with the appropriate matching degree for effective execution; at the same time, the terminal data that has been executed and passed quality control is transmitted back to the cloud.

[0058] S30: The data download and collection module of the terminal user receives the specific task package and task analysis assigned by the edge end, and then collects geographic information data according to relevant requirements (such as change map extraction), selects or optimizes the corresponding tools through the tool embedding and optimization module, and uses the relevant tools to execute the corresponding task package. The quality control and data return module performs automated quality inspection on the collected information data according to the quality control model provided by the edge end, and returns the data that has passed the quality inspection (referred to as quality inspection) to the edge end; at the same time, the real-time feedback and support module is used to provide real-time feedback to the edge end and obtain technical support.

[0059] S40: The up-down feedback collaboration module at the edge collects the execution status and feedback information of the terminal users in real time so that the edge can dynamically schedule matching resources and continuously iterate and optimize the adaptive resource allocation model. The up-down feedback collaboration module also regularly reports the task execution data and available resource data to the cloud so that the cloud can dynamically schedule other matching edge terminals according to actual needs.

[0060] S50: The real-time monitoring module in the cloud monitors the execution process of the task package in real time and makes necessary interventions based on the data reported by the upper and lower feedback collaboration modules; at the same time, the data integration and quality control module integrates the data of the returned task package and performs overall quality control on the integrated data to ensure that the geographic information data products meet the predetermined quality standards.

[0061] Therefore, the crowdsourcing surveying and mapping method of the present invention is implemented through steps S10 to S50. In S10, the method first analyzes and performs multi-layer dimensionality reduction processing on the original task through the geographic information data dimensionality reduction processing module to disassemble and reduce the complex large-scale surveying and mapping task package into a lightweight task package with low difficulty and low threshold, and then completes the task execution of the task package through steps S20 to S40, and finally integrates the data of the returned task package through S50, and performs overall quality control on the integrated data to ensure that the geographic information data product meets the predetermined quality standards. This crowdsourcing surveying and mapping method based on the cloud edge realizes the crowdsourcing completion of large-scale complex professional geographic information work, provides a set of efficient and secure collaborative solutions for the surveying and mapping geographic information industry, and builds a new flexible production model for surveying and mapping.

[0062] Based on the above general description of the crowdsourcing surveying and mapping method of the present invention, the key technologies in specific steps are further introduced in detail below:

[0063] In the S10, when reducing the dimension of the original task into a lightweight task package that is easy to use and process in the geographic information data dimension reduction processing module, the following method is used:

[0064] 1. Dimensionality reduction processing technology for geographic information data

[0065] like Figures 1 to 3 As shown in the figure, the dimension reduction processing module of geographic information data automatically analyzes and extracts key processes and key elements through AI analysis (also known as artificial intelligence), aggregation analysis or multi-scale analysis, combined with the technical process of the original task, and flexibly combines classification conversion technology and time series clustering technology to design standardized and refined decomposition models, simplify data structure and attribute information, optimize spatial and temporal data expression, and gradually simplify large-scale, high-dimensional, multi-attribute geographic information data tasks into lightweight, low-dimensional, easy-to-process lightweight task packages. The decomposition model mainly includes data volume dimension reduction, spatial dimension reduction, semantic dimension reduction and temporal dimension reduction. Therefore, using the dimension reduction processing module of geographic information data to reduce the dimension of the task package can greatly optimize the cloud task package process, and improve the efficiency and accuracy of complex geographic information data task processing on the basis of ensuring the consistency and integrity of data in the distribution and return process of the task package.

[0066] At the same time, the geographic information data dimensionality reduction processing module also designs an outer ring topology fidelity editing technology to ensure the consistency of the topological structure of each task package throughout the entire execution process.

[0067] The dimension reduction method adopted by the geographic information data dimension reduction processing module adopts the following method:

[0068] 1) Data reduction: Use AI algorithms (artificial intelligence algorithms) to analyze the data of key processes and key elements, extract important geographic information elements that require manual processing, and distribute the streamlined data to end users for processing, reducing the data load and redundancy of end users and improving end user processing efficiency.

[0069] 2) Spatial dimensionality reduction: Based on the specific process of the production task of the original task, that is, the specific process of the original task data, combined with multi-scale analysis and spatial clustering technology (aggregation technology), the spatial dimension of the original task is gradually simplified, the technical threshold and difficulty of the end user are reduced, and the processing efficiency of the end user is improved.

[0070] 3) Semantic dimensionality reduction: The complex attribute codes and related information of the original task are simplified and classified through classification conversion technology, converted into easy-to-understand text descriptions, and classified according to the text descriptions and provided to the edge end (also called the edge end team) who are familiar with the relevant attributes, thereby improving the comprehensibility and operability of the task package and lowering the threshold for entry for end users.

[0071] 4) Time dimension reduction: Through time series clustering and other technologies, the time series data of the original task is simplified, key time point data is extracted, and then distributed to appropriate edge terminals and end users to reduce the complexity and difficulty of end users' tasks and improve the processing efficiency of end users.

[0072] When the geographic information data dimension reduction processing module realizes the crowdsourcing surveying and mapping method of maintaining the topological structure of the task package through the outer expansion ring topology fidelity editing technology, the following method is adopted:

[0073] Aiming at the maintenance of topological structure in large-scale feature editing, the present invention designs an outward expansion ring topological fidelity editing technology in the process, which ensures the consistency of the topological structure before and after editing by expanding the boundaries of the ground feature elements.

[0074] In this process, if Figure 4 As shown in the figure, through topological technology, the boundaries of the terrain features that need to be edited in the original task are expanded. The area with expanded boundaries is called the expansion ring. In principle, the expansion ring area does not need to be edited, ensuring that the overall topological structure will not be destroyed due to local modifications after the internal features are edited, so that the returned terrain features can be seamlessly connected with the original topological structure.

[0075] Specifically, when an element in the outer expansion ring of a task package is identified by the end user as needing to be edited, the outer expansion ring is marked, and the editing operation is temporarily not performed in the current task package. After the task package is completed and transmitted back to the cloud and passes the topological verification of data integration, the system automatically performs dynamic task regeneration for the marked outer expansion ring elements and generates a new task package to handle the editing task of the outer expansion elements in the marked outer expansion ring.

[0076] For ease of understanding, combined Figure 4 As shown, the boundary of the expanded element is defined as: P′ i =P i +△n i , where P i is the boundary point coordinate vector of the original feature of the original task; P′ i is the coordinate vector of the boundary point after expansion. △ is the expansion distance (constant), indicating the expansion amplitude. i is the unit normal vector at the boundary point, indicating the expansion direction.

[0077] Among them, when the task package is completed and transmitted back to the cloud, the following topology verification formula can be used for data integration topology verification:

[0078]

[0079] Among them, P' i and P i They represent the boundary points of the features before and after editing respectively; ε is the topological one-time threshold, which is used to determine whether the difference in topological structure before and after editing is within an acceptable range.

[0080] At the same time, in some special cases, the elements in the outer expansion ring may need to be edited, which may cause the complexity of the topological relationship. To solve this problem, the present invention further designs a delay processing and dynamic task regeneration mechanism based on the outer expansion ring topology fidelity editing technology.

[0081] Under this mechanism, when an element in the outer expansion ring of a task package is identified by the end user as needing to be edited, the outer expansion ring is marked, for example, marked as "delayed processing", and the editing operation is temporarily not performed in the current task package. After the task package is completed and transmitted back to the cloud, and after the topological verification of data integration is passed, the system automatically performs dynamic task regeneration for the outer expansion ring elements with the mark, and generates a new task package to handle the outer expansion element editing task in the marked outer expansion ring (that is, the outer expansion ring marked for delayed processing). Through this mechanism, it is ensured that even if complex ground features are processed in multiple rounds of tasks, the overall topological consistency of the ground features can still be maintained; at the same time, this technology can effectively avoid topological errors and confusion in the process of large-scale feature editing, simplify topological verification, reduce the difficulty and complexity of end-user operations, and significantly improve work efficiency.

[0082] 2. Data encryption and decryption technology

[0083] In view of the confidentiality characteristics of surveying and mapping data, the present invention also designs a "one package one key" random transformation matrix encryption technology to solve the bottleneck of surveying and mapping data services that cannot be "crowd-sourced" due to the current constraints of geographic information security and confidentiality. The core principle of this technology is that when each task package is distributed, the data encryption and security management module dynamically generates a unique matching random transformation matrix to encrypt the geographic location information in the task package. The random transformation matrix is ​​constructed by random parameters and transformation models. The transformation models mainly include four-parameter models, seven-parameter models, and grid models. By configuring an independent encryption key for each task package, a "one package one key" security mechanism is formed.

[0084] like Figures 1 to 5 As shown, based on the crowdsourcing mapping system of the present invention, random transformation matrix technology is used in the data encryption and security management module in the cloud to perform encryption and decryption. Before the cloud distributes the task package to the edge end, an independently generated random transformation matrix is ​​used for encryption; when the edge end returns the task package to the cloud, a random transformation matrix uniquely matched with it is used for decryption, thereby realizing the encryption and decryption process of the task data; wherein the random transformation matrix is ​​uniformly stored in the key management database of the cloud (such as Figure 5 ).

[0085] When the data encryption technology is applied to the crowdsourcing surveying and mapping method of the present invention, the following method is adopted:

[0086] ① Generation of random transformation matrix: According to the data type and technical requirements of the task package, select an appropriate transformation model and set appropriate random parameters to meet the corresponding encryption strength without affecting the human readability of the data.

[0087] ② In S10, before the task package that has been processed by dimension reduction is distributed to the corresponding edge end, the geographic location information is linearly or nonlinearly transformed using a random transformation matrix so that the original geographic coordinates are randomly distorted, so that even if the data is obtained during the transmission process, it cannot be easily decrypted.

[0088] The encryption process can be expressed by the following formula: P'=M*P; where P is the coordinate vector of the original geographic location information (such as (x, y) coordinates); M is the random transformation matrix; and P' is the encrypted new coordinate vector (such as (x', y') coordinates).

[0089] Taking the four-parameter model as an example, the specific form of the random transformation matrix M can be expressed as:

[0090]

[0091] Where s is the random scaling factor; θ is the random rotation angle; ΔX is the random translation along the x-axis; and ΔY is the random translation along the y-axis.

[0092] ③ In the above ①, after each task package independently generates the corresponding random transformation matrix, it is uniformly stored in the key management library in the cloud. The key management database adopts hierarchical permission management and identity authentication mechanism to ensure that only authorized users can access specific transformation matrices for decryption operations of task packages, thereby ensuring the security of random transformation matrices.

[0093] ④ In S20, when the task package is executed and transmitted back to the cloud, a decryption operation is performed through a random transformation matrix that uniquely matches the task package to restore the original geographic location information.

[0094] In the decryption process, the decryption process is the inverse operation of encryption, which can be expressed as: P = M -1 *P'; where M -1 It is the inverse matrix of the transformation matrix M, which is used to restore the encrypted coordinate vector P' to the original coordinate P.

[0095] 3. Dynamic Intelligent Scheduling Technology

[0096] like Figure 1 and Figure 2As shown, in S10, the cloud distributes the task packages after dimension reduction processing to the matching edge terminals through the intelligent scheduling module; in S20, the edge terminals further distribute the task packages to the matching terminal users through the task allocation and feedback module; in S40, the edge terminals dynamically schedule the matching terminals through the upper and lower feedback collaboration modules, and at the same time, the cloud terminals dynamically schedule other matching edge terminals according to actual needs through the upper and lower feedback collaboration modules.

[0097] In the above three steps, dynamic scheduling technology is applied. The present invention establishes a multi-layer intelligent scheduling mechanism, and intelligently allocates the task package to the edge end (also called edge team) with corresponding fields or matching capabilities according to the type and work content of the task package. The edge end further allocates the task package to the appropriate terminal user for processing according to the specific requirements of the task package and the processing capabilities of the subordinate terminal users, monitors the task execution status in real time at each level and provides timely feedback to dynamically adjust the task allocation.

[0098] However, on the basis of this dynamic intelligent scheduling, the present invention specifically designs an automatic destruction and reconstruction mechanism for task packages in response to feedback information such as edge task timeout, or failure of topology verification of returned task packages, as well as temporary content adjustments of projects, so as to achieve rapid redistribution of tasks while ensuring task continuity and data security, and ensure effective scheduling of resources and efficient completion of tasks.

[0099] The automatic destruction and reconstruction mechanism of task packages adopts the following method: Under this mechanism, in S10, the real-time monitoring module in the cloud establishes a dynamic task package lifecycle management model and records whether there are any task timeouts, failures in task package topology verification during transmission, or temporary adjustments to project content during the entire process of task package creation, allocation, execution, and transmission. If so, the task is destroyed and rebuilt, and the rebuilt task package is reallocated to the matching edge using the intelligent scheduling module in the cloud. As a result, this mechanism can effectively improve the flexibility and efficiency of task processing, and also provides fault tolerance and exception handling strategies for the "delay processing and dynamic task regeneration mechanism" mentioned above. For topology verification failures caused by unreasonable editing of the outer expansion ring elements, dynamic destruction and reconstruction can be performed to further ensure the overall topological consistency and integrity of the task data.

[0100] In summary, the crowdsourcing surveying and mapping method based on the cloud edge of the present invention uses AI analysis (i.e. artificial intelligence analysis), spatial analysis, and outer ring topology fidelity editing and other technologies according to the specific content, technical standards, and data requirements of the surveying and mapping project to analyze and extract key processes and key elements of the original task, build a decomposition model, and perform dimensionality reduction processing on geographic information data, reducing the dimension of data tasks with large data volume and high professional threshold into lightweight task packages that are easy to use and easy to handle. At the same time, the "one package, one key" random transformation matrix encryption technology is used to encrypt the task package and decrypt it after transmission, and the returned task package is topologically verified based on preset rules to complete data integration and overall quality control. The edge receives the task package and performs specific distribution and real-time monitoring. At the same time, a quality control model is built to implement quality control. The end user loads the customized editing tool designed in the cloud to collect, modify, and inspect the task package in a standardized, semi-automatic, and easy-to-use manner. During the implementation process, the cloud, edge and terminal are monitored in real time at each level, intelligent scheduling is implemented according to the task execution status, terminal computing resources are dynamically updated based on the end-user's operating efficiency and accuracy, and task allocation is optimized and adjusted in real time to ensure the efficiency and quality of task execution, and finally complete the project tasks efficiently, with high quality and low cost.

[0101] The implementation methods of the present invention are not limited to this. According to the above content of the present invention, using common technical knowledge and customary means in the field, without departing from the above basic technical ideas of the present invention, the present invention can also make other various forms of modification, replacement or combination, all of which fall within the scope of protection of the present invention.

Claims

1. A crowdsourcing mapping system based on cloud-edge, characterized in that: Including cloud, edge, and end users; The cloud includes a geographic information data dimension reduction processing module, a data encryption and security management module, an intelligent scheduling module, a real-time monitoring module, and a data integration and quality control module; The geographic information data dimensionality reduction processing module is responsible for automatically analyzing and extracting key processes and key elements according to the original task content, technical standards, data requirements and application scenarios, using artificial intelligence, topological analysis, and multi-scale analysis technology, simplifying data structure and attribute information, optimizing spatial and temporal data expression, and gradually simplifying large-volume, high-dimensional, and multi-attribute geographic information data tasks into lightweight, low-dimensional, and easy-to-process task packages; The data encryption and security management module is responsible for data encryption and decryption, authority management and access control of the task package; The intelligent scheduling module is responsible for distributing and intelligently scheduling the task packages that have undergone dimensionality reduction processing and data encryption according to the resource conditions of the edge end; The real-time monitoring module is responsible for monitoring the task execution process and intervening; The data integration and quality control module is responsible for decrypting and integrating the data of each returned task package, and performing overall quality control on the integrated data product to ensure that the complete geographic information data product meets the predetermined quality standards; The edge end includes a task allocation and feedback module, a dynamic scheduling management module, a quality and safety management module, and an up-down feedback coordination module; the edge end is responsible for receiving task packages distributed from the cloud, and performing specific task allocation, dynamic scheduling, and quality management of the task packages; The terminal user includes a data download and acquisition module, a tool embedding and optimization module, a quality control and data return module, and a real-time feedback and support module; the terminal user is responsible for the specific execution of the task package.

2. The cloud-edge-end based crowdsourcing mapping system according to claim 1, characterized in that: The task allocation and feedback module receives the complete task package from the cloud, performs task analysis, understands the specific requirements, priorities and resource requirements of the task package, automatically calculates the matching degree of subordinate terminal users by establishing an adaptive resource allocation model based on terminal user capabilities, historical execution data, and task requirement indicators, and further allocates the task package to appropriate terminal users for effective execution; and also transmits the terminal feedback data that has passed quality control back to the cloud; The dynamic scheduling management module monitors and records the execution progress of the terminal user in real time, issues timely warnings for any progress risks found, calculates and evaluates the matching terminal resources in real time, and dynamically schedules the available matching resources as needed; The quality and safety management module builds a standardized quality control model based on the technical quality requirements of the task package sent by the cloud, implements strict and unified quality management on subordinate terminal users, and ensures that the data returned by the terminal meets the technical quality standards; The up-down feedback collaboration module is used for the edge end to regularly report task execution data and available resource data to the cloud, so that the cloud can dynamically schedule other matching edge teams according to actual needs; the edge end collects the execution status and feedback information of terminal users in real time, so that the edge end can dynamically schedule matching resources and continuously iterate and optimize the adaptive resource allocation model.

3. The cloud-edge-end based crowdsourcing mapping system according to claim 1, characterized in that: The data download and collection module is used to receive the specific task package and task analysis assigned by the edge terminal, and collect geographic information data according to relevant requirements; The tool embedding and optimization module is used to provide an embeddable, installation-free, and easy-to-operate acquisition tool; The quality control and data return module performs automatic quality inspection on the collected data through the quality control model provided by the edge end, and automatically returns the passed data to the edge end according to the specified rules; The real-time feedback and support module provides real-time feedback and technical support to help end users solve problems encountered during the production process.

4. A cloud-edge-based crowdsourcing mapping method, using the cloud-edge-based crowdsourcing mapping system as described in any one of claims 1 to 3, characterized in that: The following steps are involved: S10: In the cloud, the geographic information data dimensionality reduction processing module analyzes the original task and extracts key processes and key elements, builds a decomposition model, and performs dimensionality reduction processing on the geographic information data, thereby reducing the dimensionality of data tasks with large data volumes and high professional thresholds into lightweight task packages that are easy to use and easy to handle; the data encryption and security management module encrypts the task packages using random transformation matrix encryption technology; the intelligent scheduling module distributes each task package after dimensionality reduction processing and data encryption to the edge end with matching processing capabilities through resource evaluation algorithms and intelligent scheduling algorithms; S20: After receiving the task package, the task allocation and return module at the edge end performs task analysis on the received task package, establishes an adaptive resource allocation module, automatically calculates the matching degree of the corresponding terminal user, and then allocates the task package to the terminal user with the appropriate matching degree for effective execution; and returns the terminal data that has been executed and passed the quality control to the cloud; S30: The data download and collection module of the terminal user receives the specific task package and task analysis assigned by the edge end, and then collects geographic information data according to relevant requirements, selects or optimizes the corresponding tools through the tool embedding and optimization module, and uses the relevant tools to execute the corresponding task package. The quality control and data return module performs automatic quality inspection on the collected information data according to the quality control model provided by the edge end, and returns the data that passes the quality inspection to the edge end; at the same time, the real-time feedback and support module is used to feedback problems to the edge end in real time and obtain technical support; S40: The up-down feedback coordination module at the edge collects the execution status and feedback information of the terminal users in real time, so that the edge can dynamically schedule matching resources and continuously iterate and optimize the adaptive resource allocation model. At the same time, it regularly reports the task execution data and available resource data to the cloud, so that the cloud can dynamically schedule other matching edge terminals according to actual needs. S50: The real-time monitoring module in the cloud monitors the execution process of the task package in real time and intervenes according to the data reported by the upper and lower feedback collaboration modules; at the same time, the data integration and quality control module integrates the data sent back by the task package and performs overall quality control on the integrated data to ensure that the geographic information data products meet the predetermined quality standards.

5. The cloud-edge-based crowdsourcing mapping method according to claim 4, characterized in that: In the S10, when reducing the dimension of the original task into a lightweight task package that is easy to use and process in the geographic information data dimension reduction processing module, the following method is used: The geographic information data dimensionality reduction processing module automatically analyzes and extracts key processes and key elements through AI analysis, aggregation analysis or multi-scale analysis technology, combined with the technical process of the original task. At the same time, it flexibly combines classification conversion technology and time series clustering technology to design standardized and refined decomposition models, simplify data structure and attribute information, optimize spatial and temporal data expression, and gradually simplify large-volume, high-dimensional, multi-attribute geographic information data tasks into lightweight, low-dimensional, and easy-to-process task packages; The decomposition model mainly includes data dimension reduction, spatial dimension reduction, semantic dimension reduction and temporal dimension reduction; In addition, the geographic information data dimensionality reduction processing module also designs an outer ring topology fidelity editing technology to ensure the consistency of the topological structure of each task package throughout the entire execution process.

6. The cloud-edge-based crowdsourcing mapping method according to claim 5, characterized in that: The dimensionality reduction method used by the decomposition model adopts the following method: 1) Data reduction: Use AI algorithms to analyze the data of key processes and key elements, extract important geographic information elements that require manual processing, and distribute the streamlined data to end users for processing, reducing the data load and redundancy of end users; 2) Spatial dimensionality reduction: Based on the specific process of the production task of the original task data, combined with multi-scale analysis and aggregation technology, the spatial dimension of the original task is gradually simplified to reduce the technical threshold and difficulty of the end user; 3) Semantic dimensionality reduction: The complex attribute codes and related information of the original task are simplified and classified through classification conversion technology, converted into easy-to-understand text descriptions, and classified according to the text descriptions and provided to the edge end that is familiar with the relevant attributes; 4) Time dimension reduction: Through time series clustering technology, the time series data of the original task is simplified, key time point data is extracted, and then distributed to appropriate edge terminals and end users.

7. The cloud-edge-based crowdsourcing mapping method according to claim 5, characterized in that: The geographic information data dimension reduction processing module realizes the crowdsourcing mapping method of maintaining the topological structure of the task package through the outer expansion ring topology fidelity editing technology, using the following methods: The boundary of the feature to be edited is expanded through the expansion ring topology fidelity editing technology. The area of ​​the expanded boundary is called the expansion ring. When the elements in the outer expansion ring of the task package are identified by the end user as needing to be edited, the outer expansion ring will be marked, and the editing operation will not be performed in the current task package temporarily. After the task package is completed and transmitted back to the cloud, and passes the topological verification of data integration and quality control, the system automatically performs dynamic task regeneration for the outer expansion ring elements with the mark, and generates a new task package to handle the editing task of the outer expansion elements in the marked outer expansion ring.

8. The cloud-edge-based crowdsourcing mapping method according to claim 4, characterized in that: In S10 and S40, when the task package is transmitted from the cloud to the edge and when the task package is transmitted from the edge to the cloud, the task package needs to be encrypted and decrypted. The encryption and decryption adopt the following method: ①Generation of random transformation matrix: According to the data type and technical requirements of the task package, select the appropriate transformation model, set independent random parameters for each task package, and obtain the unique matching random transformation matrix; ② In S10, before distributing the task package that has undergone dimension reduction processing to the matching edge end, the random transformation matrix obtained in ① is used to perform linear or nonlinear transformation on the geographic location information, so that the original geographic coordinates are randomly offset and distorted; ③ In the above ①, after the random transformation matrix corresponding to each task package is independently generated, it is uniformly stored in the key management database in the cloud; ④ In S30 and S20, the terminal user's quality control and data feedback module transmits the terminal data that has passed the automated quality inspection back to the edge end. When the task allocation and feedback module at the edge end transmits the task package that has passed the quality control back to the cloud, it performs a decryption operation through a random transformation matrix that uniquely matches the task package to restore the original geographic location information.

9. The cloud-edge-based crowdsourcing mapping method according to claim 5, characterized in that: In S10, the real-time monitoring module in the cloud establishes a dynamic task package lifecycle management model and records the entire process of task package creation, allocation, execution, and return, to see whether there is a task timeout, failure of task package topology verification during return, or temporary adjustment of project content. If so, the task is destroyed and rebuilt, and the intelligent scheduling module is used in the cloud to reallocate the rebuilt task package to a matching edge end, and the edge end has the corresponding task package execution capability.

Images