AI-powered IoT systems and methods based on human-like vision and sensor data governance

Abstract

This invention discloses an AI-powered IoT system and method based on human-like vision and sensor data governance. The system includes a human-like vision module, a shared exchange application module, a shared exchange management and control module, and a shared exchange service module. The human-like vision module includes an image sensing module and an edge computing module. The image sensing module is connected to the edge computing module. The edge computing module extracts static target information and dynamic object motion information based on environmental images through brain-like dual-path computation, and fuses the static target information and dynamic object motion information to form the final output as sensor data. On the one hand, this invention obtains more accurate image features as shared sensor data through the fusion processing of static and dynamic features. On the other hand, users can receive, parse, and store data for new brands and models of devices by performing configuration operations on the platform without coding.

Description

Technical Field

[0001] This invention belongs to the field of intelligent Internet of Things (IoT) technology, and specifically relates to an artificial intelligence IoT system and method based on human-like vision and sensor data governance. Background Technology

[0002] For manufacturing enterprises and industrial parks under construction, there will typically be a greater need for device connectivity in the future. For security management platforms in these enterprises and industrial parks, the ability to connect to future, uncertain IoT device protocols is required. Different brands, models, and manufacturers of devices may have different data protocols. Currently, the traditional approach is that whenever a new device needs to be connected, backend engineers need to write new device data protocol parsing code to update the backend program. This makes users heavily reliant on software developers, and frequent program updates mean more testing work, which is detrimental to system stability and makes later operation and maintenance very inconvenient. Furthermore, new code development usually incurs new costs for users.

[0003] Safety management platforms for manufacturing enterprises and industrial parks typically need to access data from the enterprise's DCS, AMS, SIS&GDS, loading systems, and monitoring systems to monitor, analyze, display, and issue alarms on the operational status of various equipment in real time. Furthermore, to facilitate inter-business collaboration, they usually need to interface with other business systems via the internet. In most cases, the third-party internet platforms (other business systems) are developed by different vendors with different interface protocols. For each third-party platform integrated, the user needs to invest in having the software vendor develop a new interface for data integration. This integration with third-party internet platforms (and subsequent maintenance) is very inconvenient.

[0004] As can be seen from the above, the security management platform for manufacturing enterprises and industrial parks needs to solve the problem of convenient access to data protocols of various IoT devices, and the problem of convenient conversion and access to data interface protocols of various Internet platforms (other business systems). Summary of the Invention

[0005] The purpose of this invention is to provide an AI-powered Internet of Things system and method based on human-like vision and sensor data governance, in order to solve the technical problems raised in the background art.

[0006] To achieve the above objectives, this invention provides an AI-powered Internet of Things system and method based on human-like vision and sensor data governance, the technical solution of which is as follows:

[0007] The first aspect of this application provides an AI-powered Internet of Things (IoT) system based on human-like vision and sensor data governance, comprising:

[0008] A human-like vision module includes an image sensing module and an edge computing module. The image sensing module is connected to the edge computing module to feed the environmental image captured by the image sensing module to the edge computing module. The edge computing module is used to extract static target information and dynamic object motion information based on the environmental image through brain-like dual-path computing, and fuse the static target information and dynamic object motion information to form the final output result as sensing data.

[0009] The shared exchange application module serves as the entry point for various applications to access sensor data, thereby enabling the openness and sharing of sensor data.

[0010] A shared switching management module, which is used to manage and control the accessed applications;

[0011] The shared exchange service module is used to manage sensor data.

[0012] Furthermore, the shared exchange service module includes a resource catalog management module, an information resource management module, an ETL tool, a multi-source heterogeneous information consistency access AIoT fusion module, an information exchange module, and a service bus module.

[0013] Furthermore, the resource catalog management module is used to realize data aggregation and hierarchical centralized management, establish a shared data library and resource catalog, and provide data services for upper-level business.

[0014] Furthermore, the information resource management module is used to realize resource sorting, resource management, resource publishing, and resource review.

[0015] Furthermore, the ETL tool is used for data processing, data cleaning, quality inspection, and data transmission.

[0016] Furthermore, the multi-source heterogeneous information consistency access AIoT fusion module includes:

[0017] The adaptive computing power allocation subsystem is configured to implement:

[0018] 1) Logical workspace allocation: Users select the available computing power components online and submit a computing power request based on their computing power component needs;

[0019] 2) Physical workspace allocation: Components built on the cloud platform are differentiated through virtual data centers, and the same user can distinguish between different components by publishing different elastic cloud servers; components built on the data middle platform are developed by users using different workspaces for different components.

[0020] The full-dimensional analysis subsystem is configured to collect structured and unstructured data through full-domain perception, decouple the data, and decompose it into processes and scenarios so that the data runs in a unified network and resource management system.

[0021] The IoT device protocol intelligent configurable subsystem is configured to use MQTT as a lightweight IoT transmission protocol, providing message acknowledgment and selective message retransmission mechanisms. This subsystem includes a user data format definer, a code generator, and a message subscription and data converter. It supports MQTT message parsing. The user data format definer includes definitions for four message formats: hexadecimal / binary, string, JSON, and XML. The code generator supports generating parsing code for the message formats defined in the user data format definer. The message subscription and data converter selects topics to subscribe to. After a client successfully subscribes to a topic, the server sends messages to the subscriber. The message subscription and data converter's processing steps include Kafka configuration, Kafka server address setting, consumer group creation, setting data key and value serialization processing classes, consumer instance creation, listening queue consumption, sensor data format conversion, and saving data from different types of sensors to corresponding databases.

[0022] The Internet platform business interface data intelligence configurable subsystem includes a JSON structure editor and a code generator. The JSON structure editor is configured to read the corresponding values ​​based on the names in the user-defined JSON, and supports the definition of names for unordered key-value pairs and ordered arrays. The code generator is configured to automatically generate corresponding JAVA code based on the custom results of the JSON structure editor to read the JSON data, obtain the corresponding values, and store them in the database.

[0023] A full-link data governance subsystem is configured to implement the following processing flow:

[0024] 1) Data access to form the raw library: Based on data access and aggregation, relevant raw data of interest are acquired as needed according to the project's data requirements. After data access, the raw library of the project is formed. The data sources of the raw library include meteorological data, geological disaster data, and internal data of the Emergency Management Bureau. The data types include structured data, semi-structured data, and unstructured data.

[0025] 2) Data processing to form a resource library: Based on the original library, a resource library is formed through data extraction, data cleaning, and data standardization;

[0026] 3) Data processing forms a subject library. The subject library is constructed by extracting elements scattered in various business data tables of the resource library, based on emergency object elements, element characteristics, etc., and is used to deeply associate and integrate the data in the storage layer of the resource library.

[0027] 4) Data processing forms a thematic library. The thematic library is designed to meet the routine and emergency management business needs. It reorganizes the data by extracting and loading data from the resource library and thematic library, and then re-integrates the data according to the application needs of different event topics to form the thematic library.

[0028] 5) Data processing forms a business database, which serves as the data source for comprehensive emergency management applications.

[0029] Furthermore, the information exchange module includes:

[0030] The exchange monitoring subsystem is configured to monitor the status, services, and logs of data sharing and exchange. It provides functions such as registration and authorization, status query, information statistics, updates, and remote deployment for the front-end exchange system. It coordinates the operation of the departmental exchange front-end and the central exchange front-end and manages and monitors the operation of the exchange platform.

[0031] The front-end switching subsystem is configured to provide rules and patterns for front-end data exchange, utilize middleware for data exchange and sharing, realize data format conversion and bridging, and transmit information data.

[0032] The bridging and switching subsystem is configured to provide data exchange under physically isolated conditions, supporting cross-domain data exchange between two isolated network segments for various field types (such as, etc.) of data sources from heterogeneous systems. The field types include at least one of the following database table types: character, date, numeric, Blob, Clob, and stream.

[0033] The exchange and transmission subsystem is configured to provide full lifecycle file transfer and processing services. It provides exchange services to perform format conversion, filtering, and mapping of files such as database tables, business systems, XML, and files; it provides transmission services to ensure reliable file transfer, transfer of changed files, compression, and encrypted transmission; it provides exchange services and shell invocation services to integrate files with databases, business systems, etc.; it provides FTP services to integrate with external systems; and it provides file classification and retrieval services to enable file classification and retrieval.

[0034] A second aspect of this application provides an AI-powered Internet of Things (IoT) method based on human-like vision and sensor data governance, the method comprising:

[0035] The system acquires environmental images and, based on these images, extracts static target information and dynamic object motion information through brain-like dual-pathway computation. The static target information and dynamic object motion information are then fused to form the final output as sensing data.

[0036] Connect to various applications to enable the openness and sharing of sensor data;

[0037] During the data sharing process, the shared data is managed through the shared exchange business module.

[0038] Furthermore, the shared exchange service module includes a resource catalog management module, an information resource management module, an ETL tool, a multi-source heterogeneous information consistency access AIoT fusion module, an information exchange module, and a service bus module.

[0039] Furthermore, the information exchange module includes:

[0040] The exchange monitoring subsystem is configured to monitor the status, services, and logs of data sharing and exchange. It provides functions such as registration and authorization, status query, information statistics, updates, and remote deployment for the front-end exchange system. It coordinates the operation of the departmental exchange front-end and the central exchange front-end and manages and monitors the operation of the exchange platform.

[0041] The front-end switching subsystem is configured to provide rules and patterns for front-end data exchange, utilize middleware for data exchange and sharing, realize data format conversion and bridging, and transmit information data.

[0042] The bridging and switching subsystem is configured to provide data exchange under physically isolated conditions, supporting cross-domain data exchange between two isolated network segments for various field types (such as, etc.) of data sources from heterogeneous systems. The field types include at least one of the following database table types: character, date, numeric, Blob, Clob, and stream.

[0043] The exchange and transmission subsystem is configured to provide full lifecycle file transfer and processing services. It provides exchange services to perform format conversion, filtering, and mapping of files such as database tables, business systems, XML, and files; it provides transmission services to ensure reliable file transfer, transfer of changed files, compression, and encrypted transmission; it provides exchange services and shell invocation services to integrate files with databases, business systems, etc.; it provides FTP services to integrate with external systems; and it provides file classification and retrieval services to enable file classification and retrieval.

[0044] The beneficial effects of this invention are:

[0045] 1. This invention, through the design of a human-like vision module, adopts a dual-channel computing method to accurately process the environmental images collected by the sensing module. By fusing static and dynamic features, more accurate image features are obtained as shared sensing data.

[0046] 2. This invention provides a functional platform that enables users to configure / apply for IoT access protocols during the later stages of platform operation and maintenance.

[0047] 2. Based on the IoT device data protocol, this invention allows users to configure / apply for the protocol on the platform, thereby enabling access to IoT device data without requiring programmers to develop backend access programs.

[0048] 3. Users can perform configuration operations on the platform without coding, and can receive, parse, and store data for new brands and models of devices.

[0049] 4. A functional platform that allows users to configure / apply for access to data from third-party internet platforms. Based on the rules of the data interface of the third-party internet platform, users can configure the rules on the platform to achieve access to data from the third-party internet platform without developing a backend access program. Attached Figure Description

[0050] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0051] Figure 1 A structural diagram of an AI-powered Internet of Things system based on human-like vision and sensor data governance, according to an embodiment of the present invention, is shown.

[0052] Figure 2 A data processing flowchart of an edge computing module according to an embodiment of the present invention is shown.

[0053] Figure 3 A schematic diagram of the composition structure of an AI middleware platform based on AIoT fusion according to an embodiment of the present invention is shown.

[0054] Figure 4 A schematic diagram of the composition structure of a multi-source heterogeneous information consistency access AIoT fusion module according to an embodiment of the present invention is shown.

[0055] Figure 5 A flowchart illustrating the message subscription and data converter processing of the IoT device protocol intelligent configurable subsystem according to an embodiment of the present invention is shown.

[0056] Figure 6 A flowchart of the end-to-end data governance subsystem processing according to an embodiment of the present invention is shown. Detailed Implementation

[0057] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.

[0058] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples.

[0059] This invention provides an AI-powered Internet of Things (IoT) system based on human-like vision and sensor data governance, such as... Figure 1 As shown, the system 100 includes:

[0060] The humanoid vision module 101 includes an image sensing module 1011 and an edge computing module 1012. The image sensing module 1011 is connected to the edge computing module 1012 to feed the environmental image captured by the image sensing module 1011 to the edge computing module 1012. The edge computing module 1012 is used to extract static target information and dynamic object motion information based on the environmental image through brain-like dual-path computing, and fuse the static target information and dynamic object motion information to form the final output result as sensing data.

[0061] The shared exchange application module 102 serves as the entry point for various applications to access sensor data, thereby enabling the openness and sharing of sensor data.

[0062] A shared exchange management and control module 103 is used to manage and control the accessed applications;

[0063] The shared exchange service module 104 is used to manage sensor data.

[0064] In this embodiment, the image sensing module 1011 in the humanoid vision module 101 can be implemented as a camera, and the edge computing module 1012 can be implemented as a microchip to process the corresponding images captured by the camera. The processing flow can be referred to Figure 2 As shown.

[0065] Specifically, the brain-like dual-channel system mainly includes the dorsal and ventral pathways. These pathways are used to process motion and detail respectively, obtaining corresponding static target information and dynamic object motion information. The processing methods for "motion" and "detail" differ: "motion" detection uses "edges" for processing, while "detail" recognition uses "shape" and "color." The ventral pathway ("detail") is deeper than the dorsal pathway ("motion"). Motion detection: <200ms; detail recognition: 400~500ms.

[0066] Specifically, in the back-side path processing flow, the intensity of a single pixel x is considered to establish the event generation model as follows:

[0067] ,

[0068] Therefore, each event can be represented as:

[0069] ,

[0070] Where pixel coordinates x = (x, y), t is the timestamp (with us resolution); the indicators of brightness change are: brightness increase ("ON") event and brightness decrease ("OFF") event, with polarity values ​​represented as follows:

[0071] ,

[0072] For example, taking a certain image as an example, the image processed by the back-side path has the following advantages: redundancy removal, sparse output, low bandwidth, low storage cost, power saving due to sparsity, high speed, low latency (asynchronous), high temporal resolution, almost no motion blur, and high dynamic range (HDR).

[0073] In this embodiment, the core of image data processing lies in the extraction of motion features. Static features can be extracted using existing methods. The extracted motion features and static features are fused in the time dimension to obtain sensor data with high resolution and low storage space, which is beneficial for subsequent data sharing applications.

[0074] Based on the acquired sensor data, this embodiment aims to address the challenges faced by security management platforms in manufacturing enterprises and industrial parks during operation, including the access issues of IoT devices from various brands, models, and manufacturers, as well as the access issues of business data from third-party internet platforms. These challenges are solved by designing a novel microservice flexible architecture and an AI platform based on AIoT convergence.

[0075] Specifically, the sharing and interaction of sensor data processed by the humanoid vision module is achieved by introducing a shared exchange application module, a shared exchange control module, and a shared exchange business module. These modules are collectively integrated into an AI platform based on AIoT fusion, with the following structure: Figure 3 As shown. The humanoid vision module, as a data acquisition and processing device, communicates with the AI ​​platform. The communication connection can be wired or wireless. The humanoid vision module is deployed in the relevant environmental area where the data needs to be collected (usually outdoors), while the AI ​​platform is generally deployed indoors.

[0076] Shared Exchange Application Module: A unified entry point for various applications, including data openness and data sharing.

[0077] Shared exchange management module: Provides unified management of the platform, including modules such as exchange management, monitoring and statistics, directory management, and service statistics.

[0078] Shared exchange service module: Provides resource catalog management, information resource management, ETL tools, AIoT fusion module for consistent access of multi-source heterogeneous information, information exchange module and service bus, etc.

[0079] The resource catalog management in the shared exchange business module is a supporting mechanism for data sharing and exchange. It mainly realizes data aggregation and hierarchical centralized management, establishes a shared data repository and resource catalog, and provides data services for upper-level businesses.

[0080] The information resource management in the shared exchange business module includes resource sorting, resource management, resource publishing, and resource review.

[0081] The ETL tools in the shared exchange service module include data processing, data cleaning, quality inspection, and data transmission.

[0082] The information exchange module within the shared exchange service module comprises an exchange monitoring subsystem, a front-end exchange subsystem, a bridging exchange subsystem, and an exchange transmission subsystem. This information exchange module is a crucial component of the data sharing and exchange platform. It employs a message middleware software design based on JAVA technology, leveraging the data storage and processing capabilities of big data and cloud computing to support the data exchange platform in achieving massive, multi-format, highly reliable, high-concurrency, and highly stable data acquisition and distribution. Service nodes are deployed at the center of the shared exchange platform to ensure stable and reliable information transmission between the front-end exchange system and the platform. Appropriate technical means are selected to ensure reliable information transmission between the front-end exchange systems, achieving efficient transmission of exchanged information content with "no loss, no error, and no duplication." It supports routing and backup routing functions between exchange nodes, provides breakpoint resume functionality, and offers data packaging, transmission, conversion, and unpacking functions.

[0083] The information exchange module includes an exchange monitoring subsystem. This subsystem monitors the status, services, and logs of data sharing and exchange. It provides functions such as registration and authorization of the front-end exchange system, status query, information statistics, updates, and remote deployment. It coordinates the operation of departmental and central exchange front-end machines and manages and monitors the operation of the exchange platform.

[0084] The information exchange module includes a front-end exchange subsystem. This front-end exchange subsystem acts as a bridge for data exchange between business applications and the platform. It is responsible for providing rules and patterns for front-end data exchange, enabling data exchange and sharing using middleware, and facilitating data format conversion and bridging, as well as information data transmission.

[0085] The bridging and switching subsystem within the information exchange module includes the bridging service runtime environment and configuration tools. It provides data exchange in physically isolated environments. It supports cross-domain data exchange between two isolated network segments using various field types (such as character, date, numeric, Blob, Clob, and stream types) from heterogeneous system data sources (such as Oracle, MySQL, and SQL Server).

[0086] The information exchange module's exchange and transmission subsystem provides full lifecycle file transfer and processing services. It offers exchange services for format conversion, filtering, and mapping of database tables, business systems, XML, and other files. It provides transmission services for reliable file (folder) transfer, transfer of changed files, compression, and encrypted transmission. It provides exchange and shell call services for file integration with databases and business systems; FTP services for integration with external systems; and file categorization and retrieval services for categorized file access.

[0087] The structure of the multi-source heterogeneous information consistency access AIoT fusion module in the shared exchange service module is as follows: Figure 4 As shown, it consists of a computing power adaptive allocation subsystem, a full-dimensional analysis subsystem, an IoT device protocol intelligent configurable subsystem, an Internet platform business interface data intelligent configurable subsystem, a full-link data governance subsystem, a data access service, a data acquisition and application service, an integrated development designer, and a unified management platform.

[0088] The data access service provides functions such as sensing device management, sensing node authorization, business interface management, and business node authorization.

[0089] The data acquisition application service provides authentication and authorization, monitoring and management interfaces, extension interfaces, data bridging, data synchronization, data extraction, data storage, and processing workflows.

[0090] The integrated development designer includes data source access, process development, task configuration, task deployment, debugging and tracing, task execution, monitoring output, and runtime logs.

[0091] The unified management platform mainly realizes the supervision and control of the data acquisition process, including the control and management of devices, nodes and permissions, the operation and status monitoring during the acquisition process, and the performance statistics and analysis of data acquisition.

[0092] Computing power adaptive allocation subsystem. 1) Logical workspace allocation: Users can select available computing power components online (computing resources are configured on the computing power components) and submit computing power request applications according to their computing power component needs. 2) Physical workspace allocation: Components built on the cloud platform are differentiated through virtual data centers. When the same user uses different components, different ECS (Elastic Cloud Servers) are deployed to differentiate them. Components built on the data middle platform are developed using different workspaces for different users.

[0093] The full-dimensional analysis subsystem collects both structured and unstructured data (such as ledger data, form data, and text data) through comprehensive perception. It decouples the data, breaking it down into processes and scenarios, ensuring data is managed within a unified network and resource management framework, thus promoting scientific data governance. Taking video surveillance data as an example, it not only addresses the "storage, processing, and management" of video surveillance data but also maintains the "retrieval, recording, broadcasting, and control" of basic data, enabling video analysis, data clustering, and the analysis and judgment of related events.

[0094] The IoT device protocol intelligent configurable subsystem utilizes MQTT, a lightweight IoT transmission protocol, providing message acknowledgment and selective message retransmission mechanisms. It not only enables secure and reliable message delivery but also supports bidirectional data synchronization, information routing, and message addressing. This subsystem includes a user data format definer, a code generator, and a message subscription and data converter, supporting the parsing of MQTT messages. The user data format definer defines four message formats: hexadecimal / binary, string, JSON, and XML. The code generator generates parsing code for the message formats defined in the user data format definer. The message subscription and data converter selects topics to subscribe to; after a client successfully subscribes to a topic, the server sends messages to the subscriber. Figure 5 As shown, the message subscription and data converter processing steps include Kafka configuration, Kafka server address setting, creating consumer groups, setting serialization processing classes for data keys and values, creating consumer instances, listening to queues for consumption, performing format conversion on sensor data, and saving data from different types of sensors to corresponding databases.

[0095] This is a configurable subsystem for data intelligence in internet platform business interfaces. It includes a JSON structure editor and a code generator. Currently, most internet platform business interfaces use JSON as the data exchange format. JSON primarily has two data structures: unordered key-value pairs and ordered arrays. Unordered key-value pairs consist of a collection of "name / value" pairs. An object begins with "{" (left bracket) and ends with "}" (right bracket). Each "name" is followed by a ":" (colon); "name / value" pairs are separated by "," (commas). Arrays are ordered collections of values. An array begins with "[" (left square bracket) and ends with "]" (right square bracket). Values ​​are separated by "," (commas), and values ​​can be strings enclosed in double quotes, numbers, true, false, null, objects, or arrays. These value structures can be nested. The JSON structure editor allows users to define names in the JSON to read the corresponding values. The JSON structure editor supports defining names for both unordered key-value pairs and ordered arrays. The code generator automatically generates corresponding Java code based on the custom results of the JSON structure editor, which is used to read JSON data, retrieve the corresponding values, and store them in the database.

[0096] End-to-end data governance subsystem, such as Figure 6 As shown, its processing flow includes the following 5 steps:

[0097] 1) Data Access to Form the Raw Data Library. Based on data access and aggregation, relevant raw data of interest are acquired as needed according to the project's data requirements, forming the project's raw data library. The data sources for this raw data library include data from other municipal departments (e.g., meteorological data, geological disaster data, etc.) and internal data from the Emergency Management Bureau. Data types include structured data, semi-structured data, and unstructured data.

[0098] 2) Data processing to form a resource repository. Based on the original repository, a resource repository is formed through data extraction, data cleaning, and data standardization.

[0099] 3) Data processing forms a thematic library. The thematic library is constructed by extracting elements scattered in various business data tables of the resource library, based on emergency object elements, element characteristics, etc., and is mainly used for deep association and integration of data in the resource library storage layer.

[0100] 4) Data processing forms thematic databases. Thematic databases cater to both routine and emergency management needs. Data is reorganized through a secondary extraction and loading process from resource and thematic databases, and then re-integrated according to the application requirements of different event topics. Thematic databases are an extension of resource and thematic databases, storing specialized business data from different sources and on different themes, serving specific business applications in a particular field. Thematic databases can be categorized as hazardous chemical accident databases, earthquake databases, and geological disaster databases, among others.

[0101] 5) Data processing forms a business database. The business database is the data source for integrated emergency management applications (emergency command and rescue systems, emergency unified maps, etc.).

[0102] The above embodiments are only used to illustrate the present invention and are not intended to limit the present invention. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, all equivalent technical solutions also fall within the scope of the present invention, and the patent protection scope of the present invention should be defined by the claims.

Claims

1. An AI-powered Internet of Things (IoT) system based on human-like vision and sensor data governance, characterized in that: include: A human-like vision module includes an image sensing module and an edge computing module. The image sensing module is connected to the edge computing module to feed the environmental image captured by the image sensing module to the edge computing module. The edge computing module is used to extract static target information and dynamic object motion information based on the environmental image through brain-like dual-path computing, and fuse the static target information and dynamic object motion information to form the final output result as sensing data. The shared exchange application module serves as the entry point for various applications to access sensor data, thereby enabling the openness and sharing of sensor data. A shared switching management and control module, which is used to manage and control the accessed applications; A shared exchange service module, which is used to manage sensor data; The shared exchange service module includes a resource catalog management module, an information resource management module, an ETL tool, a multi-source heterogeneous information consistency access AIoT fusion module, an information exchange module, and a service bus module; The multi-source heterogeneous information consistency access AIoT fusion module includes: The adaptive computing power allocation subsystem is configured to implement: 1) Logical workspace allocation: Users select the available computing power components online and submit a computing power request based on their computing power component needs; 2) Physical workspace allocation: Components built on the cloud platform are differentiated through virtual data centers, and the same user can distinguish between different components by publishing different elastic cloud servers; components built on the data middle platform are developed using different workspaces for different users. The full-dimensional analysis subsystem is configured to collect structured and unstructured data through full-domain perception, decouple the data, and decompose it into processes and scenarios so that the data runs in a unified network and resource management system. The IoT device protocol intelligent configurable subsystem is configured to use MQTT as a lightweight IoT transmission protocol, providing message acknowledgment and selective message retransmission mechanisms. This subsystem includes a user data format definer, a code generator, and a message subscription and data converter. It supports MQTT message parsing. The user data format definer includes definitions for four message formats: hexadecimal / binary, string, JSON, and XML. The code generator supports generating parsing code for the message formats defined in the user data format definer. The message subscription and data converter selects topics to subscribe to. After a client successfully subscribes to a topic, the server sends messages to the subscriber. The message subscription and data converter's processing steps include Kafka configuration, Kafka server address setting, consumer group creation, setting data key and value serialization processing classes, consumer instance creation, listening queue consumption, sensor data format conversion, and saving data from different types of sensors to corresponding databases. The Internet platform business interface data intelligence configurable subsystem includes a JSON structure editor and a code generator. The JSON structure editor is configured to read the corresponding values ​​based on the names in the user-defined JSON, and supports the definition of names for unordered key-value pairs and ordered arrays. The code generator is configured to automatically generate corresponding JAVA code based on the custom results of the JSON structure editor to read the JSON data, obtain the corresponding values, and store them in the database.

2. The AI-powered IoT system based on human-like vision and sensor data governance according to claim 1, characterized in that, The resource catalog management module is used to realize data aggregation and hierarchical centralized management, establish a shared data library and resource catalog, and provide data services for upper-level business.

3. The AI-powered IoT system based on human-like vision and sensor data governance according to claim 1, characterized in that, The information resource management module is used to realize resource sorting, resource management, resource publishing and resource review.

4. The AI-powered IoT system based on human-like vision and sensor data governance according to claim 1, characterized in that, The ETL tool is used for data processing, data cleaning, quality inspection, and data transmission.

5. The AI-powered IoT system based on human-like vision and sensor data governance according to claim 1, characterized in that, The multi-source heterogeneous information consistency access AIoT fusion module also includes: A full-link data governance subsystem is configured to implement the following processing flow: 1) Data access to form the raw library: Based on data access and aggregation, relevant raw data of interest are acquired as needed according to the project's data requirements. After data access, the raw library of the project is formed. The data sources of the raw library include meteorological data, geological disaster data, and internal data of the Emergency Management Bureau. The data types include structured data, semi-structured data, and unstructured data. 2) Data processing to form a resource library: Based on the original library, a resource library is formed through data extraction, data cleaning, and data standardization; 3) Data processing forms a thematic library, which is constructed by extracting elements scattered in various business data tables of the resource library, based on emergency object elements and element characteristics, and is used to deeply associate and integrate the data in the storage layer of the resource library; 4) Data processing forms a thematic library. The thematic library is designed to meet the routine and emergency management business needs. It reorganizes the data by extracting and loading data from the resource library and thematic library, and then re-integrates the data according to the application needs of different event topics to form the thematic library. 5) Data processing forms a business database, which serves as the data source for comprehensive emergency management applications.

6. The AI-powered IoT system based on human-like vision and sensor data governance according to claim 1, characterized in that, The information exchange module includes: The exchange monitoring subsystem is configured to monitor the status, services, and log information of data sharing and exchange. It provides functions such as registration and authorization, status query, information statistics, updates, and remote deployment of the front-end exchange system. It coordinates the operation of the departmental exchange front-end and the central exchange front-end and manages and monitors the operation of the exchange platform. The front-end switching subsystem is configured to provide rules and patterns for front-end data exchange, utilize middleware for data exchange and sharing, realize data format conversion and bridging, and transmit information data. The bridging and switching subsystem is configured to provide data exchange under physically isolated conditions, supporting cross-domain data exchange between two isolated network segments for various field types of data sources from heterogeneous systems. The field types include at least one of the following database table types: character, date, numeric, Blob, Clob, and stream. The exchange and transmission subsystem is configured to provide full lifecycle file transfer and processing services. It provides exchange services to implement format conversion, filtering, and mapping processing of database tables, business systems, XML, and files; it provides transmission services to implement reliable file transfer, transfer of changed files, compression, and encrypted transfer; it provides exchange services and shell invocation services to achieve file-to-database and business system integration; it provides FTP services to achieve integration with external systems; and it provides file classification and retrieval services to achieve file classification and retrieval.

7. An AI-powered Internet of Things (IoT) method based on human-like vision and sensor data governance, implemented according to the system described in claim 1, characterized in that... The method includes: The system acquires environmental images and, based on these images, extracts static target information and dynamic object motion information through brain-like dual-pathway computation. The static target information and dynamic object motion information are then fused to form the final output as sensing data. Connect to various applications to enable the openness and sharing of sensor data; During the data sharing process, the shared data is managed through the shared exchange business module.

8. The method according to claim 7, characterized in that, The shared exchange service module includes a resource catalog management module, an information resource management module, an ETL tool, a multi-source heterogeneous information consistency access AIoT fusion module, an information exchange module, and a service bus module.

9. The method according to claim 7, characterized in that, The information exchange module includes: The exchange monitoring subsystem is configured to monitor the status, services, and log information of data sharing and exchange. It provides functions such as registration and authorization, status query, information statistics, updates, and remote deployment of the front-end exchange system. It coordinates the operation of the departmental exchange front-end and the central exchange front-end and manages and monitors the operation of the exchange platform. The front-end switching subsystem is configured to provide rules and patterns for front-end data exchange, utilize middleware for data exchange and sharing, realize data format conversion and bridging, and transmit information data. The bridging and switching subsystem is configured to provide data exchange under physically isolated conditions, supporting cross-domain data exchange between two isolated network segments for various field types of data sources from heterogeneous systems. The field types include at least one of the following database table types: character, date, numeric, Blob, Clob, and stream. The exchange and transmission subsystem is configured to provide full lifecycle file transfer and processing services. It provides exchange services to implement format conversion, filtering, and mapping processing of database tables, business systems, XML, and files; it provides transmission services to implement reliable file transfer, transfer of changed files, compression, and encrypted transfer; it provides exchange services and shell invocation services to achieve file-to-database and business system integration; it provides FTP services to achieve integration with external systems; and it provides file classification and retrieval services to achieve file classification and retrieval.

Images