Data updating method and device of end-side camera, electronic equipment and medium

Abstract

Embodiments of the present application disclose a data updating method and device of an end-side camera, electronic equipment and a storage medium, comprising: classifying and packing update data of the end-side camera into data update packages according to data types and chip adaptation information; a mobile terminal collects chip brands, chip models and types and version information of currently deployed data of a bound camera, sends a query request to a server, the server queries an adapted incremental data update package file in a corresponding directory structure according to the query request, and returns a query result to the mobile terminal; the mobile terminal sends a data download update instruction to the camera according to the query result, and the camera downloads the corresponding incremental data update package file from the server and completes end-side data updating according to the data download update instruction.

Description

Technical Field

[0001] This application relates to the field of data processing technology, specifically to a data update method, apparatus, electronic device, and storage medium for an end-side camera. Background Technology

[0002] With the rapid development of artificial intelligence and machine vision technologies, edge cameras, as intelligent sensing terminals, have been widely used in various fields such as intelligent security, smart communities, industrial inspection, and traffic monitoring. The core functions of edge cameras not only rely on edge AI algorithms but also require the support of various supporting operational data. This data includes, but is not limited to, target recognition feature libraries, image detection templates, algorithm operation configuration parameters, scene-based rule data, and device adaptation parameters. These data form the crucial foundation for edge AI algorithms to complete local data processing, intelligent analysis, and target detection.

[0003] The operational data of edge cameras needs to be continuously updated according to changes in actual application scenarios. For example, the facial feature database of community security cameras needs to be updated as residents change, the vehicle recognition template of traffic monitoring cameras needs to be updated with vehicle models, and the defect feature data of industrial inspection cameras needs to be updated with product process adjustments. The existing data update methods for edge cameras mainly suffer from the following technical problems: First, full updates waste resources. In existing technologies, camera-side data updates mostly use the method of downloading the entire data packet. Even if only a small amount of data needs to be updated, the entire data packet still needs to be downloaded. This not only consumes a lot of network bandwidth resources, but also consumes the camera's limited on-side storage and computing resources. For scenarios with poor network environments, such as outdoor and industrial sites, full updates are extremely inefficient and prone to update interruption.

[0004] Second, the data classification and adaptability are poor. The existing server-side storage of camera update data simply classifies it by chip brand or camera model, without combining data type and data version for fine-grained directory deployment. This results in low efficiency when mobile terminals query and adapt data, and it is easy for the downloaded data package to be incompatible with the AI ​​algorithm and hardware chip on the camera side, affecting the implementation of functions after data update.

[0005] Third, there is a lack of incremental update and data management mechanisms. Existing technologies do not implement incremental data updates based on the version and type of data already deployed by the camera, and cannot push appropriate incremental data packets based on the existing data on the device side. At the same time, there is no sound data cleanup and rollback mechanism on the camera side. Redundant and erroneous data after updates cannot be deleted in a timely manner, and data version rollback operations are complex. After long-term use, data accumulation on the device side is likely to occur, further occupying storage resources and reducing equipment operating efficiency.

[0006] Fourth, the data update and algorithm integration are poor. Currently, after data updates, the camera simply loads the data without establishing an adaptation loading and computation feedback mechanism between the data and the edge AI algorithm. This makes it impossible to monitor the computing power consumption and matching effect during data processing in real time, and also prevents the accurate feedback of data computation results to the mobile terminal. Users cannot promptly grasp the actual application effect after the data update. Poor data-algorithm compatibility can easily lead to problems such as decreased detection accuracy of the edge AI algorithm and wasted computing power.

[0007] Fifth, it cannot meet personalized data needs. Different users have significantly different requirements for operational data from edge cameras. For example, for the same model of camera, community security scenarios require facial feature database data, while traffic monitoring scenarios require vehicle recognition data. However, existing technologies pre-install fixed basic data when the camera leaves the factory, and the server does not provide diversified and scenario-based data packages for users to choose from. This makes it difficult to meet the personalized data needs of different scenarios, and the pre-installed data cannot be updated in a timely manner as the scenario changes, further limiting the functional expansion of edge cameras.

[0008] In summary, existing data update methods for edge cameras suffer from problems such as resource waste, poor adaptability, low update efficiency, lack of data management, insufficient collaboration, and inability to meet personalized needs. Summary of the Invention

[0009] This application provides a data update method, electronic device, apparatus, and storage medium for edge cameras, aiming to solve the problems of existing edge camera data update methods, such as waste of resources in full-volume updates, poor data classification adaptability, lack of incremental update mechanism, lack of data management, low synergy with algorithms after updates, and difficulty in meeting users' personalized data needs. It proposes an edge camera data update method, apparatus, system, and camera to achieve incremental, accurate, and on-demand updates of edge camera data, optimize edge resource utilization, improve data update efficiency and adaptability, and establish a sound data management and algorithm collaboration mechanism to meet the personalized data needs of different scenarios.

[0010] In a first aspect, embodiments of this application provide a data update method for an end-side camera, including: Update data from the edge camera is categorized and packaged into data update packages according to data type and chip compatibility information. After the data update packages are uploaded to the server, a directory structure is deployed on the server. The directory structure includes, in sequence, the chip brand directory, chip model directory, data type directory, and data version directory of the camera. Multiple data update package files in the data update package are placed in the corresponding data version directory. The mobile terminal collects the chip brand, chip model, and type and version information of the currently deployed data from the bound camera and sends a query request to the server. The server queries the corresponding directory structure for the appropriate incremental data update package file based on the query request and returns the query result to the mobile terminal. The mobile terminal sends a data download and update command to the camera based on the query results. The camera then downloads and updates the corresponding incremental data update package file from the server based on the data download and update command.

[0011] Optionally, in some embodiments of this application, the data update package file includes a data ontology file and a data adaptation description file. The data ontology file is used to store and represent the update data of the end-side camera, and the data adaptation description file is used to describe the adaptation information and update rules of the update data.

[0012] Optionally, in some embodiments of this application, the data adaptation description file is used to describe the chip adaptation parameters, data verification rules, incremental update range, and data activation conditions of the updated data in JSON data format.

[0013] Optionally, in some embodiments of this application, the camera downloads and completes the end-side data update of the corresponding incremental data update package file from the server according to the data download and update instruction, specifically including: After receiving the data download and update instruction, the camera downloads the corresponding incremental data update package file from the corresponding directory structure on the server. After the download is completed, the data update package file is decompressed and the integrity of the data and chip compatibility are verified according to the rules of the data adaptation description file. If the verification passes, the original data on the end side is overwritten or updated based on the incremental data, and the updated data is stored in the preset data directory.

[0014] Optionally, in some embodiments of this application, the method further includes: The mobile terminal obtains the type, version, and storage status of the currently updated data from the bound camera, and sends a data cleanup or data rollback command to the camera according to the user's needs. The camera deletes redundant data on the end side or rolls back to the original data of the specified version according to the data cleanup or data rollback command.

[0015] Optionally, in some embodiments of this application, the method further includes: After the camera completes the data update, the updated data is loaded using the edge data processing framework, and the edge AI algorithm is called to perform data matching calculations. At the same time, the data matching results and edge computing power usage data during the calculation process are sent to the mobile terminal.

[0016] Optionally, in some embodiments of this application, the edge data processing framework includes a chip NPU driver layer, a data adaptation abstraction layer, a data computation layer, and a data interaction layer. The data adaptation abstraction layer is used to automatically identify the updated data type and complete the adaptation and loading of the data with the edge AI algorithm. The data computation layer is used to perform matching and computation between the loaded updated data and the on-site data collected by the camera. The data interaction layer is used to upload the data matching results and computing power usage data to the data cloud. The data cloud sends the processed computation results to the mobile terminal.

[0017] Secondly, embodiments of this application provide a data update device for an end-side camera, comprising: The packaging module is used to classify and package the updated data of the end-side camera into data update packages according to data type and chip compatibility information. After the data update package is uploaded to the server, a directory structure is deployed on the server. The directory structure includes, in sequence, the chip brand directory, the chip model directory, the data type directory, and the data version directory of the camera. Multiple data update package files in the data update package are placed in the corresponding data version directory. The sending module is used by the mobile terminal to collect the chip brand, chip model and the type and version information of the currently deployed data of the bound camera and send a query request to the server. The server queries the corresponding directory structure for the appropriate incremental data update package file according to the query request and returns the query result to the mobile terminal. The update module is used by the mobile terminal to send a data download update command to the camera based on the query results. The camera then downloads and completes the corresponding incremental data update package file from the server based on the data download update command to update the data on the device side.

[0018] Accordingly, this application also provides an electronic device, including a memory, a processor, and a processor program stored in the memory and executable on the processor, wherein the processor executes the program as described in any of the preceding methods.

[0019] This application also provides a storage medium storing a processor program that, when executed by a processor, implements the method described in any of the preceding claims.

[0020] This application provides a data update method, apparatus, electronic device, and storage medium for edge cameras. The update data from the edge camera is categorized and packaged into data update packages according to data type and chip compatibility information. After the data update packages are uploaded to a server, a directory structure is deployed on the server. This directory structure sequentially includes a chip brand directory, a chip model directory, a data type directory, and a data version directory. Multiple data update package files within the data update package are placed in the corresponding data version directory. A mobile terminal collects the chip brand, chip model, and currently deployed data type and version information of the bound camera and sends a query request to the server. The server queries the corresponding directory structure for the appropriate incremental data update package file based on the query request and returns the query result to the mobile terminal. The mobile terminal sends a data download update command to the camera based on the query result. The camera downloads and completes the edge data update of the corresponding incremental data update package file from the server according to the data download update command. In the edge camera data update scheme provided in this application, the server pushes appropriate incremental data packets based on the type and version of the deployed data of the camera, replacing the traditional full update method. This significantly reduces network bandwidth usage, lowers the storage and computing power consumption of the camera edge, and improves data update efficiency, making it particularly suitable for scenarios with poor network environments. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a flowchart illustrating the data update method for an end-side camera provided in an embodiment of this application; Figure 2 This is a schematic diagram of the data update device for the end-side camera provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0023] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] In the description of this application, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships based on the orientation or positional relationships shown in the accompanying drawings, are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "a plurality of" means two or more, unless otherwise explicitly specified.

[0025] See Figure 1 , Figure 1 This is a flowchart illustrating the data update method for an end-side camera provided in an embodiment of this application, as detailed below: The technical solution of this application embodiment is applicable to all edge cameras with edge AI function, including but not limited to smart security cameras, smart community face capture cameras, traffic monitoring vehicle recognition cameras, industrial defect detection cameras, supermarket customer flow statistics cameras, etc. It can realize incremental, accurate and on-demand updates of various edge operation data. The edge operation data includes target recognition feature library (face, vehicle, object, defect, etc.), image detection template, algorithm operation configuration parameters, scene-based rule data, device hardware adaptation parameters, abnormal warning threshold data, etc.

[0026] The core design concept of this application embodiment is as follows: to achieve classified storage of updated data through refined directory deployment on the server side, to achieve the acquisition of adapted incremental data packets through precise querying on the mobile terminal, to achieve efficient deployment of data through incremental updates, data verification, management and maintenance on the camera side, and to achieve collaborative operation and result feedback of data and AI algorithms through the edge data processing framework, so as to solve various technical problems of existing data update methods and realize the full life cycle management and efficient utilization of edge camera data.

[0027] The data update method, apparatus, system, and camera of the end-side camera according to embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0028] The main flow of the data update method for the end-side camera provided in this application embodiment is as follows: Figure 1As shown, the core includes three main steps: server-side data packet deployment, mobile terminal data query, and camera-side incremental update. It also expands upon these steps with processes such as data cleanup and rollback, and data loading and computational feedback. The detailed implementation is as follows: S1. The update data of the end-side camera is classified and packaged into data update packages according to data type and chip compatibility information. After the data update package is uploaded to the server, a directory structure is deployed on the server. The directory structure includes, in sequence, the chip brand directory, the chip model directory, the data type directory, and the data version directory of the camera. Multiple data update package files in the data update package are placed in the corresponding data version directory.

[0029] In this embodiment, the packaging and server directory deployment of updated data are fundamental to achieving accurate data querying and incremental updates. The principles of categorized packaging and layered deployment must be followed. Specific implementation details are as follows: Update data classification and packaging First, the various update data from the edge cameras are classified according to data type. Data types can be divided according to the application scenarios and functional requirements of the edge cameras, such as facial feature database data, vehicle recognition template data, object detection feature data, industrial defect feature data, algorithm configuration parameter data, and equipment adaptation parameter data. Each data type corresponds to a unique identifier code, which facilitates identification by the server and mobile terminal.

[0030] Secondly, for each type of data, we filter it based on the camera's chip compatibility information. The chip compatibility information includes the chip brand (such as HiSilicon, Ascend, Rockchip, NVIDIA, etc.) and the chip model (such as HiSilicon Hi3516DV300, Ascend310, Rockchip RK3588, etc.) to ensure that each type of data only contains content that the corresponding chip hardware can support, thus avoiding data incompatibility with hardware.

[0031] Finally, the categorized updated data is packaged into data update packages according to data version. Under the same data type and the same chip adaptation information, different update versions correspond to different data update packages. For example, the face feature library data of the HiSilicon Hi3516DV300 chip, version V1.0 is the basic feature library, version V1.1 is the incremental update library of V1.0, and version V1.2 is the incremental update library of V1.1. They are packaged into independent data update packages.

[0032] The data update package contains two core files: a data ontology file and a data adaptation description file. These two files have a one-to-one correspondence and together constitute a complete updatable data package. The data ontology file is a binary or structured file used to store and represent the actual updated data from the edge camera, such as feature value data from a facial feature library or image feature data from a vehicle recognition template. Its file format is adapted to the reading requirements of the edge AI algorithm based on the data type. The data adaptation description file is a text file written in JSON format, used to describe in detail the adaptation information and update rules of the updated data. JSON format is characterized by its simple structure, ease of reading and writing, independence from programming languages, and cross-platform parsing capabilities, effectively improving the compatibility and maintainability of the data update package.

[0033] The core content that the data adaptation description file should include is: chip adaptation parameters (corresponding supported chip brands, chip models, chip computing power requirements, etc.), data verification rules (data hash value, checksum, data size, integrity verification method, etc.), incremental update scope (the original data version corresponding to the incremental data, the field / content range of the updated data, whether the incremental update method is overwrite or append, etc.), and data activation conditions (the triggering method for data loading, the version matching requirements with the edge AI algorithm, and the scenario conditions for data activation, etc.). It can also include auxiliary information such as data name, data type identifier, data version number, data update time, and applicable scenarios for the data.

[0034] Server-side directory structure deployment After the data update package is packaged, it is uploaded to a dedicated algorithm data server. The server deploys all data update packages in a four-level hierarchical directory structure. The four levels of directories are, in order, chip brand directory, chip model directory, data type directory, and data version directory. Each level of directory is named with a corresponding identifier, forming a unique directory path, which facilitates accurate subsequent querying and downloading.

[0035] Level 1: Chip Brand Directory, named after the chip brand, such as "HiSilicon", "Ascend", "Rockchip", "NVIDIA", etc., storing all camera update data for the corresponding brand of chip; Level 2: Chip Model Directory, which belongs to the chip brand directory and is named after the chip model, such as "Hi3516DV300", "Ascend 310", "RK3588", etc. It stores all the updated data of cameras with the corresponding chip model. The third level is the data type directory, which belongs to the chip model directory. It is named by data type identifier or name, such as "face_feature", "car_recognition", "device_config", etc., and stores all version update data of the corresponding data type. Level 4: Data Version Directory, which belongs to the Data Type Directory and is named after the data version number, such as "V1.0", "V1.1", "V1.2", etc. It stores the data update package files (data body file + data adaptation description file) for the corresponding version.

[0036] An example of a unique path formed by a fourth-level directory is: / HiSilicon / Hi3516DV300 / face_feature / V1.1 / , which contains the V1.1 version face feature library data update package file for the Hisilicon Hi3516DV300 chip camera.

[0037] The server-side establishes a unified data index for all data update packages. This index records information such as the directory path, data name, data type, chip compatibility information, data version, update time, incremental update base version, file size, and hash value for each update package. This information is used to quickly respond to query requests from mobile terminals and improve query efficiency. Simultaneously, the server provides a unified API interface that supports query requests from mobile terminals and data packet download requests from cameras. The interface uses HTTPS encrypted transmission to ensure data transmission security.

[0038] S2. The mobile terminal collects the chip brand, chip model, and currently deployed data type and version information of the bound camera and sends a query request to the server. The server queries the corresponding directory structure for the appropriate incremental data update package file based on the query request and returns the query result to the mobile terminal.

[0039] The mobile terminal serves as the interaction medium between the user, the server, and the camera, undertaking the functions of data collection, query request initiation, and query result reception. The server performs precise retrieval and incremental matching based on the query request from the mobile terminal and then returns the query results. The specific implementation details are as follows: Binding and information collection between mobile terminals and cameras Mobile terminals include, but are not limited to, smartphones, tablets, personal computers, and smart central control devices. They need to be pre-bound to the end-side camera via network. Binding methods include WIFI, Bluetooth, LAN, 4G / 5G, etc. After binding, a unique device communication connection is established. The mobile terminal can send commands to the camera and obtain the camera's hardware and data information through this connection.

[0040] The camera information collected by the mobile terminal includes two parts: basic hardware information and edge data deployment information. Basic hardware information includes the camera's chip brand and model; this is an inherent hardware parameter of the camera and can be obtained through the camera's firmware, hardware identifier, or communication protocol. Edge data deployment information includes the type and version information of all currently deployed running data on the camera. This information is maintained in real-time by the data management module on the camera's edge and can be obtained from the camera's edge via data query commands. This information includes data type identifier, data version number, data installation time, and data storage path.

[0041] The mobile terminal encapsulates all the collected information in a structured manner to form a device information data packet, providing a data foundation for subsequent query requests.

[0042] Initiation and transmission of query requests Based on the user's operation instructions, the mobile terminal sends a data query request to the server using the unified API interface provided by the server. The query request contains a data packet of device information of the bound camera, and can also contain the target data type selected by the user (such as adding a type identifier if the user only needs to update the facial feature database data), so as to achieve accurate contextual query.

[0043] The query request is sent to the server using an encrypted transmission protocol. After receiving the query request, the server first verifies the validity of the request, including verifying the binding relationship between the mobile terminal and the camera, the signature information of the request, and the integrity of the device information. If the verification fails, a query failure message is returned; if the verification passes, the data retrieval process begins.

[0044] Incremental data retrieval and query result feedback from the server The server performs a layer-by-layer search based on the device information data packet in the query request, following a four-level directory structure. Simultaneously, it combines incremental update matching rules to filter suitable incremental data update packet files. The specific steps of the search and matching are as follows: The first step is to locate the corresponding first-level and second-level directories on the server side based on the chip brand and chip model, and exclude all update data for incompatible chips. The second step is to locate the corresponding third-level data type directory based on the type of data currently deployed by the camera, and exclude all updated data that is not the target data type. The third step is to filter out data packages that are higher than the current version and are incremental update packages in the fourth-level data version directory under the corresponding data type directory, based on the version number of the currently deployed data of the camera. That is, filter out incremental update packages based on the current version and exclude full update packages and low version data packages. The fourth step involves performing a second verification on the selected incremental data packets, based on the chip adaptation parameters and data activation conditions in the data adaptation description file. This ensures that the data packets match the camera's hardware computing power and the edge AI algorithm version, and excludes data packets that do not meet the activation conditions.

[0045] After the server completes the retrieval and matching, it organizes the adapted incremental data update package information into query results. The query results are a structured list, which includes information such as the name, data type, version number, incremental update range, file size, update time, server directory path, and download link of each incremental data package. It can also include auxiliary information such as the applicable scenarios of the data package and its matching status with the AI ​​algorithm.

[0046] The server returns the query results to the mobile terminal via the API interface. After receiving the query results, the mobile terminal displays them to the user in a visual form, such as listing all updatable incremental data packets in the APP interface for the user to select.

[0047] S3. The mobile terminal sends a data download and update instruction to the camera based on the query results. The camera downloads and updates the corresponding incremental data update package file from the server based on the data download and update instruction.

[0048] The mobile terminal sends a data download and update command to the camera based on the user's selection. The camera, as the executing entity, completes the core operations of data packet download, decompression and verification, and incremental update. This step is crucial for realizing on-device data updates. The specific implementation details are shown in Figure 2, as follows: Sending and receiving data download and update commands The mobile terminal receives the user's selection operation in the visual interface (such as the user selecting to download the V1.1 version of the face feature library incremental data package). Based on the data package information in the query results, it generates a data download update instruction, which contains core information such as the server download link of the data package to be updated, data type, version number, and data verification rules.

[0049] The mobile terminal sends a data download and update command to the camera through a binding communication connection. After receiving the command, the camera first verifies the legality of the command, including verifying the signature of the command and the compatibility of the data packet to be updated with its own hardware and data. If the verification fails, the camera returns a command failure message to the mobile terminal. If the verification passes, the camera enters the data packet download process.

[0050] Downloading and decompressing incremental data update packages According to the server download link in the data download update instruction, the camera initiates a data packet download request to the server through its own network module (WIFI, 4G / 5G, LAN, etc.). The download request contains the camera's device identifier, binding information, etc. After the server verifies the information, it transmits the corresponding incremental data update package file to the camera.

[0051] The camera uses a resume-from-break method to download data packets, avoiding repeated downloads due to network interruptions and improving download efficiency. After the data packets are downloaded, the camera stores them in a temporary storage directory on the device side, and then decompresses the data packets to extract the data body file and data adaptation description file.

[0052] Data integrity and chip compatibility verification The camera performs integrity checks on the decompressed data packets according to the data verification rules in the data adaptation description file. Specific verification methods include: verifying whether the hash value / checksum of the data body file is consistent with that in the description file, verifying whether the file size matches, and verifying whether the file format is correct. If the verification fails, it means that the data packet was corrupted during transmission. The camera deletes the temporarily stored data packet, returns a download failure message to the mobile terminal, and requests a re-download. If the verification passes, it proceeds to the chip compatibility verification stage.

[0053] Chip compatibility verification refers to the camera comparing the chip compatibility parameters in the data compatibility description file with its own hardware parameters to verify whether its chip brand, chip model, and computing power meet the data packet compatibility requirements. At the same time, it verifies whether the version of its own edge AI algorithm meets the data activation conditions. If the verification fails, it means that the data packet is incompatible with the camera, and the camera returns a "data mismatch" prompt message to the mobile terminal. If the verification passes, it enters the incremental update stage.

[0054] Incremental update and storage of edge data The camera incrementally updates the existing data on the device side according to the incremental update range and update method in the data adaptation description file. The update method is divided into two types: overwrite update and append update. Append update is suitable for scenarios that require adding new content, such as feature library and template library. For example, when adding resident feature data to the existing face feature library, the camera appends the content of the incremental data body file to the original data file. Overwrite update is suitable for scenarios that require modifying the original content, such as configuration parameters and warning thresholds. For example, when modifying the running threshold parameters of the algorithm, the camera overwrites the corresponding fields of the original data with the content of the incremental data body file.

[0055] During incremental updates, the camera backs up the original data and stores the backup data in a backup directory on the device. The backup data includes the data version number and backup time, providing a basis for subsequent data rollback. After the incremental update is complete, the camera stores the updated data in a preset data directory, which is the default data reading directory for the device's AI algorithm, ensuring that the AI ​​algorithm can quickly load the updated data.

[0056] After the data update is complete, the camera updates the data deployment information in the edge data management module, including the updated data version number, update time, storage path, etc. At the same time, it sends a successful update feedback message to the mobile terminal, including the data update type, version, update method, etc. If an error occurs during the update process, the camera returns an error message to the mobile terminal and restores the backed-up original data to ensure the integrity of the edge data.

[0057] S4. Data Cleaning and Data Rollback Extension Process To address the issues of data accumulation and the inability to roll back erroneous data in existing edge-side cameras, the method in this application embodiment further includes a data cleaning and data rollback process, enabling refined management of edge-side data. Specific implementation details are as follows: Acquisition of edge data information The mobile terminal can send data query commands to the camera at any time through the bound communication connection. After receiving the command, the camera extracts information such as the type, version, storage status, installation time, and data size of the currently updated data through the end-side data management module and returns it to the mobile terminal. The mobile terminal displays the information in a visual form, allowing users to view the overall deployment status of the end-side data.

[0058] Execution of data cleanup commands If a user discovers redundant data on the device side (such as outdated data or mismatched contextual data), they can select the data to be cleaned on the mobile terminal. The mobile terminal then generates a data cleanup command and sends it to the camera. Upon receiving the command, the camera first verifies whether the data to be cleaned is currently in use. If it is, it returns a "cleanup failed" message to the mobile terminal. If it is redundant data, it deletes the data's body file and description file, updates the information in the device-side data management module, releases the corresponding storage resources, and sends feedback information to the mobile terminal after the cleanup is complete.

[0059] Execution of data rollback instructions If a user discovers errors in the updated data (such as decreased recognition accuracy due to incorrect feature library data, or abnormal device operation due to incorrect configuration parameters), they can select the data type and target version to be rolled back on the mobile terminal. The mobile terminal then generates a data rollback command and sends it to the camera. Upon receiving the command, the camera retrieves the corresponding data type and target version backup data from the backup directory, replaces the currently used data with the backup data, deletes the erroneous updated data, updates the information in the edge data management module, and sends feedback information to the mobile terminal after the rollback is complete, ensuring the camera returns to normal operating status.

[0060] S5, Data Loading, Calculation and Result Feedback Extension Process To address the issue of low synergy between updated data and edge AI algorithms, the method in this application further includes data loading, computation, and result feedback processes to construct a collaborative operation mechanism between edge data and AI algorithms, enabling real-time monitoring of update effects. Specific implementation details are as follows: Construction of the edge data processing framework A pre-built edge data processing framework is constructed on the camera side. This framework is a software-layer architecture, developed based on the camera's chip hardware and edge AI algorithms, and deeply adapted to the camera's NPU (Neural Processing Unit). The core structure comprises four layers: a chip NPU driver layer, a data adaptation abstraction layer, a data processing layer, and a data interaction layer. Each layer adopts a modular design with low coupling and high cohesion, facilitating development and maintenance. The specific functions and interaction relationships of each layer are as follows: Chip NPU Driver Layer: This is the bottom layer of the framework, which directly interacts with the camera's NPU hardware, provides NPU driver support, enables the scheduling and management of hardware resources, provides computing power support for upper-layer data processing and computation, and is responsible for converting upper-layer computation instructions into hardware instructions that can be executed by the NPU. Data Adaptation Abstraction Layer: This is the core adaptation layer of the framework, located between the NPU driver layer and the data processing layer. It is used to automatically identify updated data types and complete the adaptation and loading of data with edge AI algorithms. This layer parses the data adaptation description file to obtain information such as data type, format, and activation conditions. It automatically matches the corresponding AI algorithm on the camera edge, converts the updated data into a format readable by the AI ​​algorithm, and loads it into the algorithm's runtime memory, achieving seamless data-algorithm integration without manual intervention. Data Computation Layer: This is the core computation layer of the framework, located above the data adaptation abstraction layer. It is used to perform data matching operations between the loaded updated data and the on-site data collected by the camera. The camera's image acquisition module transmits real-time captured on-site images and video data to this layer. This layer calls the edge AI algorithm to match, detect, and analyze the on-site data with updated feature libraries, template libraries, and other data to complete intelligent recognition and judgment. At the same time, it collects the data matching results (such as recognition success rate, target matching degree, anomaly warning information, etc.) and edge computing power usage data (such as NPU computing power utilization, memory usage, computation time, etc.) in real time during the operation. Data Interaction Layer: This is the upper layer of the framework, responsible for uploading and feedback of data. It encapsulates the matching results and computing power usage data collected by the data processing layer in a structured manner, uploads them to the data cloud through an encrypted transmission protocol, and simultaneously receives the processing results and instructions from the data cloud.

[0061] Feedback and display of data processing results After receiving the matching results and computing power usage data uploaded by the camera, the data cloud performs simple processing and analysis, such as calculating the recognition success rate, average computing power usage, and number of abnormal warnings. Then, the processed calculation results are sent to the mobile terminal via the network.

[0062] The mobile terminal displays the computation results to the user in a visual format, such as real-time recognition success rate, NPU computing power utilization, and anomaly warning information displayed in the APP interface. Users can use this information to promptly understand the actual application effect after data updates. If problems such as excessively low recognition success rate, excessively high computing power utilization, or frequent anomaly warnings are detected, users can promptly initiate data cleanup, rollback, or re-update operations via the mobile terminal to ensure the normal operation of the camera.

[0063] Meanwhile, the cloud-based data storage system stores the camera's computational data long-term, generating data update effect analysis reports to provide data support for subsequent data update optimization and edge AI algorithm iteration.

[0064] This application provides a data update method for edge cameras. The updated data from the edge camera is categorized and packaged into data update packages based on data type and chip compatibility information. After the data update packages are uploaded to a server, a directory structure is deployed on the server. This directory structure sequentially includes a chip brand directory, a chip model directory, a data type directory, and a data version directory. Multiple data update package files within the data update package are placed in the corresponding data version directory. A mobile terminal collects the chip brand, chip model, and currently deployed data type and version information of the bound camera and sends a query request to the server. The server queries the corresponding directory structure for the appropriate incremental data update package file based on the query request and returns the query result to the mobile terminal. The mobile terminal sends a data download update command to the camera based on the query result. The camera downloads and updates the corresponding incremental data update package file from the server according to the data download update command. In the edge camera data update scheme provided in this application, the server pushes appropriate incremental data packets based on the type and version of the deployed data of the camera, replacing the traditional full update method. This significantly reduces network bandwidth usage, lowers the storage and computing power consumption of the camera edge, and improves data update efficiency, making it particularly suitable for scenarios with poor network environments.

[0065] To facilitate better implementation of the data update method for end-side cameras in the embodiments of this application, the embodiments of this application also provide a data update device for end-side cameras, wherein the meanings of the terms are the same as those in the data update method for end-side cameras described above, and specific implementation details can be found in the description in the system embodiments.

[0066] Please see Figure 2 , Figure 2 This is a schematic diagram of the structure of a data update device for an end-side camera provided in an embodiment of this application. Specifically, the data update device for the end-side camera may include a packetization module 201, a transmission module 202, and an update module 203, as follows: The packaging module 201 is used to classify and package the update data of the end-side camera into data update packages according to data type and chip compatibility information. After the data update package is uploaded to the server, a directory structure is deployed on the server. The directory structure includes, in sequence, the chip brand directory, the chip model directory, the data type directory, and the data version directory of the camera. Multiple data update package files in the data update package are placed in the corresponding data version directory. The sending module 202 is used for the mobile terminal to collect the chip brand, chip model and the type and version information of the currently deployed data of the bound camera and send a query request to the server. The server queries the corresponding directory structure for the appropriate incremental data update package file according to the query request and returns the query result to the mobile terminal. The update module 203 is used by the mobile terminal to send a data download update instruction to the camera based on the query results. The camera downloads and updates the corresponding incremental data update package file from the server based on the data download update instruction.

[0067] This application provides a data update device for an edge camera. The update data from the edge camera is categorized and packaged into data update packages according to data type and chip compatibility information. After the data update packages are uploaded to a server, a directory structure is deployed on the server. This directory structure sequentially includes a chip brand directory, a chip model directory, a data type directory, and a data version directory. Multiple data update package files within the data update package are placed in the corresponding data version directory. A mobile terminal collects the chip brand, chip model, and currently deployed data type and version information of the bound camera and sends a query request to the server. The server queries the corresponding directory structure for a compatible incremental data update package file based on the query request and returns the query result to the mobile terminal. The mobile terminal sends a data download update command to the camera based on the query result. The camera downloads and completes the edge data update of the corresponding incremental data update package file from the server according to the data download update command. In the edge camera data update scheme provided in this application, the server pushes compatible incremental data packets based on the type and version of the deployed data of the camera, replacing the traditional full update method. This significantly reduces network bandwidth usage, lowers the storage and computing power consumption of the camera edge, and improves data update efficiency, making it particularly suitable for scenarios with poor network environments. Furthermore, embodiments of this application also provide an electronic device, such as... Figure 3 As shown, it illustrates a structural schematic diagram of the electronic device involved in the embodiments of this application, specifically: The electronic device may include components such as a processor 301 with one or more processing cores, a memory 302 with one or more processor-readable storage media, a power supply 303, and an input unit 304. Those skilled in the art will understand that... Figure 3 The electronic device structure shown does not constitute a limitation on the electronic device and may include more or fewer components than shown, or combine certain components, or have different component arrangements. Wherein: Processor 301 is the control center of the electronic device. It connects various parts of the electronic device via various interfaces and lines. By running or executing software programs and / or modules stored in memory 302, and by calling data stored in memory 302, it performs various functions and processes data, thereby providing overall monitoring of the electronic device. Optionally, processor 301 may include one or more processing cores; preferably, processor 301 may integrate an application processor and a modem processor. The application processor mainly handles the operating system, user interface, and applications, while the modem processor mainly handles data updates from the wireless end-side camera. It is understood that the modem processor may not be integrated into processor 301.

[0068] The memory 302 can be used to store software programs and modules. The processor 301 executes various functional applications and data update methods for the end-side camera by running the software programs and modules stored in the memory 302. The memory 302 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device, etc. In addition, the memory 302 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory 302 may also include a memory controller to provide the processor 301 with access to the memory 302.

[0069] The electronic device also includes a power supply 303 that supplies power to various components. Preferably, the power supply 303 can be logically connected to the processor 301 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system. The power supply 303 may also include one or more DC or AC power supplies, recharging systems, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0070] The electronic device may also include an input unit 304, which can be used to receive input digital or character information and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function control.

[0071] Although not shown, the electronic device may also include a display unit, etc., which will not be described in detail here. Specifically, in the embodiments of this application, the processing 301 in the electronic device loads the executable files corresponding to the processes of one or more applications into the memory 302 according to the following instructions, and the processing 301 runs the applications stored in the memory 302 to realize various functions, as follows: Updated data from the edge cameras is categorized and packaged into data update packages based on data type and chip compatibility information. These data update packages are uploaded to the server, where a directory structure is deployed. This directory structure sequentially includes a chip brand directory, a chip model directory, a data type directory, and a data version directory. Multiple data update package files within each data update package are placed in their respective data version directories. The mobile terminal collects the chip brand, chip model, and currently deployed data type and version information of the bound camera and sends a query request to the server. The server queries the corresponding directory structure for compatible incremental data update package files based on the query request and returns the query results to the mobile terminal. The mobile terminal then sends a data download and update command to the camera based on the query results. The camera downloads and updates the corresponding incremental data update package file from the server according to the data download and update command, completing the edge data update.

[0072] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0073] In this embodiment, the update data of the edge camera is categorized and packaged into data update packages according to data type and chip compatibility information. After the data update packages are uploaded to the server, a directory structure is deployed on the server. The directory structure includes, in sequence, a chip brand directory, a chip model directory, a data type directory, and a data version directory for the camera. Multiple data update package files in the data update package are placed in the corresponding data version directory. The mobile terminal collects the chip brand, chip model, and currently deployed data type and version information of the bound camera and sends a query request to the server. The server queries the corresponding directory structure for the appropriate incremental data update package file according to the query request and returns the query result to the mobile terminal. The mobile terminal sends a data download update instruction to the camera according to the query result. The camera downloads and updates the edge data according to the data download update instruction from the server and completes the edge data update of the corresponding incremental data update package file. In the edge camera data update scheme provided in this application, the server pushes appropriate incremental data packets according to the type and version of the deployed data of the camera, replacing the traditional full update method, which greatly reduces network bandwidth occupation, reduces the storage and computing power consumption of the camera edge, and improves data update efficiency, especially suitable for scenarios with poor network environment.

[0074] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be performed by instructions, or by instructions controlling related hardware. These instructions can be stored in a processor-readable storage medium and loaded and executed by a processor.

[0075] Therefore, embodiments of this application provide a storage medium storing multiple instructions that can be loaded by a processor to execute steps in any of the data update methods for end-side cameras provided in embodiments of this application. For example, the instructions can execute the following steps: Updated data from the edge cameras is categorized and packaged into data update packages based on data type and chip compatibility information. These data update packages are uploaded to the server, where a directory structure is deployed. This directory structure sequentially includes a chip brand directory, a chip model directory, a data type directory, and a data version directory. Multiple data update package files within each data update package are placed in their respective data version directories. The mobile terminal collects the chip brand, chip model, and currently deployed data type and version information of the bound camera and sends a query request to the server. The server queries the corresponding directory structure for compatible incremental data update package files based on the query request and returns the query results to the mobile terminal. The mobile terminal then sends a data download and update command to the camera based on the query results. The camera downloads and updates the corresponding incremental data update package file from the server according to the data download and update command, completing the edge data update.

[0076] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0077] The storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0078] Since the instructions stored in the storage medium can execute the steps in any of the data update methods for end-side cameras provided in the embodiments of this application, the beneficial effects that any of the data update methods for end-side cameras provided in the embodiments of this application can achieve can be realized. For details, please refer to the previous embodiments, which will not be repeated here.

[0079] The data update method, apparatus, electronic device, and storage medium for an end-side camera provided in the embodiments of this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A data update method for an end-side camera, characterized in that, include: Update data from the edge camera is categorized and packaged into data update packages according to data type and chip compatibility information. After the data update packages are uploaded to the server, a directory structure is deployed on the server. The directory structure includes, in sequence, the chip brand directory, chip model directory, data type directory, and data version directory of the camera. Multiple data update package files in the data update package are placed in the corresponding data version directory. The mobile terminal collects the chip brand, chip model, and type and version information of the currently deployed data from the bound camera and sends a query request to the server. The server queries the corresponding directory structure for the appropriate incremental data update package file based on the query request and returns the query result to the mobile terminal. The mobile terminal sends a data download and update command to the camera based on the query results. The camera then downloads and updates the corresponding incremental data update package file from the server based on the data download and update command.

2. The data update method for an end-side camera according to claim 1, characterized in that, The data update package file includes a data ontology file and a data adaptation description file. The data ontology file is used to store and represent the updated data from the end-side camera, and the data adaptation description file is used to describe the adaptation information and update rules of the updated data.

3. The data update method for an end-side camera according to claim 2, characterized in that, The data adaptation description file is used to describe the chip adaptation parameters, data verification rules, incremental update range, and data activation conditions of the updated data in JSON data format.

4. The data update method for an end-side camera according to claim 1, characterized in that, The camera downloads and updates the corresponding incremental data update package file from the server according to the data download and update instructions. This includes: After receiving the data download and update instruction, the camera downloads the corresponding incremental data update package file from the corresponding directory structure on the server. After the download is completed, the data update package file is decompressed and the integrity of the data and chip compatibility are verified according to the rules of the data adaptation description file. If the verification passes, the original data on the end side is overwritten or updated based on the incremental data, and the updated data is stored in the preset data directory.

5. The data update method for an end-side camera according to claim 1, characterized in that, The method further includes: The mobile terminal obtains the type, version, and storage status of the currently updated data from the bound camera, and sends a data cleanup or data rollback command to the camera according to the user's needs. The camera deletes redundant data on the end side or rolls back to the original data of the specified version according to the data cleanup or data rollback command.

6. The data update method for an end-side camera according to claim 1, characterized in that, The method further includes: After the camera completes the data update, the updated data is loaded using the edge data processing framework, and the edge AI algorithm is called to perform data matching calculations. At the same time, the data matching results and edge computing power usage data during the calculation process are sent to the mobile terminal.

7. The data update method for an end-side camera according to claim 6, characterized in that, The edge data processing framework includes a chip NPU driver layer, a data adaptation abstraction layer, a data computation layer, and a data interaction layer. The data adaptation abstraction layer is used to automatically identify the updated data type and complete the adaptation and loading of the data with the edge AI algorithm. The data computation layer is used to match and compute the loaded updated data with the on-site data collected by the camera. The data interaction layer is used to upload the data matching results and computing power usage data to the data cloud. The data cloud sends the processed computation results to the mobile terminal.

8. A data update device for an end-side camera, characterized in that, include: The packaging module is used to classify and package the updated data of the end-side camera into data update packages according to data type and chip compatibility information. After the data update package is uploaded to the server, a directory structure is deployed on the server. The directory structure includes, in sequence, the chip brand directory, the chip model directory, the data type directory, and the data version directory of the camera. Multiple data update package files in the data update package are placed in the corresponding data version directory. The sending module is used by the mobile terminal to collect the chip brand, chip model and the type and version information of the currently deployed data of the bound camera and send a query request to the server. The server queries the corresponding directory structure for the appropriate incremental data update package file according to the query request and returns the query result to the mobile terminal. The update module is used by the mobile terminal to send a data download update command to the camera based on the query results. The camera then downloads and completes the corresponding incremental data update package file from the server based on the data download update command to update the data on the device side.

9. An electronic device, characterized in that, include: A memory, a processor, and a processor program stored in the memory and executable on the processor, wherein the processor executes the program as steps of the data update method for an end-side camera according to any one of claims 1 to 7.

10. A storage medium, characterized in that, The computer processing program is stored in which it can be loaded by a processor and executed as a data update method for an end-side camera according to any one of claims 1 to 7.

Images