Video data management method, system and storage medium for railway passenger service and production control platform

Through two-level access and three-level storage data management methods, the problems of video data access bandwidth limitation and equipment management data acquisition in railway passenger transport systems are solved, efficient access and security supervision of video data is realized, and the station is improved Safety guarantee level.

CN118014511BActive Publication Date: 2024-12-03INST OF COMPUTING TECH CHINA ACAD OF RAILWAY SCI +2
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410008362.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-01-02
Publication Date
2024-12-03
Estimated Expiration
2044-01-02

AI Technical Summary

Technical Problem

In the railway passenger transport system, there is bandwidth limitation on the access of video data between the intelligent station and the integrated video surveillance system, and the device management data cannot be obtained, resulting in naked video churn and delayed, and does not meet the railway production safety requirements.

Method used

Through two-level access and three-level storage data management methods, video data access to different systems and networks can be achieved, device management information is obtained, key scenario analysis models are optimized, scene alarm fragment files are generated, and index models are built for unified data management. .

Benefits of technology

It improves video data access efficiency and security supervision efficiency, improves station security guarantee level, solves the problems of naked video frame loss and delay, meets the needs of 100 megabit bandwidth, and realizes data sharing and unified resource management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118014511B_ABST
    Figure CN118014511B_ABST
Patent Text Reader

Abstract

The present invention provides a video data management method, system and storage medium for a railway passenger service and production control platform, the method comprising: obtaining video code streams and equipment management data of stations under the jurisdiction of the bureau and other bureaus from a comprehensive video monitoring system by a railway bureau-level platform; obtaining video code streams of the station and stations under the jurisdiction of the station from a comprehensive video monitoring system by a station-level platform; generating scene alarm fragment files based on key scenes by the station-level platform, and forming a scene alarm fragment file set, an optimized analysis model set and a relational database on the railway bureau-level platform; and indexing resources based on metadata of the railway bureau-level platform by a national railway group-level platform. The present invention adopts a two-level access and three-level storage application architecture, based on different independent systems to meet the video data access, intelligent video analysis application and data management requirements under bandwidth constraints, and realizes full-line data sharing while avoiding large data volume transmission, while improving safety supervision efficiency and service quality.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of data governance technology, and in particular to a video data management method and system for a railway passenger service and production control platform. Background Art

[0002] Railway passenger transport has special industry characteristics such as complex environment, high density of people flow, and strong mobility. Video surveillance is an important part of station construction and an important means of railway passenger transport safety. Intelligent identification and analysis of various key scenes can effectively improve safety supervision efficiency and service quality, and enhance the level of station safety assurance. In addition, the railway passenger transport section is responsible for the service of passenger trains within the jurisdiction of the bureau, and it is also necessary to timely track the operation of trains within the bureau and other bureaus according to the arrival and departure plan, and conduct video linkage to better serve passenger train crew and other work.

[0003] In the railway passenger transport system, the railway passenger service and production control platform (hereinafter referred to as the smart station) and the integrated video surveillance system are independent systems, belonging to the integrated video network and office network in the safety production network respectively. The railway integrated video surveillance system manages the information of equipment such as servers, routers, network switches and cameras in the railway bureau at the railway bureau level, and the railway bureau-level system and the station-level system with access nodes of the system can provide the video information required by various railway business departments and information systems, and realize the storage and sharing of network and video information resources. In addition, the internal network of the integrated video network where the integrated video surveillance system is located is a gigabit bandwidth, and the external interface currently adopts the national standard 28181 protocol. At present, the smart station uses the camera direct connection method to access the video code stream at the station, but this method has the situation of one machine and two networks, which does not meet the requirements of railway safety production, and cannot obtain the equipment management data of node directory, camera status and fault list data; secondly, the network connection between the national railway group-level platform, railway bureau-level platform and station-level platform of the smart station is a 100M bandwidth, and the transmission of video data in the smart station is limited by bandwidth; in addition, the accessed video code stream is a video naked stream, which has frame loss and delay.

[0004] Therefore, how to achieve the access of video data from different networks and systems, the acquisition of equipment management information, the optimization of analysis models for key scenarios, and the unified management of data are issues that need to be addressed urgently. Summary of the invention

[0005] In view of this, an embodiment of the present invention provides a video data management method and system for a railway passenger service and production control platform, so as to enable different systems and different networks to access video data, obtain equipment management information, and perform data management with the goal of data sharing, so as to ensure station operation and passenger travel safety.

[0006] One aspect of the present invention provides a video data management method for a railway passenger service and production control platform, wherein the railway passenger service and production control platform includes a station-level platform, a railway bureau-level platform, and a national railway group-level platform, wherein the national railway group-level platform is connected to a plurality of railway bureau-level platforms, and each railway bureau-level platform is connected to a plurality of station-level platforms, and the method includes the following steps:

[0007] The railway bureau-level platform obtains the equipment management data and video code streams of the railway bureau and other railway bureau-level stations from the integrated video surveillance system through the first interface, and manages the obtained equipment management data and video code streams for the passenger section video terminal to access and realize video linkage; the equipment management data includes the following information: camera node directory, camera online status and camera fault list data;

[0008] The station-level platform obtains the station video code stream of the station and its subordinate stations from the integrated video surveillance system with video access nodes through the second interface, and obtains the equipment management data of the station and its subordinate stations through the railway bureau to which the station belongs;

[0009] The station-level platform automatically identifies multiple pre-set key scenes based on the station video code stream and equipment management data of the station and the stations managed by the station, using different analysis models corresponding to different key scenes, and extracts the scene alarm fragments corresponding to each identified key scene according to the identification results, generates a scene alarm fragment file, and uploads the scene alarm fragment file to the railway bureau-level platform to which it belongs in a specific time period to generate a scene alarm fragment file set; wherein the scene alarm fragment file set includes a correct scene alarm fragment file and a false alarm scene alarm fragment file, and the multiple key scenes include all or part of the following scenes: platform intrusion scene, face recognition scene, elevator reverse scene, queue detection scene, leftover object detection scene and headcount scene, and the scene alarm file includes a scene alarm record and a scene alarm fragment;

[0010] The road bureau-level platform optimizes the analysis models corresponding to different key scenarios based on the false alarm scene alarm fragment files to obtain an optimized analysis model set, and forms a relational database based on the scene alarm fragment file set and the key point extraction algorithm;

[0011] The National Railway Group-level platform extracts the metadata from the optimization analysis model collection, scene alarm fragment file collection and relational database for classification and cataloging, generates a resource directory model of the directory structure for providing resource services, and builds an index model. The index model is used to establish a unified positioning of data resources based on the resource directory model; wherein the metadata includes all or part of the following information: file name, model name, field name, field type, field length, field definition description, field business meaning description, the user to which the current data belongs, and the group to which it belongs.

[0012] In some embodiments of the present invention, the method further includes: deploying the optimization analysis model to the corresponding station-level platform by the railway bureau-level platform.

[0013] In some embodiments of the present invention, the relational database includes a first relational database and a second relational database; the scene alarm record includes all or part of the following information: alarm sequence number, alarm location, alarm time, key scene, scene alarm segment storage location and scene alarm segment size;

[0014] A relational database is formed based on the scene alarm fragment file collection and key point extraction algorithm, including:

[0015] Extracting key point features of the scene alarm segment in the scene alarm segment file set based on a key point extraction algorithm, and forming a first relational database based on the extracted key point features;

[0016] A second relational database is formed based on the scene alarm records.

[0017] In some embodiments of the present invention, the method also includes: the railway bureau-level platform and the integrated video surveillance system are securely isolated by firewalls and security isolation devices; the station-level platform and the integrated video surveillance system are isolated by firewalls; wherein the throughput of the firewall and the security isolation device is determined by the number of video access channels and the average storage flow of a single media stream.

[0018] In some embodiments of the present invention, the national railway group-level platform further includes:

[0019] Metadatabase, used to store the optimized analysis model set of the railway bureau-level platform, the scene alarm fragment file set and the metadata in the relational database;

[0020] Data version model, used to store metadata database update and processing records.

[0021] In some embodiments of the present invention, the scene alarm fragment file set is a file set formed by uploading the scene alarm fragment files of each station-level platform under the railway bureau-level platform to the railway bureau-level platform based on the mapping generation definition rule; wherein the mapping generation definition rule includes all or part of the following information: data source definition, data source data domain definition, data update mechanism definition, timestamp definition, data update condition, submission data range and data access permission range definition;

[0022] The metadata database is a database that stores the optimized analysis model set of the bureau-level platform, the scene alarm fragment file set and the metadata in the relational database.

[0023] In some embodiments of the present invention, the metadata database is obtained by:

[0024] Based on the metadata format, the metadata of the optimization analysis model, scene alarm fragment file collection and relational database are standardized to form a metadata model, which exists in the form of key-value pairs, and the metadata model is stored in the national railway group-level platform to generate a metadata database.

[0025] In some embodiments of the present invention, the railway bureau-level platform obtains the equipment management data and video code stream of the stations under the railway bureau and other railway bureaus from the integrated video monitoring system through the first interface, including:

[0026] The railway bureau-level system of the integrated video surveillance system obtains the video code stream and equipment management data of stations under the jurisdiction of other railway bureaus through the video core node and sends it to the railway bureau;

[0027] The railway bureau-level platform obtains the video code stream and equipment management data of the stations under the jurisdiction of the railway bureau and other railway bureaus from the railway bureau-level system corresponding to the comprehensive video monitoring system through the first interface.

[0028] Another aspect of the present invention provides a video data management system for a railway passenger service and production control platform, comprising a processor and a memory, wherein the memory stores computer instructions, and the processor is used to execute the computer instructions stored in the memory. When the computer instructions are executed by the processor, the system implements the steps of the method described in any of the above embodiments.

[0029] Another aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method described in any of the above embodiments when the program is executed by a processor.

[0030] The video data management method and system of the railway passenger service and production control platform of the present invention adopt a two-level access and three-level storage data management method, which can realize different systems and different networks to access video data, obtain equipment management information, optimize the analysis model corresponding to key scenarios and unified data management, and ensure station operation and passenger travel safety.

[0031] Additional advantages, purposes, and features of the present invention will be described in part in the following description, and will become apparent to those skilled in the art after studying the following, or may be learned from the practice of the present invention. The purposes and other advantages of the present invention may be achieved and obtained by the structures specifically indicated in the specification and the accompanying drawings.

[0032] Those skilled in the art will appreciate that the objectives and advantages that can be achieved with the present invention are not limited to the above specific description, and the above and other objectives that can be achieved by the present invention will be more clearly understood from the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The drawings described herein are used to provide a further understanding of the present invention, constitute a part of the present application, and do not constitute a limitation of the present invention. In the drawings:

[0034] Figure 1 The figure is a flow chart of a video data management method for a railway intelligent station in one embodiment of the present invention.

[0035] Figure 2 The topological structure diagram of video data access and data management of a railway intelligent station and an integrated video surveillance system in one embodiment of the present invention.

[0036] Figure 3 The logical structure diagram of data access between the railway bureau-level platform of the intelligent station and the railway bureau-level system of the integrated video monitoring system in one embodiment of the present invention.

[0037] Figure 4 The structure diagram of the station-level platform and the small stations managed by the railway intelligent station in one embodiment of the present invention.

[0038] Figure 5 It is a structural diagram of the three-level video data management of a railway intelligent station in one embodiment of the present invention. DETAILED DESCRIPTION

[0039] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments and the accompanying drawings. Here, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0040] It should also be noted that, in order to avoid obscuring the present invention due to unnecessary details, only structures and / or processing steps closely related to the solutions according to the present invention are shown in the accompanying drawings, while other details that are not closely related to the present invention are omitted.

[0041] It should be emphasized that the term “include / comprises” when used herein refers to the presence of features, elements, steps or components, but does not exclude the presence or addition of one or more other features, elements, steps or components.

[0042] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In the accompanying drawings, the same reference numerals represent the same or similar components, or the same or similar steps.

[0043] In railway passenger transport, smart stations and integrated video surveillance systems are independent systems, and their networks belong to the office network and integrated video dedicated network respectively, which limits the network bandwidth between the railway bureau and the station for these two systems. Currently, stations usually access video streams by directly connecting to the camera, which does not meet the requirements of railway safety production. In addition, due to the small number of video data samples at the station, the analysis accuracy of the analysis model for key scenes is also affected to a certain extent.

[0044] Based on this, the embodiment of the present invention provides a video data management method for a railway passenger service and production control platform, which implements two-level video data access between the intelligent station and the integrated video surveillance system based on a specific data access logic, and performs three-level storage data management in the intelligent station. The method can meet the needs of video data access, analysis of key scenes in the video, and data management with data indexing as the goal, and can effectively improve safety supervision efficiency and service quality, and enhance the safety level of the station.

[0045] Figure 1 FIG. 1 is a flow chart of a video data management method for a railway intelligent station according to an embodiment of the present invention. Figure 1 The method shown may specifically include steps S110 to S150.

[0046] The railway passenger service and production management and control platform includes station-level platforms, bureau-level platforms and national railway group-level platforms. The national railway group-level platform connects all bureau-level platforms, and each bureau-level platform is connected to multiple station-level platforms.

[0047] Step S110: The railway bureau-level platform obtains the equipment management data and video code streams of the stations under the jurisdiction of the railway bureau and other railway bureaus from the integrated video surveillance system through the first interface, and manages the obtained equipment management data and video code streams for viewing by the passenger section video terminal to achieve video linkage; among which, the equipment management data includes the following information: camera node directory, camera online status and camera fault list data.

[0048] In some embodiments of the present invention, the railway bureau-level platform obtains the equipment management data and video code stream of the stations under the railway bureau and other railway bureaus from the integrated video monitoring system through the first interface, including:

[0049] The railway bureau-level system of the integrated video surveillance system obtains the video code stream and equipment management data of stations under the jurisdiction of other railway bureaus through the video core node and sends it to the railway bureau;

[0050] The railway bureau-level platform obtains the video code streams of the stations under the jurisdiction of the railway bureau and other railway bureaus from the railway bureau-level system corresponding to the comprehensive video monitoring system through the first interface.

[0051] More specifically, Figure 2FIG. 1 is a topological diagram of video data access and management of a railway intelligent station according to an embodiment of the present invention. Figure 2 As shown, the intelligent station adopts a two-level access and three-level storage application architecture, including: the railway bureau-level platform and the station-level platform can obtain video code streams and equipment management data from the integrated video surveillance system; the station-level platform forms scene alarm fragment files based on key scene analysis, the railway bureau-level platform can form scene alarm fragment file sets, optimized analysis model sets and relational databases based on scene alarm fragment files, and the railway bureau-level platform can index the data of the railway bureau-level platform at the national railway group-level platform. Furthermore, the video core node of the integrated video surveillance system obtains the equipment management data of other railway bureaus from the video area nodes of other railway bureaus and transmits it to the video area node of the railway bureau of the integrated video surveillance system, and then transmits it to the railway bureau-level platform corresponding to the intelligent station through the firewall, security isolation equipment and firewall through the first interface (at this time, each railway bureau has the equipment management data of the railway bureau and other railway bureaus). Afterwards, the railway bureau-level platform of the intelligent station can obtain the real-time video code stream of the designated camera of the railway bureau and other railway bureaus from the integrated video surveillance system according to the equipment number in the camera node directory. Among them, the equipment management data includes the following information: camera node directory, camera online status and camera fault list data.

[0052] In some embodiments of the present invention, the method also includes: the bureau-level platform and the integrated video surveillance system are securely isolated through firewalls and security isolation devices; wherein the throughput of the firewall and the security isolation device is determined by the number of video access channels and the average storage flow of a single media stream.

[0053] Furthermore, to ensure the safety of the station, when the railway bureau obtains equipment management data from the integrated video surveillance system, it should set up firewalls and security isolation devices (gateways) to isolate the network between the smart station and the integrated video system. The throughput of the firewall and security isolation equipment should be determined based on the number of video access channels and the average storage flow of a single media stream. Among them, the number of access video channels is affected by the bandwidth between the railway bureau-level platform and the passenger section, the encoding type of the video stream (H.264 or H.265), the image resolution and frame rate of the camera, etc. For example, when the 100M bandwidth supports a frame rate of 25fps, uses the H.264 compression encoding format, and the image resolution is 4K, the number of access video channels is no more than 1024÷16=64.

[0054] Furthermore, the railway bureau-level platform sends the acquired equipment management data and real-time video code streams of the stations under the jurisdiction of this railway bureau and other railway bureaus to the video access unit of the railway bureau-level platform. The video access unit encapsulates the video code stream, and the video application unit manages the equipment management data and video code stream. The video terminal of the passenger section sends a specific video request to the video application unit of the railway bureau through the switch, and the video access unit forwards the encapsulated video code stream to the video terminal for linkage display. The passenger section is used to view the real-time stop status of a train of this railway bureau or other railway bureaus at the platform in real time based on the train arrival and departure plan using the video code streams of the stations under the jurisdiction of this railway bureau and other railway bureaus.

[0055] Step S120: The station-level platform obtains the station video code stream of the station and the stations under its jurisdiction from the integrated video surveillance system with a video access node through the second interface, and obtains the equipment management data of the station and the stations under its jurisdiction through the railway bureau to which the station belongs.

[0056] More specifically, the data accessed by the station-level platform with video access nodes from the integrated video surveillance system station through the second interface includes the real-time station video code stream and historical station video code stream of the station and the small stations managed by the station. And the station-level platform with video access nodes can obtain the equipment management data of the station and the small stations managed by the station from the railway bureau-level platform to which it belongs through the video application unit, which corresponds one-to-one with the station video code stream through the equipment coding (i.e., the equipment ID obtained by the railway equipment based on specific coding rules).

[0057] As an example, a station-level platform with a video access node of a smart station accesses data from a station-level system with a video access node of a comprehensive video surveillance system.

[0058] In some embodiments of the present invention, the method further includes: the station-level platform and the integrated video surveillance system are isolated by a firewall; wherein the throughput of the firewall and the security isolation device is determined by the number of video access channels and the average storage flow of a single media stream.

[0059] More specifically, the intelligent station-level platform and the integrated video surveillance system-level system should be isolated by setting up a firewall. The throughput of the firewall is determined based on the number of video access channels and the average storage flow of a single media stream.

[0060] As an example, the first interface and the second interface in the present invention are different interfaces. The first interface refers to the interface for data transmission between the railway bureau-level platform of the smart station and the railway bureau-level system of the integrated video surveillance system, and the second interface refers to the interface for data transmission between the station-level platform of the smart station and the station-level system of the integrated video surveillance system. However, the first interface and the second interface can connect the smart station and the integrated video surveillance system based on the same interface protocol, such as the GB / T28181-2016 standard interface protocol.

[0061] As an example, the logical structure of data access between the railway bureau-level platform of the intelligent station and the railway bureau-level system of the integrated video surveillance system of an embodiment of the present invention is as follows: Figure 3 As shown, including:

[0062] The interface of the smart station is connected to the integrated video system through the GB / T 28181-2016 standard interface. Through the national standard interface, basic information and video code streams such as camera node directory, camera information query, camera status query, etc. can be obtained. Through the streaming media service, the reception of the original code stream, the encoding and decoding of the video code stream, transcoding, distribution, media packaging, etc. can be analyzed, and the frame rate, maximum code rate, minimum code rate, cache, resolution, color reference, encoding format, etc. of the code stream can be modified; at the same time, the best encoding format can be obtained based on the best code stream situation and the realized code stream format, which can optimize the phenomenon of frame supplementation and delay to a large extent. The surveillance image code stream format (including RTSP, FLV-H.264 / H.265 or other standard encoding formats) obtained by transcoding the video code stream supports the GB / T 28181-2016 standard protocol, supports the http protocol to obtain device management information, and can realize node directory resource synchronization, camera status query and device status reporting, and provide push streaming services according to the specified RTSP / FLV streaming media protocol.

[0063] Step S130: The station-level platform automatically identifies multiple pre-set key scenes based on the station video code stream and equipment management data of the station and the stations managed by the station, using different analysis models corresponding to different key scenes, and extracts scene alarm clips corresponding to each identified key scene according to the identification results, generates scene alarm clip files, and uploads the scene alarm clip files to the railway bureau-level platform to which they belong in a specific time period to generate a scene alarm clip file set; wherein the scene alarm clip file set includes correct scene alarm clip files and false alarm scene alarm clip files, and the multiple key scenes include all or part of the following scenes: platform intrusion scene, face recognition scene, elevator reverse scene, queue detection scene, residue detection scene and headcount scene, and the scene alarm file includes scene alarm record and scene alarm clip.

[0064] More specifically, if Figure 2 As shown, the station-level platform obtains the station video code stream from the station-level system where the video access node of the integrated video surveillance system is located through the firewall. The video access unit of the station-level platform encapsulates the obtained station video code stream and equipment management data of the station and the small stations managed by the station and forwards them to the video analysis unit of the station-level platform respectively. The video analysis unit pre-stores different analysis models corresponding to different key scenes. It can perform intelligent analysis on the key scenes in the station video code stream based on the analysis model, generate scene alarm segment files, and send the scene alarm segment files to the video application service unit for access and call by the video terminals of the station and the small stations managed by the station. The video application service unit provides a visualization application program for intelligent video analysis applications and video linkage. The video terminal realizes real-time video access of the platform, ticket gate, and exit gate according to the arrival and departure plan by connecting to the camera through the video application service unit.

[0065] Furthermore, the video analysis unit performs intelligent analysis on the key scenes in the station video stream based on the analysis model to generate a scene alarm clip file, including: inputting the station video stream into different analysis models corresponding to different key scenes respectively; for multiple pre-set key scenes, the different analysis models corresponding to different key scenes first automatically identify each key scene in the station video stream, and extract the scene alarm clips of a period of time before and after the time point of each identified key scene (for example, obtaining a 10s scene alarm video clip); the scene alarm clips and scene alarm records together constitute the scene alarm clip file.

[0066] Among them, the key scenarios include: face recognition scenarios (such as face recognition at the entrance), elevator reverse scenarios (or escalator reverse), platform intrusion scenarios, queue detection scenarios (such as queue detection at the entrance and ticket gate), people counting scenarios (such as sudden gatherings), relics detection scenarios and waiting room fight scenarios.

[0067] As an example, the scene alarm segment includes the scene alarm segment corresponding to each key scene extracted from the station video code stream of the station and the stations under the jurisdiction of the station.

[0068] As an example, the structure of the intelligent station station-level platform and the small stations it manages is as follows: Figure 4 The station-level platform of a larger station has a video access unit, a video analysis unit, a video application unit and a video terminal, which can analyze key scenes of the acquired video data and display real-time video images. The small stations under the jurisdiction of the large station (usually large stations and small stations are distinguished according to the station scale) only have video terminals, which can be used to access and retrieve data, but cannot analyze key scenes. Therefore, the large station usually conducts intelligent analysis of different key scenes on the station video code streams of the station and the small stations under its jurisdiction.

[0069] In some embodiments of the present invention, a scene alarm fragment file set is a file set formed by uploading scene alarm fragment files of each station-level platform under the railway bureau-level platform to the railway bureau-level platform based on a mapping generation definition rule; wherein the mapping generation definition rule includes all or part of the following information: data source definition, data source data domain definition, data update mechanism definition, timestamp definition, data update conditions, submission data range, and data access permission range definition.

[0070] Furthermore, the video application unit of each station-level platform stores each scene alarm segment file information to the corresponding railway bureau-level platform based on the mapping generation definition rule, and generates a scene alarm segment file set in the video storage unit of the railway bureau-level platform.

[0071] As an example, since the analysis model may have scene recognition errors or scene failures when performing intelligent recognition and analysis on key scenes, such scene alarm clip files are classified as false alarm scene alarm clip files; scene alarm clip files that correctly recognize the scene and give an alarm are classified as correct scene alarm clip files. Therefore, the scene alarm clip file set includes correct scene alarm clip files and false alarm scene alarm clip files.

[0072] As an example, the data stored in each station-level platform (referred to as station data for short) are all entity data. The entity data of the station-level platform include different analysis models and scenario alarm fragment files corresponding to different key scenarios.

[0073] As an example, each station-level platform uploads the entity data generated every day to the railway bureau-level platform for storage during the idle passenger period at night through the safety production network between the station-level platform and the railway bureau-level platform.

[0074] As an example, both the station-level platform and the railway bureau-level platform store different analysis models corresponding to different key scenarios. The analysis model of the station-level platform is used to obtain scene alarm fragment files, and the analysis model of the railway bureau-level platform is used to optimize the false alarm scene alarm fragment files to obtain an optimized analysis model, and deploy the optimized analysis model to the station-level platform.

[0075] Step S140: The railway bureau-level platform optimizes the analysis models corresponding to different key scenarios based on the false alarm scene alarm fragment files to obtain an optimized analysis model set, and forms a relational database based on the scene alarm fragment file set and the key point extraction algorithm.

[0076] More specifically, the railway bureau-level platform uses the false alarm scene alarm clip files of the stations under the jurisdiction of the railway bureau within a certain period of time (such as 90 days) to optimize the training of different analysis models corresponding to different key scenes in the video algorithm training unit to improve the accuracy of the analysis model's recognition and alarm of key scenes, and obtain different optimized analysis models (i.e., optimized analysis model sets) corresponding to different key scenes. In addition, the railway bureau-level platform deploys each optimized analysis model to the corresponding station-level platform, which can be used to replace the unoptimized analysis model in the station-level platform.

[0077] In some embodiments of the present invention, the relational database includes a first relational database and a second relational database; the scene alarm record includes all or part of the following information: alarm sequence number, alarm location, alarm time, key scene, scene alarm segment storage location and scene alarm segment size;

[0078] A relational database is formed based on the scene alarm fragment file collection and key point extraction algorithm, including:

[0079] Extracting key point features of the scene alarm segment in the scene alarm segment file set based on a key point extraction algorithm, and forming a first relational database based on the extracted key point features;

[0080] A second relational database is formed based on the scene alarm records.

[0081] More specifically, the railway bureau-level platform stores a collection of scene alarm fragment files, and the alarm fragment files include scene alarm records and scene alarm fragments. Therefore, the key point features of the monitoring images of the scene alarm fragments can be extracted based on the key point extraction algorithm to form a first relational database (for example, the key point features (i.e., feature vectors) of the scene alarm fragments of the platform intrusion scene are extracted based on the ISS algorithm (intrinsic shape signatures) to obtain a key point relationship model); a second relational database is formed based on the scene alarm records (for example, a relationship model is formed based on the alarm position and key scenes of the platform intrusion scene of Station A; the relationship model of the alarm record can be formed according to the key scene classification to form a relationship model, and can also be formed according to the station to which the scene alarm fragment file belongs, and so on).

[0082] The data stored in the railway bureau-level platform (referred to as railway bureau data) is entity data, including a set of optimization analysis models, a set of scene alarm fragment files, and a relational database. The number of key scenarios, models, and different models corresponding to different key scenarios on the railway bureau-level platform should not be less than that on the station-level platform.

[0083] Step S150: The National Railway Group-level platform extracts the metadata from the optimization analysis model set, the scene alarm fragment file set and the relational database for classification and cataloging, generates a resource directory model of a directory structure for providing resource services, and constructs an index model, which is used to establish a unified positioning of data resources based on the resource directory model; wherein the metadata includes all or part of the following information: file name, model name, field name, field type, field length, field definition description, field business meaning description, the user to whom the current data belongs, and the group to which it belongs.

[0084] More specifically, the resource directory model is a hierarchical and classified management directory formed by collecting metadata of road bureau data (that is, extracting metadata of road bureau data of each road bureau-level platform to form a metadata database and resource directory model); and an index model, which is used to index and locate the metadata of road bureau data based on the resource directory model to realize virtualized data link management.

[0085] As an example, the resource directory model can be expressed in the form of a directory tree or a table, and the resource directory model can also generate a management directory through the metadata of the metadata database, and then establish a unified resource location for the road bureau data based on the management directory, and then perform hierarchical and classified management of the road bureau data. The index model can locate specific road bureau data based on the resource directory model.

[0086] In some embodiments of the present invention, the national railway group-level platform further includes:

[0087] Metadatabase, used to store the optimized analysis model set of the railway bureau-level platform, the scene alarm fragment file set and the metadata in the relational database;

[0088] The data version model is used to store the update records and processing records of the metadata database.

[0089] In some embodiments of the present invention, the metadata database is a database that stores a set of optimization analysis models of a railway bureau-level platform, a set of scene alarm fragment files, and metadata in a relational database.

[0090] In some embodiments of the present invention, the metadata database is obtained by:

[0091] Based on the metadata format, the metadata of the railway bureau data is standardized to form a metadata model so that it exists in the form of key-value pairs, and the metadata model is stored in the national railway group-level platform to generate a metadata database.

[0092] More specifically, the railway bureau-level platform can extract metadata of the optimization analysis model set, metadata of the scene alarm fragment file set, and metadata in the relational database to generate a metadata model. A metadata database is generated on the national railway group-level platform, and the metadata database is used to store the metadata of the extracted railway bureau data.

[0093] like Figure 2 As shown, the video metadata storage unit of the national railway group-level platform stores metadata information of railway bureau data (which can be used to describe data attributes, storage location and usage). It is a distributed data lake formed by mapping the railway bureau data of each railway bureau-level platform to different domain databases based on mapping definition rules, which can cover the monitoring video resources of all railway bureaus.

[0094] As an example, metadata includes the following information: file name, model name, field name, field type, field length, field definition description, field business meaning description, user or group to which the current data belongs, etc. The above information is only an example, and the present invention is not limited thereto.

[0095] In a specific embodiment of the present invention, Figure 5 As shown in the figure, the three-level video data management structure of the railway intelligent station includes:

[0096] (1) Station-level platform

[0097] The entity data model of the station-level platform is A. Model A includes scene alarm fragment files C and video analysis model D obtained based on key scene analysis, where A = {C1, C2, …, C t}∪{D1,D2…,D t}, t is the number of key scenes at the station level platform. i is the scene alarm fragment file of the i-th key scene on the station-level platform, D i It is the analysis model corresponding to the i-th key scenario in the station-level platform.

[0098] (2) Railway Bureau-level Platform

[0099] The entity data model of the railway bureau-level platform is B, which includes a scene alarm segment file set G, a video optimization analysis model set H, and a relational database set L. Among them, G = {G1, G2, ..., G k}, k is the number of stations under the railway bureau, G i is the scene alarm clip file of the i-th station; H = {H1, H2, ..., H n}, H i ={D m1 ,D m2 ,…,D mn}, where m is the number of algorithm models for the t-th key scenario, m = 1, 2, ..., n, n is the number of types of optimization analysis models for the bureau-level platform, H n is the optimization analysis model corresponding to the nth key scenario of the railway bureau-level platform, D mn is the number of algorithm models of the nth type of key scenarios, and n ≥ t (the number of key scenarios corresponding to the optimization analysis model of the railway bureau-level platform is not less than the number of key scenarios corresponding to the analysis model of the station-level platform); L is a data organization consisting of a two-dimensional table of all alarm records of k stations in a certain period of time (such as 90 days) and a two-dimensional table of key point features of scene alarm fragments, and the relationship between their two-dimensional tables.

[0100] For example, the scene alarm clip file C can be represented as a collection of column vectors, C i ={h1,h2,…,h x}, h x There are x files for the i-th key scene of the station. The scene alarm clip file set G can also be expressed as a set of column vectors, G i ={q1,q2,…,q k}, q k is the number of files of the kth station of the i-th key scene of the railway bureau. G actually constitutes a three-dimensional record numerical model. In addition, in B, a mapping from C to G is generated according to the entity model and the data mapping model to generate a definition rule R. Under the action of rule R, different C i Integration can generate G i , thus producing G.

[0101] The metadata model P is obtained by extracting metadata from the scene alarm fragment file set G, the optimization analysis model set H and the relational database model set L.

[0102] (3) National Railway Group-level platform

[0103] The metadata database W of the national railway group-level platform is a standardized data set obtained by the metadata model P based on the mapping definition rule R. The data version model is a record of the orderly storage, update and processing of data. The resource directory model is used to establish a unified data resource directory management system, and the virtualized data link can establish a unified resource locator for the entire railway data through the index model, including but not limited to user name and password.

[0104] In a specific embodiment of the present invention, the process of alarming for a platform intrusion scenario at a station-level platform is as follows:

[0105] (1) Before the train enters the station and stops, focus on monitoring the yellow line entry area of ​​the platform;

[0106] (2) A pedestrian re-identification algorithm is used to identify a person. If a pedestrian crosses the yellow line and the distance from the yellow line exceeds a first set threshold (e.g., 10 cm), an alarm is issued directly; if a pedestrian crosses the yellow line and the distance from the yellow line is greater than a second set threshold (the second set threshold is less than the first set threshold, e.g., 5 cm), and less than the first set threshold, and the distance range is maintained for more than 10 seconds, a loitering alarm is issued;

[0107] (3) Capture the scene alarm clips of 5 seconds before and after the alarm, and record the alarm record information such as the alarm location and alarm time to generate the alarm clip file of the platform intrusion.

[0108] The video data management method for railway intelligent stations proposed by the present invention has the following advantages:

[0109] 1. A unified data management method with two-level video data access and three-level video data application is adopted, fully considering the impact of different systems, different networks and bandwidth limitations on access video data, and fully considering video data analysis applications and storage methods to ensure station operations and passenger travel safety.

[0110] 2. The two-level access video data access can meet the 100M bandwidth requirements between passenger sections, stations and railway bureaus, and at the same time meet the business requirements of passenger section video linkage and passenger station intelligent video analysis. At the same time, it is designed and solves the problems of frame loss, delay and inconsistent data code stream application format in video naked stream. By changing the existing station video to use the camera direct connection to access video data, the problem of being unable to obtain node directory, camera status and fault list data is overcome, and it does not meet the requirements of one machine and one network in railway safety production.

[0111] 3. The tertiary storage application realizes the full routing and coordination of all physical nodes of the national railway bureau database, alarm clip files and video analysis models by building a virtualized data index network. The radial and distributed data lake formed solves the problems of large storage requirements, large bandwidth consumption, and long time in the centralized construction of the data lake and realizes data sharing.

[0112] 4. By using a large number of scene alarm video samples on the railway bureau-level platform to train and optimize the video analysis model, and deploying the optimized analysis model on the station-level platform, the accuracy of identifying key scenes can be improved, and the real-time requirements of the station can be better met.

[0113] Corresponding to the above method, the present invention also provides a system, which includes a computer device, the computer device includes a processor and a memory, the memory stores computer instructions, the processor is used to execute the computer instructions stored in the memory, and when the computer instructions are executed by the processor, the system implements the steps of the method described above.

[0114] The embodiment of the present invention also provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the steps of the aforementioned edge computing server deployment method are implemented. The computer-readable storage medium can be a tangible storage medium, such as a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a floppy disk, a hard disk, a removable storage disk, a CD-ROM, or any other form of storage medium known in the technical field.

[0115] It should be understood by those skilled in the art that the exemplary components, systems and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software or a combination of the two. Whether it is performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present invention are programs or code segments used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link via a data signal carried in a carrier.

[0116] It should be clear that the present invention is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present invention.

[0117] In the present invention, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with features of other embodiments or replace features of other embodiments.

[0118] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the embodiments of the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A video data management method for a railway passenger service and production control platform, wherein the railway passenger service and production control platform includes a station-level platform, a railway bureau-level platform and a national railway group-level platform, wherein the national railway group-level platform is connected to multiple railway bureau-level platforms, and each railway bureau-level platform is connected to multiple station-level platforms, characterized in that: The method comprises the following steps: The railway bureau-level platform obtains the equipment management data and video code streams of the railway bureau and other railway bureau-level stations from the integrated video surveillance system through the first interface, and manages the obtained equipment management data and video code streams for the passenger section video terminal to access and realize video linkage; wherein the equipment management data includes the following information: camera node directory, camera online status and camera fault list data; The station-level platform obtains the station video code stream of the station and its subordinate stations from the integrated video surveillance system with video access nodes through the second interface, and obtains the equipment management data of the station and its subordinate stations through the railway bureau to which the station belongs; The station-level platform automatically identifies multiple pre-set key scenes based on the station video code stream and equipment management data of the station and the stations managed by the station, using different analysis models corresponding to different key scenes, and extracts the scene alarm fragments corresponding to each identified key scene according to the identification results, generates a scene alarm fragment file, and uploads the scene alarm fragment file to the railway bureau-level platform to which it belongs in a specific time period to generate a scene alarm fragment file set; wherein the scene alarm fragment file set includes a correct scene alarm fragment file and a false alarm scene alarm fragment file, and the multiple key scenes include all or part of the following scenes: platform intrusion scene, face recognition scene, elevator reverse scene, queue detection scene, relic detection scene and headcount scene, and the scene alarm file includes a scene alarm record and a scene alarm fragment; The railway bureau-level platform optimizes the analysis models corresponding to different key scenarios based on the false alarm scenario alarm fragment files to obtain an optimized analysis model set, and forms a relational database based on the scenario alarm fragment file set and the key point extraction algorithm; The national railway group-level platform extracts the metadata from the optimization analysis model set, the scene alarm fragment file set and the relational database for classification and cataloging, generates a resource directory model of the directory structure for providing resource services, and constructs an index model, which is used to establish a unified positioning of data resources based on the resource directory model; wherein the metadata includes all or part of the following information: file name, model name, field name, field type, field length, field definition description, field business meaning description, the user to which the current data belongs, and the group to which it belongs.

2. The method according to claim 1, characterized in that The method also includes: the railway bureau-level platform deploys the optimization analysis model to the corresponding station-level platform.

3. The method according to claim 1, characterized in that The relational database includes a first relational database and a second relational database; the scene alarm record includes all or part of the following information: alarm sequence number, alarm location, alarm time, key scene, scene alarm segment storage location and scene alarm segment size; The relational database formed based on the scene alarm segment file set and the key point extraction algorithm includes: Extracting key point features of the scene alarm segment in the scene alarm segment file set based on a key point extraction algorithm, and forming a first relational database based on the extracted key point features; A second relational database is formed based on the scene alarm records.

4. The method according to claim 1, characterized in that: The method also includes: the railway bureau-level platform and the integrated video surveillance system are securely isolated by firewalls and security isolation devices; the station-level platform and the integrated video surveillance system are isolated by firewalls; wherein the throughput of the firewall and the security isolation device is determined by the number of video access channels and the average storage flow of a single media stream.

5. The method according to claim 1, characterized in that The national railway group-level platform also includes: Metadatabase, used to store the optimized analysis model set of the railway bureau-level platform, the scene alarm fragment file set and the metadata in the relational database; Data version model, used to store metadata database update and processing records.

6. The method according to claim 1 or 5, characterized in that: The scene alarm fragment file set is a file set formed by uploading the scene alarm fragment files of each station-level platform under the railway bureau-level platform to the railway bureau-level platform based on the mapping generation definition rule; wherein the mapping generation definition rule includes all or part of the following information: data source definition, data source data domain definition, data update mechanism definition, timestamp definition, data update condition, submission data range and data access permission range definition; The metadata database is a database for storing a set of optimization analysis models of a railway bureau-level platform, a set of scene alarm fragment files, and metadata in a relational database.

7. The method according to claim 5, characterized in that The metadata database is obtained by: Based on the metadata format, the metadata of the optimization analysis model, scene alarm fragment file collection and relational database are standardized to form a metadata model, which exists in the form of key-value pairs, and the metadata model is stored in the national railway group-level platform to generate a metadata database.

8. The method according to claim 1, characterized in that The railway bureau-level platform obtains the equipment management data and video code streams of the stations under the jurisdiction of the railway bureau and other railway bureaus from the integrated video monitoring system through the first interface, including: The railway bureau-level system of the integrated video surveillance system obtains the video code stream and equipment management data of stations under the jurisdiction of other railway bureaus through the video core node and sends it to the railway bureau; The railway bureau-level platform obtains the video code stream and equipment management data of the stations under the jurisdiction of the railway bureau and other railway bureaus from the railway bureau-level system corresponding to the comprehensive video monitoring system through the first interface.

9. A video data management system for a railway passenger service and production control platform, comprising a processor and a memory, characterized in that: The memory stores computer instructions, and the processor is used to execute the computer instructions stored in the memory. When the computer instructions are executed by the processor, the system implements the steps of the method as claimed in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.