CN113612970A - Safety event intelligent analysis management and control platform for industrial monitoring video - Google Patents
Safety event intelligent analysis management and control platform for industrial monitoring video Download PDFInfo
- Publication number
- CN113612970A CN113612970A CN202110874184.4A CN202110874184A CN113612970A CN 113612970 A CN113612970 A CN 113612970A CN 202110874184 A CN202110874184 A CN 202110874184A CN 113612970 A CN113612970 A CN 113612970A
- Authority
- CN
- China
- Prior art keywords
- algorithm
- management
- alarm
- platform
- analysis
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004458 analytical method Methods 0.000 title claims abstract description 80
- 238000012544 monitoring process Methods 0.000 title claims abstract description 49
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 108
- 238000000034 method Methods 0.000 claims abstract description 21
- 238000001914 filtration Methods 0.000 claims abstract description 17
- 238000001514 detection method Methods 0.000 claims abstract description 10
- 238000013480 data collection Methods 0.000 claims abstract description 9
- 238000004806 packaging method and process Methods 0.000 claims abstract description 4
- 230000008569 process Effects 0.000 claims description 19
- 238000012545 processing Methods 0.000 claims description 6
- 230000007246 mechanism Effects 0.000 claims description 4
- 230000000694 effects Effects 0.000 claims description 3
- 210000001503 joint Anatomy 0.000 claims description 2
- 238000007726 management method Methods 0.000 description 80
- 230000006870 function Effects 0.000 description 9
- 238000003032 molecular docking Methods 0.000 description 6
- 238000003860 storage Methods 0.000 description 6
- 239000000243 solution Substances 0.000 description 5
- 238000004519 manufacturing process Methods 0.000 description 4
- 238000012216 screening Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 230000006835 compression Effects 0.000 description 2
- 238000007906 compression Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 230000009191 jumping Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000002688 persistence Effects 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000019771 cognition Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 238000002347 injection Methods 0.000 description 1
- 239000007924 injection Substances 0.000 description 1
- 238000012432 intermediate storage Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
- H04N7/181—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a plurality of remote sources
Abstract
The invention discloses an industrial monitoring video oriented security event intelligent analysis and control platform, and belongs to the field of video analysis. The method comprises the following steps: the user authority management module is used for managing user information and user authority; the data collection and collection module is used for automatically pulling the industrial monitoring video according to the configuration of the camera and the collection time, matching and binding the camera with the filtering algorithm, filtering and packaging the pulled video and providing a corresponding downloading interface; the algorithm management module is used for managing versions of a filtering algorithm and an industrial monitoring video analysis algorithm; the alarm strategy management module is used for automatically pulling the monitoring video and running a selected analysis algorithm for detection according to the configured camera and the warning time, and pushing when a target event is detected; and the alarm information management module is used for inquiring, managing and visualizing alarm information and supporting an open API (application program interface) to inquire alarm records. The invention improves the flexibility of algorithm management while ensuring the safety of the industrial monitoring video.
Description
Technical Field
The invention belongs to the field of intelligent video analysis, and particularly relates to an intelligent analysis and control platform for industrial monitoring video-oriented security events.
Background
In the safety production management of industrial enterprises, most of the video analysis systems on the market basically realize the functions of detecting and analyzing abnormal events in specific images or videos, and can be pushed to a platform or terminal equipment according to the detection result, so that the production management efficiency is greatly improved. However, by investigating and analyzing the intelligent video analysis platform proposed by the academia and industry at present, some problems are still found to exist:
1. data security and early warning real-time problems. In the video analysis solution provided by part of service enterprises, video data are uploaded to a public network cloud end for analysis and processing in a network calling mode, and processing results are returned to an enterprise early warning platform to generate an alarm. The intelligent video analysis algorithm belongs to a core algorithm related to enterprise safety production management activities, has complete independent intellectual property rights and software copyright rights, and is not allowed to be deployed to a service platform in a public network for operation processing. Many scene information in traditional industrial enterprises and equipment point location information under the monitoring network relate to internal secrets of the enterprises, and data security needs to be considered at the same time. The mode is difficult to ensure on the information security, and the mode based on API call has larger network transmission delay and is not suitable for the scene with higher requirements on early warning real-time property.
2. Monitoring the deployment and management of video analysis algorithms. Although the mainstream solutions proposed in academia and industry at present solve many problems of large-scale monitoring video processing in complex scenes, algorithms are embedded in a platform, and third-party algorithm deployment and online problems except for platform providers and enterprise users are not considered. The actual scene requirements of specific industrial enterprises are not considered in the aspects of setting algorithm threshold parameters, flexibly matching and binding an analysis algorithm and a camera, managing algorithm versions under similar analysis tasks and the like.
3. And the video analysis task management granularity is low. Most of the existing video analysis algorithms mostly adopt a 24-hour uninterrupted operation mode, and continuously acquire video stream data and output results when an abnormal detection event occurs. The management granularity of the video analysis is low, and although the video analysis meets the actual scene requirements to a certain extent, long-term online operation can cause meaningless occupation of server computing resources, so that the overall use experience of the platform is influenced. Considering the requirements such as analysis in a specific time period of a certain camera in a monitoring network, the time fine granularity of task management is improved, and the purposes of reasonable resource scheduling and peak-shifting use are achieved.
4. And the alarm information is not fully utilized. The alarm information is important information in the safety production management process of the industrial enterprise, and the full utilization of the alarm information is the purpose of intelligent analysis of the video data by the industrial enterprise. At present, platforms developed by service providers such as Haicanwei and the like mainly provide general video analysis services, do not perform personalized customization services aiming at alarm information services of specific enterprises, and do not provide linkage of alarm information in combination with other third-party application platforms such as intelligent broadcast push platforms and the like which are already put into use or planned to be put into use by the specific enterprises.
5. The use requirements of a third party are not considered. At present, a mainstream video analysis platform mainly faces to specific analysis requirements of enterprises, mainly completes self functional services, and does not consider the use requirements of third-party algorithm developers. After investigation and analysis, a large amount of mechanical and repeated data set acquisition and screening work exists in the algorithm pre-training process of an algorithm developer, and the process consumes excessive human resources, so that the development period of the analysis algorithm is prolonged. The intelligent analysis platform is used by considering the requirements of algorithm developers, dividing different user use permissions and standardizing the platform use process of different roles such as the algorithm developers, enterprise security management personnel and the like.
Disclosure of Invention
Aiming at the defects and improvement requirements of the prior art that algorithm management is not flexible, task management granularity is coarse, alarm information is not fully utilized, third party requirements are not considered, and the like, the invention provides an industrial monitoring video-oriented security event intelligent analysis management and control platform, and aims to improve the flexibility of algorithm management while ensuring the security of an industrial monitoring video.
In order to achieve the above object, according to an aspect of the present invention, there is provided an intelligent analysis management and control platform for security events of industrial surveillance videos, including:
the user authority management module is used for managing user information and user authority, and defining three different roles, wherein each role has corresponding authority;
the data collection and collection module is used for automatically pulling the industrial monitoring video under the monitoring network according to the configured camera and the collection time, matching and binding the pull camera with the filtering algorithm, filtering and packaging the pulled industrial monitoring video and providing a corresponding downloading interface;
the algorithm management module is used for managing versions of the filtering algorithm and the industrial monitoring video analysis algorithm, including uploading, deployment and management of the algorithm and supporting parallel operation of different versions of the algorithm;
the alarm strategy management module is used for automatically pulling the corresponding industrial monitoring video under the monitoring network according to the configured camera and the warning time, running a selected analysis algorithm for detection, pushing when a target event is detected, and supporting the self-definition of an alarm strategy;
and the alarm information management module is used for inquiring, managing and visualizing alarm information and supporting an open API (application program interface) to inquire alarm records.
Has the advantages that: the data set acquisition module considers the requirements of third-party algorithm developers, perfects the data acquisition process facing the industrial monitoring video, adds a certain filtering strategy to help filter and screen the data set, reduces the data acquisition work difficulty of the algorithm developers through the data set acquisition and filtering process, and accelerates the model pre-training process and the algorithm development cycle.
Preferably, the three roles governing the platform include:
a hypervisor, which has the following privileges: data set collection, user authority definition, algorithm uploading and management, alarm strategy configuration, alarm record inquiry and API interface butt joint with a third-party platform;
the general administrator, who has the following rights: uploading and managing an algorithm, configuring an alarm strategy and inquiring an alarm record;
the ordinary user, it possesses the following rights: and inquiring alarm records.
Preferably, in the algorithm management module, in the algorithm uploading process, the running environment, the program entity and the parameters of the application program are decoupled, wherein the running environment is stored in a mirror warehouse in a docker mirror image form, and is locally deployed in a docker container form during use, the program entity algorithm is placed in a server root directory in a source code or executable program form, and is uploaded to the container and packaged into a mirror image during use, and the parameters are stored in the server root directory in a json file form and are sent to the program entity in a network package form during use.
Has the advantages that: decoupling the algorithm program file from the parameter file is one of the ways to increase the flexibility of the algorithm operation. And the intelligent analysis control platform reads the parameter file uploaded by the user, analyzes the corresponding parameter and imports the parameter file when the algorithm starts file calling.
Preferably, the industrial monitoring video analysis algorithm is distributed in the cloud in the form of micro-services and supports GPU scheduling.
Preferably, the configuration of data acquisition comprises: a certain camera or group of cameras, a collection date, a collection time point, a collection time period, and/or a collection interval.
Has the advantages that: the detailed management of the alarm strategy on the time granularity is completed through a timing task mechanism, and the alarm strategy management problem is simplified into the protocol request sending opportunity management problem. The fine-grained division of the alarm strategy in the time dimension is more in line with the management and use habits and cognition of a user on the common alarm strategy, the video analysis task performed by using server resources in a peak-off mode is also the great optimization of resource utilization, and under the condition of the same server hardware configuration, the wider and flexible task requirements are met.
Preferably, the security event intelligent analysis management and control platform uses a distributed message dump system at the bottom layer, and the distributed message dump system is used for monitoring, forwarding and storing processes of the bottom alarm information.
Has the advantages that: the message dump service adopts a distributed system architecture solution, reduces the coupling between the algorithm service cluster and the control platform, and improves the service expandability and flexibility of the control platform.
Preferably, the security event intelligent analysis management and control platform uses an open API interface of the comprehensive security management platform downwards, and is configured to obtain basic information maintained by the platform from the comprehensive security management platform, and provide a relevant API interface for other third party platforms and video annotation systems located at the same service level upwards.
Preferably, the distributed architecture of the kurbernets cloud platform is used for processing multi-channel video stream data, and for each delivery, when the load factors of POD instances in the cluster are all larger than a first threshold, a dynamic capacity expansion mechanism takes effect, api provided by the kubernets cluster is called to increase the number of the delivery instances, and a rehash process is carried out again; and when the load on a certain POD in the cluster is smaller than a second threshold value, closing the current POD instance, and re-delivering the running video analysis task to other POD instances for running, wherein the load factor of the POD is the number of cameras currently running in a single POD/the maximum number of cameras allowed to run simultaneously by the single POD, and the first threshold value is larger than the second threshold value.
Has the advantages that: the invention preferably selects the mode to realize the elastic expansion of the calculation service volume.
Generally, by the above technical solution conceived by the present invention, the following beneficial effects can be obtained:
the invention provides an industrial monitoring video oriented security event intelligent analysis and control platform which integrates a user authority management module, a data collection module, an algorithm management module, an alarm strategy management module and an alarm information management module, supports flexible configuration of a pull camera and acquisition time of video data, supports matching and binding of the pull camera and an algorithm, supports uploading, deployment and management of different versions of a filtering algorithm and a pre-trained industrial monitoring video analysis algorithm, supports customization of an alarm strategy, guarantees the security of an industrial monitoring video, and improves the flexibility of algorithm management.
Drawings
FIG. 1 is a general architecture diagram of an intelligent analysis management and control platform for industrial surveillance video-oriented security events provided by the present invention;
FIG. 2 is a flowchart of a file listening service routine provided by the present invention;
FIG. 3 is an overview of algorithm management by the algorithm management module provided by the present invention;
FIG. 4 is a flow chart of the alarm policy management module timing task execution provided by the present invention;
FIG. 5 is a data collection flow diagram of a data collection module provided by the present invention;
FIG. 6 is a diagram illustrating different types of user permissions of the user management module according to the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In addition, the technical features involved in the embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
As shown in fig. 1, the present invention provides an intelligent analysis management and control platform for security events of industrial surveillance videos, including: the system comprises a distributed message dump system, an algorithm management module, an alarm strategy management module, an alarm information management module, a data collection and collection module, a third-party platform docking module, a user management module and a human-computer interface.
And the distributed message dump system is used for monitoring, forwarding and storing the bottom early warning information. And the algorithm management module is used for the classified management of the intelligent analysis tasks and the uploading and management of the versions of the intelligent analysis algorithms in the analysis tasks. And the alarm strategy management module is used for configuring and managing an alarm strategy and realizing matching and binding of original video streams of different pulled monitoring equipment and an intelligent analysis algorithm. And the alarm information management module is used for inquiring and managing alarm information. And the data collection and collection module is used for filtering and screening the video data in the process of pulling the selected original video stream of the monitoring equipment and providing a data downloading function. And the third-party platform docking module is used for providing monitoring equipment level information, camera inspection information, original video stream address information and an alarm information retrieval function of the third-party platform docking platform for the alarm strategy management module. And the user management module is used for the authority hierarchical management of the platform login role based on the specific module function. And the human-computer interface is used for inputting parameter configuration and outputting alarm information.
Specifically, the distributed message dump system comprises: the system comprises a file monitoring service sub-module, an intermediate message storage sub-module, a message persistence sub-module and a platform local file storage system. The file monitoring service is used for monitoring and analyzing the algorithm analysis result output of the algorithm under the specified file directory, submitting the algorithm analysis result to the intermediate message storage module aiming at the early warning information through analysis and judgment as shown in fig. 2, and storing the algorithm analysis result to the platform local file system aiming at the video and other structured data in a file synchronization mode. Further, the message intermediate storage submodule is used for temporarily storing and forwarding messages, and the message middleware technology is used for ensuring the orderliness, the reliability and the safety of message transmission. And carrying out persistence operation on the early warning message in the intermediate message storage submodule after the early warning message is received, and storing the early warning message in the intermediate message storage submodule to a database.
Specifically, the algorithm management module is used for the classification management of the intelligent analysis tasks and the version uploading and management of the intelligent analysis algorithms used by the analysis tasks. As shown in fig. 3, the content of algorithm management mainly includes analysis task management and version management under the corresponding task.
The algorithm management module divides the algorithm uploading into three modules of mirror image uploading, algorithm file uploading and parameter file uploading, and the main steps of the algorithm management module comprise:
a1: the reading platform technology is used for docking a document, and defining an input/output format and a parameter configuration file format of an intelligent analysis algorithm to be uploaded;
a2: respectively uploading a mirror image, an algorithm file and a parameter configuration file;
a3: selecting a basic mirror image required by the operation of a current uploading algorithm, wherein the selected mirror image can be an existing mirror image of a platform or a mirror image uploaded by a user independently;
a4: selecting an algorithm file needed by the current task after the platform has the algorithm program file compression package or has uploaded the algorithm program file compression package;
a5: after the platform has the parameter configuration file or the algorithm configuration file is uploaded, selecting the parameter configuration file required to be used by the current task, and manually modifying and storing the parameter configuration file;
a6: and checking the operation correctness of the selected mirror image, the algorithm file, the parameter configuration file and the intelligent analysis algorithm according to the platform technology docking document specification.
The uploading abstraction of the algorithm is three parts: an operating environment, a program entity, and a parameter file. The operation environment is the environment support of the program operation, the program entity is the logic realization of the algorithm, and the parameter file influences the branch of the program operation and the output of the result. In particular, the uploaded analysis algorithm allows for operation in a unified package into a binary file. Aiming at the problem that part of analysis algorithms cannot be packed into binary files to be uploaded and run, a mode of using a pre-configured basic mirror image container is provided to solve the problem of environment dependence. Aiming at the problem of program entity loading, a mode of automatically importing the program entity file into a basic mirror image container and packaging the program entity file into an algorithm service mirror image during algorithm uploading is provided to solve the problem. Aiming at the problem of complex parameter file configuration, in the running process of a program entity, parameters are sent to an algorithm analysis service instance from a platform in the form of a network packet.
The analysis algorithm is specified as an algorithm executable file with unified input and output and a matched executable script. Specifically, the input file is a Shell start script which is allowed to run in the Docker container, and the first three parameters in the script are uniformly specified by the platform and are respectively: the camera ID, the RTSP video stream address and the algorithm output result directory, and subsequent parameters are specified according to the sequence in the parameter configuration file. The output file is a JSON format file, and the parameter items which must be contained are as follows: event start time, event end time, camera ID, event early warning video clip file name.
And (3) adopting a Docker Registry and a Kubernets cluster to complete the processes of automatic deployment, operation and maintenance management, service discovery, load balancing and elastic expansion of the algorithm analysis service. The algorithm deployment adopts a Kubernetes cluster deployment mode to provide functions of resource scheduling, deployment and operation, service discovery, elastic expansion and the like for containerization application, and algorithm analysis services in the system are all hosted in Kubernetes in a container mode. Specifically, the storage of the algorithm analysis mirror image is completed by using a Docker privatized mirror image warehouse Registry, and the method supports that any node of the Kubernets cluster pulls the Docker Registry mirror image for localization deployment.
And the alarm strategy management module is used for configuring and managing an alarm strategy and realizing matching and binding of original video streams of different pulled monitoring equipment and an intelligent analysis algorithm. The alarm strategy management module comprises a strategy configuration submodule and a strategy management submodule. The strategy management submodule is used for managing and maintaining the alarm strategy. The alarm strategy uploading module is used for configuring alarm strategy timing tasks and pushing settings, and mainly comprises the following steps:
b1: selecting a camera: acquiring and maintaining monitoring equipment hierarchy information, routing inspection information, a video preview streaming address based on an RTSP (real time streaming protocol) and the like by the platform based on a third party calling API (application programming interface), managing a hierarchy list by using a shuttle frame component and providing a camera selecting function;
b2: selecting an algorithm and configuring parameters: providing various online intelligent analysis algorithm selection functions in online analysis tasks, and supporting manual algorithm modification and uploading of configuration parameter files;
b3: date and time period of selection: there are three options for the date, including: without limitation of date, working day, custom date. There are two options for the period, including: unlimited time period and self-defined time period;
b4: and (3) push setting: the pushing setting is connected with the definition of the alarm information and is used for setting an alarm information pushing platform, pushing contents, time intervals of alarm grade division and pushing time intervals of alarms in different grades.
Specifically, the data source of the algorithm analysis is a real-time video stream of a Haikang, Dahua and other high-definition network monitoring camera managed and maintained by a bottom management platform obtained based on API call provided by a third-party docking module. The alarm strategy is bound with a group of camera IDs, and the platform calls an API interface to obtain a stream address and transmits the stream address into the algorithm analysis service. The timing task is implemented by using a Linux Inotify file monitoring mechanism, and as shown in fig. 4, the specific implementation flow includes:
s1: the crontab monitoring service is started in a daemon process mode when the platform is started;
s2: reading the appointed monitoring file every 2 seconds;
s3: judging whether a task is to be executed at the current moment; if no task is to be executed, go to S2;
s4: if the execution task is judged in the S3, executing a corresponding script sending instruction;
the injection of the WEB service is used for the response process of the alarm strategy, and the alarm strategy management is converted into the problem of API request sending opportunity management based on the Rest style API design.
The alarm information management module is used for inquiring and managing alarm information. Specifically, the alarm information management module includes: form query and alarm information management, global fuzzy search, chart query (according to days/hours/minutes) of historical alarm information, chart query of real-time alarm information, form and chart information linkage query, statistical monitoring large-screen analysis and the like.
Alarm strategy pushing setting constructs an alarm grading strategy, defines different alarm grades and the time length of event occurrence under the corresponding alarm grades for removing redundant alarm information, and defines pushing intervals under the different alarm grades for setting alarm information pushing frequency in the different alarm strategies.
And the data set acquisition module is used for filtering and screening the video data in the process of pulling the selected original video stream of the monitoring equipment. As shown in fig. 5, the main steps of the acquisition configuration include:
s1: editing and submitting the acquisition configuration information, comprising the following steps: acquisition start time, acquisition end time, acquisition camera equipment, whether to use a filtering algorithm. The acquisition start time is less than the acquisition end time;
s2: judging the time interval between the current acquisition starting time and the acquisition ending time in the step S1;
s3: in step S2, if the acquisition start time is later than the current time, a timing task delay waiting stream is set; the realization of the timing task is similar to that of the alarm strategy module, and the functions of a Crontab monitoring frame can be shared;
s4: in step S2, if the acquisition start time is earlier than the current time, it is determined whether the acquisition end time is shorter than the current time;
s5: if the acquisition ending time is less than the current time in the step S4, immediately starting to pull the stream;
s6: if the capture time is longer than the current time in step S4, the real-time stream and the video stream from the capture start time to the current time are immediately pulled simultaneously.
Specifically, the collecting and filtering submodule is used for screening a data set, and the main steps of the collecting and filtering submodule comprise:
s1: intercepting a frame in video stream data of the camera, and judging whether a human body exists in the current frame by using a human body detection algorithm;
s2: if the confidence of human body detection in the current frame is high, carrying out next detection at intervals of 100 frames and storing continuous video frames in the interval, and jumping to S1 until the acquisition ending time comes;
s3: and if the confidence of human body detection in the current frame is low, carrying out next detection at an interval of 25 frames, not storing the continuous video frames in the interval, and jumping to S1 until the acquisition ending time comes.
And the user management module is used for the authority hierarchical management of platform login roles based on the functions of the specific modules. As shown in fig. 6, three types of users are divided, including: super manager, common user. The super administrator has user authority distribution, data set collection, algorithm management, alarm strategy management and alarm information management authority. The common administrator has the authority of algorithm management, alarm strategy management and alarm information management. The common user only has the alarm information management authority.
It will be understood by those skilled in the art that the foregoing is only a preferred embodiment of the present invention, and is not intended to limit the invention, and that any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (8)
1. The utility model provides a security incident intelligent analysis management and control platform towards industry surveillance video which characterized in that includes:
the user authority management module is used for managing user information and user authority, and defining three different roles, wherein each role has corresponding authority;
the data collection and collection module is used for automatically pulling the industrial monitoring video under the monitoring network according to the configured camera and the collection time, matching and binding the pull camera with the filtering algorithm, filtering and packaging the pulled industrial monitoring video and providing a corresponding downloading interface;
the algorithm management module is used for managing versions of the filtering algorithm and the industrial monitoring video analysis algorithm, including uploading, deployment and management of the algorithm and supporting parallel operation of different versions of the algorithm;
the alarm strategy management module is used for automatically pulling the corresponding industrial monitoring video under the monitoring network according to the configured camera and the warning time, running a selected analysis algorithm for detection, and pushing the alarm strategy to support self-defining when a target event is detected;
and the alarm information management module is used for inquiring, managing and visualizing alarm information and supporting an open API (application program interface) to inquire alarm records.
2. The security event intelligent analysis management and control platform of claim 1, wherein the three roles of the management and control platform comprise:
a hypervisor, which has the following privileges: data set collection, user authority definition, algorithm uploading and management, alarm strategy configuration, alarm record inquiry and API interface butt joint with a third-party platform;
the general administrator, who has the following rights: uploading and managing an algorithm, configuring an alarm strategy and inquiring an alarm record;
the ordinary user, it possesses the following rights: and inquiring alarm records.
3. The platform of claim 1, wherein in the algorithm management module, an operating environment, a program entity and parameters of the application program are decoupled in an algorithm uploading process, wherein the operating environment is stored in a mirror warehouse in a docker mirror mode, and is locally deployed in a docker container mode during use, the program entity algorithm is placed in a server root directory in a source code or executable program mode, and is uploaded to the container and packaged as a mirror image during use, and the parameters are stored in the server root directory in a json file mode and are sent to the program entity in a network package mode during use.
4. The intelligent analysis and control platform for security events according to claim 3, wherein the industrial surveillance video analysis algorithm is distributed in the cloud by micro-services and supports GPU scheduling.
5. The security event intelligent analysis management and control platform of claim 1, wherein the configuration of data collection comprises: a certain camera or group of cameras, a collection date, a collection time point, a collection time period, and/or a collection interval.
6. The security event intelligent analysis management and control platform of claim 1, wherein the security event intelligent analysis management and control platform uses a distributed message dump system at the bottom layer, and the distributed message dump system is used for monitoring, forwarding and storing processes of the bottom alarm information.
7. The platform of claim 1, wherein the platform uses an open API of the integrated security management platform downward, and is configured to obtain basic information maintained by the platform from the integrated security management platform and provide related API interfaces for other third party platforms and video annotation systems located at the same service level upward.
8. The intelligent security event analysis and control platform according to claim 1, wherein a distributed architecture of the kurbernets cloud platform is used for processing multi-channel video stream data, and for each Deployment, when load factors of POD instances in a cluster are all larger than a first threshold, a dynamic capacity expansion mechanism takes effect, api provided by the kuberenetes cluster is called to increase the number of Deployment instances, and a rehash process is performed again; and when the load on a certain POD in the cluster is smaller than a second threshold value, closing the current POD instance, and re-delivering the running video analysis task to other POD instances for running, wherein the load factor of the POD is the number of cameras currently running in a single POD/the maximum number of cameras allowed to run simultaneously by the single POD, and the first threshold value is larger than the second threshold value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110874184.4A CN113612970A (en) | 2021-07-30 | 2021-07-30 | Safety event intelligent analysis management and control platform for industrial monitoring video |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110874184.4A CN113612970A (en) | 2021-07-30 | 2021-07-30 | Safety event intelligent analysis management and control platform for industrial monitoring video |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113612970A true CN113612970A (en) | 2021-11-05 |
Family
ID=78338825
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110874184.4A Pending CN113612970A (en) | 2021-07-30 | 2021-07-30 | Safety event intelligent analysis management and control platform for industrial monitoring video |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113612970A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114048103A (en) * | 2021-11-19 | 2022-02-15 | 重庆富民银行股份有限公司 | PMML wind control model management platform |
| CN114285970A (en) * | 2021-12-22 | 2022-04-05 | 南京国电南自轨道交通工程有限公司 | Intelligent station operation auxiliary linkage method based on video analysis technology |
| CN115243010A (en) * | 2022-07-15 | 2022-10-25 | 浪潮通信信息系统有限公司 | Bright kitchen scene intelligent detection system and device |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080106597A1 (en) * | 1999-10-12 | 2008-05-08 | Vigilos, Inc. | System and method for storing and remotely retrieving surveillance video images |
| CN101609588A (en) * | 2008-06-16 | 2009-12-23 | 云南正卓信息技术有限公司 | Full-automatic anti-intrusion intelligent video monitoring alarm system for unattended villa |
| CN103108159A (en) * | 2013-01-17 | 2013-05-15 | 新疆电力公司乌鲁木齐电业局 | Electric power intelligent video analyzing and monitoring system and method |
| US20140047107A1 (en) * | 2012-08-09 | 2014-02-13 | Rockwell Automation Technologies, Inc. | Remote industrial monitoring and analytics using a cloud infrastructure |
| US20160080703A1 (en) * | 2013-05-28 | 2016-03-17 | Huawei Technologies Co., Ltd. | Video Surveillance Method, Apparatus, and System |
| CN105450987A (en) * | 2015-11-12 | 2016-03-30 | 北京弘恒科技有限公司 | Intelligent recognition platform video monitoring early warning system |
| CN108777693A (en) * | 2018-06-29 | 2018-11-09 | 华东师范大学 | A kind of equipment safety control based on FPGA families monitor video |
| CN109509190A (en) * | 2018-12-19 | 2019-03-22 | 中国科学院重庆绿色智能技术研究院 | Video monitoring image screening technique, device, system and storage medium |
| CN112052804A (en) * | 2020-09-10 | 2020-12-08 | 公安部第三研究所 | Video intelligent analysis and alarm system and method for realizing safety management |
| US20210096911A1 (en) * | 2020-08-17 | 2021-04-01 | Essence Information Technology Co., Ltd | Fine granularity real-time supervision system based on edge computing |
-
2021
- 2021-07-30 CN CN202110874184.4A patent/CN113612970A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080106597A1 (en) * | 1999-10-12 | 2008-05-08 | Vigilos, Inc. | System and method for storing and remotely retrieving surveillance video images |
| CN101609588A (en) * | 2008-06-16 | 2009-12-23 | 云南正卓信息技术有限公司 | Full-automatic anti-intrusion intelligent video monitoring alarm system for unattended villa |
| US20140047107A1 (en) * | 2012-08-09 | 2014-02-13 | Rockwell Automation Technologies, Inc. | Remote industrial monitoring and analytics using a cloud infrastructure |
| CN103108159A (en) * | 2013-01-17 | 2013-05-15 | 新疆电力公司乌鲁木齐电业局 | Electric power intelligent video analyzing and monitoring system and method |
| US20160080703A1 (en) * | 2013-05-28 | 2016-03-17 | Huawei Technologies Co., Ltd. | Video Surveillance Method, Apparatus, and System |
| CN105450987A (en) * | 2015-11-12 | 2016-03-30 | 北京弘恒科技有限公司 | Intelligent recognition platform video monitoring early warning system |
| CN108777693A (en) * | 2018-06-29 | 2018-11-09 | 华东师范大学 | A kind of equipment safety control based on FPGA families monitor video |
| CN109509190A (en) * | 2018-12-19 | 2019-03-22 | 中国科学院重庆绿色智能技术研究院 | Video monitoring image screening technique, device, system and storage medium |
| US20210096911A1 (en) * | 2020-08-17 | 2021-04-01 | Essence Information Technology Co., Ltd | Fine granularity real-time supervision system based on edge computing |
| CN112052804A (en) * | 2020-09-10 | 2020-12-08 | 公安部第三研究所 | Video intelligent analysis and alarm system and method for realizing safety management |
Non-Patent Citations (1)
| Title |
|---|
| 王显龙;方明;: "一种智能视频监控系统架构方案设计", 长春理工大学学报(自然科学版), no. 01 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114048103A (en) * | 2021-11-19 | 2022-02-15 | 重庆富民银行股份有限公司 | PMML wind control model management platform |
| CN114285970A (en) * | 2021-12-22 | 2022-04-05 | 南京国电南自轨道交通工程有限公司 | Intelligent station operation auxiliary linkage method based on video analysis technology |
| CN115243010A (en) * | 2022-07-15 | 2022-10-25 | 浪潮通信信息系统有限公司 | Bright kitchen scene intelligent detection system and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113612970A (en) | Safety event intelligent analysis management and control platform for industrial monitoring video | |
| WO2020207266A1 (en) | Network system, instance management method, device, and storage medium | |
| WO2020207265A1 (en) | Network system, management and control method and device, and storage medium | |
| CN113342478B (en) | Resource management method, device, network system and storage medium | |
| WO2017050141A1 (en) | Distributed storage-based file delivery system and method | |
| CN101707399B (en) | Method and system for acquiring electric energy information | |
| CN112269640B (en) | Method for realizing life cycle management of container cloud component | |
| EP1361761A1 (en) | Telecommunications network management system and method for service monitoring | |
| CN114830080B (en) | Data distribution flow configuration method and device, electronic equipment and storage medium | |
| CN111752723B (en) | Visual multi-source service management system and implementation method thereof | |
| CN115858108A (en) | Cloud edge coordination system constructed based on Kubeedge edge computing framework | |
| CN102143182A (en) | Internet-based digital television multimedia information making, examination and publishing system | |
| CN116048467A (en) | Micro-service development platform and business system development method | |
| CN116489214A (en) | Unified service scheduling method, device, medium and equipment based on micro service gateway | |
| CN110971439A (en) | Policy decision method and device, system, storage medium, policy decision unit and cluster | |
| CN112199178A (en) | Cloud service dynamic scheduling method and system based on lightweight container | |
| CN114666576A (en) | Public safety edge gateway system | |
| CN111048164A (en) | Medical big data long-term storage system | |
| CN113573150A (en) | Video stream processing method and device, electronic equipment and storage medium | |
| CN112035466B (en) | External index development framework for block chain query | |
| CN115757552B (en) | Bank historical data management system based on distributed microservice | |
| CN103379155A (en) | Method and system for distributing contents and content management system | |
| Garcia et al. | NUBOMEDIA: an elastic PaaS enabling the convergence of real-time and big data multimedia | |
| CN112422349B (en) | Network management system, method, equipment and medium for NFV | |
| CN112203148B (en) | Method for managing Service of Android set top box system built-in system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |