CN116881482A - Cross-media intelligent sensing and analyzing processing method for public safety data - Google Patents
Cross-media intelligent sensing and analyzing processing method for public safety data Download PDFInfo
- Publication number
- CN116881482A CN116881482A CN202310767292.0A CN202310767292A CN116881482A CN 116881482 A CN116881482 A CN 116881482A CN 202310767292 A CN202310767292 A CN 202310767292A CN 116881482 A CN116881482 A CN 116881482A
- Authority
- CN
- China
- Prior art keywords
- data
- cross
- intelligent
- media
- modal
- 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
- 231100000279 safety data Toxicity 0.000 title claims abstract description 22
- 238000003672 processing method Methods 0.000 title claims abstract description 11
- 238000000034 method Methods 0.000 claims abstract description 37
- 230000008569 process Effects 0.000 claims abstract description 21
- 230000007246 mechanism Effects 0.000 claims abstract description 16
- 238000012545 processing Methods 0.000 claims abstract description 14
- 238000012512 characterization method Methods 0.000 claims abstract description 11
- 230000008878 coupling Effects 0.000 claims abstract description 11
- 238000010168 coupling process Methods 0.000 claims abstract description 11
- 238000005859 coupling reaction Methods 0.000 claims abstract description 11
- 238000013508 migration Methods 0.000 claims abstract description 11
- 230000005012 migration Effects 0.000 claims abstract description 11
- 238000012098 association analyses Methods 0.000 claims abstract description 7
- 238000004891 communication Methods 0.000 claims abstract description 6
- 238000004458 analytical method Methods 0.000 claims description 20
- 238000000605 extraction Methods 0.000 claims description 12
- 230000004927 fusion Effects 0.000 claims description 12
- 238000005516 engineering process Methods 0.000 claims description 10
- 238000013528 artificial neural network Methods 0.000 claims description 8
- 238000001514 detection method Methods 0.000 claims description 8
- 238000012544 monitoring process Methods 0.000 claims description 8
- 238000010586 diagram Methods 0.000 claims description 7
- 230000006399 behavior Effects 0.000 claims description 6
- 238000010801 machine learning Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 5
- 230000002776 aggregation Effects 0.000 claims description 4
- 238000004220 aggregation Methods 0.000 claims description 4
- 230000005540 biological transmission Effects 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 4
- 125000004122 cyclic group Chemical group 0.000 claims description 4
- 238000013523 data management Methods 0.000 claims description 4
- 238000013500 data storage Methods 0.000 claims description 4
- 238000001784 detoxification Methods 0.000 claims description 4
- 230000005764 inhibitory process Effects 0.000 claims description 4
- 210000001503 joint Anatomy 0.000 claims description 4
- 230000007787 long-term memory Effects 0.000 claims description 4
- 238000013507 mapping Methods 0.000 claims description 4
- 238000005259 measurement Methods 0.000 claims description 4
- 230000006855 networking Effects 0.000 claims description 4
- 230000008520 organization Effects 0.000 claims description 4
- 239000002574 poison Substances 0.000 claims description 4
- 231100000614 poison Toxicity 0.000 claims description 4
- 238000011160 research Methods 0.000 claims description 4
- 238000012827 research and development Methods 0.000 claims description 4
- 238000012502 risk assessment Methods 0.000 claims description 4
- 230000000007 visual effect Effects 0.000 claims description 4
- 230000002159 abnormal effect Effects 0.000 abstract description 3
- 230000010365 information processing Effects 0.000 abstract description 3
- 230000035515 penetration Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/40—Information retrieval; Database structures therefor; File system structures therefor of multimedia data, e.g. slideshows comprising image and additional audio data
- G06F16/43—Querying
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/40—Information retrieval; Database structures therefor; File system structures therefor of multimedia data, e.g. slideshows comprising image and additional audio data
- G06F16/48—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/483—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Computational Linguistics (AREA)
- Multimedia (AREA)
- Library & Information Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a cross-media intelligent sensing and analyzing processing method of public safety data, belongs to the technical field of public safety data, and aims to provide the cross-media intelligent sensing and analyzing processing method of public safety data, improve the accuracy of cross-media sensing and analyzing, realize convergence and layering treatment of data, enable deep application of public safety subdivision scenes, and solve the problems of unclear field sensing, abnormal early warning lag, insufficient joint linkage, lack of specialized tools and the like in the digital actual combat process of the public safety industry. Based on theories in the aspects of unified characterization, association analysis, knowledge migration and the like of cross-media data, the intelligent processing system and the method are combined with the fields of intelligent public security, intelligent traffic and the like to subdivide scenes, explore a collaboration mechanism among multiple modal data, break through the limitation of single-media information processing, and improve the intelligent processing capacity of cross-media communication, coupling and collaboration. The invention is suitable for a cross-media intelligent sensing and analyzing processing method of public safety data.
Description
Technical Field
The invention belongs to the technical field of public safety data, and particularly relates to a cross-media intelligent sensing and analyzing processing method of public safety data.
Background
Public safety is an important sign showing human social order, and occurrence of public safety events forms a serious challenge for government reputation and social management systems. However, public safety data come from different data sources and are expressed as different media types, public safety field data have the characteristics and challenges of large data scale, multiple mode types, few labeled samples, high data feature dimension, cross-mode distribution heterogeneity and the like, and public safety authorities should deal with debilitation when processing security events which are changeable under new situations, so that new data processing technical means are urgently needed to serve actual combat applications.
Therefore, how to solve the problems of 'data modal isomerism, feature distribution difference', 'semantic inconsistency', ambiguous association mechanism ', various data sources, changeable application scene' and the like in the public safety field data is a key for solving the problems of unclear field perception, abnormal early warning lag, insufficient joint linkage, lack of professional tools and the like in the digital actual combat process of the public safety industry.
Disclosure of Invention
The invention aims at: the cross-media intelligent sensing and analysis processing method for public safety data is provided, so that the accuracy rate of cross-media sensing and analysis is improved, convergence and layering treatment of the data are realized, the deep application of public safety subdivision scenes is energized, and the problems of unclear field sensing, abnormal early warning lag, insufficient joint linkage, lack of specialized tools and the like in the digital actual combat process of the public safety industry are solved.
The technical scheme adopted by the invention is as follows:
a cross-media intelligent sensing and analyzing processing method for public safety data comprises the following steps:
(1) Based on public safety data from different data sources, representing different media types, constructing an efficient data joint expression, similarity measurement and query algorithm, and based on a user-defined abstract semantic description and a search method, realizing fuzzy search by text description under the condition that an image or video sample of a target cannot be obtained; based on a cross-media hash retrieval model, unified joint expression and hash retrieval of multi-modal data of voice, video, image and text are realized, the unified expression is achieved to a common Hamming space, and the large-scale multi-modal data rapid query and retrieval facing to practical application is supported;
(2) After entity, relation and attribute knowledge elements are extracted from the original unstructured data through an information extraction map construction technology, the learned knowledge map is evaluated and filtered through knowledge reasoning, entity alignment and concept extraction subtasks, modeling of multi-modal entities and cross-modal relation is completed based on a heterogeneous graph neural network of the multi-modal knowledge map, and alignment of cross-modal information among the entities is completed;
(3) The uniform characterization of multi-modal coupling is realized by utilizing the semantic consistency among the multi-modal information so as to learn more comprehensive characteristic representation, and the cross-modal difference among the multi-modal data is reduced by improving the characteristic extraction and the public space mapping, so that the cross-modal retrieval is more accurate;
(4) Based on cross-media semantic consistency and diversity, visual language characteristic self-adaptive association and cross-modal semantic and relation fusion, a long-term memory model, a multi-modal cyclic neural network, a self-adaptive attention mechanism, a cross-modal aggregation network and a hierarchical association method model are provided, the cross-media semantic gap is reduced, and accurate understanding and efficient association of cross-media content are realized.
(5) The existing source domain knowledge is fully utilized through cross-domain knowledge migration, knowledge of target domain data is learned in an auxiliary mode, and a tie between the source domain and the target domain is established;
(6) Based on cross-media data unified characterization, association analysis and knowledge migration theory, combining public security, traffic and discipline detection subdivision scenes to construct intelligent public security, intelligent poison inhibition, intelligent traffic and intelligent discipline detection scene depth application; the collaborative mechanism among the multi-modal data is relied on, the public safety subdivision scene is energized based on cross-media retrieval, map link prediction based on a multi-modal knowledge base, analysis object behavior recognition and domain generalization pedestrian re-recognition algorithm model, iterative evolution is realized through algorithm training, actual combat application capability of a cross-media intelligent algorithm and a machine learning model is continuously improved, and intelligent processing capability of cross-media communication, coupling and collaboration is improved.
Further, in the step (6), the process of the intelligent public security scene depth application specifically includes: the analysis channel from low-value original video image data to valuable structured feature attribute data is provided, a standardized video image information data management mechanism is formed, video image information is created to be an important basic resource for supporting various public security works under dynamic and informationized conditions, and the video image information is an important support of a three-dimensional social security and protection system.
Further, in the step (6), the process of the intelligent poison-restricted scene depth application specifically includes:
based on multi-source heterogeneous big data fusion research and development, a data center is established to eliminate information island, a structured and unstructured multi-source big data center is established, data storage, transmission and calculation efficiency is improved, various data related to detoxification work are summarized and fused into different entities of human, matter, ground, things, organization and virtual identity, and a relation network is formed by the relation among the reconstructed data according to attribute relation, space-time relation, semantic relation, characteristic relation, position relation and the like.
Further, in the step (6), the process of the intelligent traffic scene depth application specifically includes:
automatically acquiring information and parameters of a vehicle number plate, a vehicle type and a driving direction, networking, comparing and alarming with a blacklist database, and automatically monitoring the driving route of the blacklist vehicle; the road traffic control system automatically acquires parameters of road junctions, road section traffic flows, saturation and occupancy, provides real-time traffic data for a traffic signal control system, participates in green signal ratio adjustment of a lamp-controlled road junction, provides real-time traffic data for a traffic intelligent guidance system, participates in regional traffic guidance, pushes text information or graphic information to intelligent terminals such as mobile phones or police service terminals, realizes dispatching and commanding of road police, organically combines road monitoring, public security bayonets and traffic parameter acquisition, and provides analysis and research of real-time video images and high-definition video recording data.
Further, in the step (6), the process of the intelligent discipline scene depth application specifically includes: analyzing large batch of data which are in butt joint with a plurality of units of public security, government affairs, banks and securities based on a large data technology, performing data arrangement and fusion on complex data and complex networks, extracting multidimensional data characteristics based on rules to form a relational network, and opening up a data island; and through analysis of the multipoint shared information and marking of the strong contact diagram, the risk analysis based on the mode is realized.
In summary, due to the adoption of the technical scheme, the beneficial effects of the invention are as follows:
1. based on theories in the aspects of unified characterization, association analysis, knowledge migration and the like of cross-media data, the invention further combines with the fields of intelligent public security, intelligent traffic and the like to subdivide scenes, explores a collaboration mechanism among a plurality of modal data, and breaks through the limitation of single media information processing; the method is characterized in that algorithm model indexes such as cross-media retrieval, map link prediction based on a multi-mode knowledge base, analysis object behavior recognition, domain generalized pedestrian re-recognition and the like are developed, verified and optimized in combination with actual services, energy is provided for public safety subdivision scenes, iterative evolution is realized through training of a new algorithm, actual combat application capacity of key technologies such as a cross-media intelligent algorithm and a machine learning model is continuously improved in project practice, and intelligent processing capacity of cross-media penetration, coupling and collaboration is improved.
Drawings
For a clearer description of the technical solutions of embodiments of the present invention, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present invention and should not be considered limiting in scope, and other related drawings can be obtained according to these drawings without inventive effort for a person skilled in the art, wherein:
FIG. 1 is a diagram of a cross-media hash retrieval model of the present invention;
FIG. 2 is a diagram of a multi-modal knowledge graph construction and learning architecture of the present invention;
fig. 3 is a schematic diagram of the application of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
It should be noted that: reference numerals and letters denote similar items throughout the following figures, and thus once an item is defined in one figure, it need not be further defined and explained in subsequent figures.
In the description of the present invention, it should be noted that, directions or positional relationships indicated by terms such as "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., are directions or positional relationships based on those shown in the drawings, or those that are conventionally put in use of the inventive product, are merely for convenience of description of the present invention, and are not indicative or implying that the apparatus or element to be referred to must have a specific direction, be constructed and operated in a specific direction, and thus should not be construed as limiting the present invention. Furthermore, the terms "first," "second," "third," and the like are used merely to distinguish between descriptions and should not be construed as indicating or implying relative importance.
Furthermore, the terms "horizontal," "vertical," and the like do not denote a requirement that the component be absolutely horizontal or overhang, but rather may be slightly inclined. As "horizontal" merely means that its direction is more horizontal than "vertical", and does not mean that the structure must be perfectly horizontal, but may be slightly inclined.
In the description of the present invention, it should also be noted that, unless explicitly specified and limited otherwise, the terms "disposed," "mounted," "connected," and "connected" are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally connected; the mechanical connection can be made or the electrical connection can be made; can be directly connected or indirectly connected through an intermediate medium, and can be the communication between the two original parts. The specific meaning of the above terms in the present invention will be understood in specific cases by those of ordinary skill in the art.
A cross-media intelligent sensing and analyzing processing method for public safety data comprises the following steps:
(1) Based on public safety data from different data sources, representing different media types, constructing an efficient data joint expression, similarity measurement and query algorithm, and based on a user-defined abstract semantic description and a search method, realizing fuzzy search by text description under the condition that an image or video sample of a target cannot be obtained; based on a cross-media hash retrieval model, unified joint expression and hash retrieval of multi-modal data of voice, video, image and text are realized, the unified expression is achieved to a common Hamming space, and the large-scale multi-modal data rapid query and retrieval facing to practical application is supported;
(2) After entity, relation and attribute knowledge elements are extracted from the original unstructured data through an information extraction map construction technology, the learned knowledge map is evaluated and filtered through knowledge reasoning, entity alignment and concept extraction subtasks, modeling of multi-modal entities and cross-modal relation is completed based on a heterogeneous graph neural network of the multi-modal knowledge map, and alignment of cross-modal information among the entities is completed;
(3) The uniform characterization of multi-modal coupling is realized by utilizing the semantic consistency among the multi-modal information so as to learn more comprehensive characteristic representation, and the cross-modal difference among the multi-modal data is reduced by improving the characteristic extraction and the public space mapping, so that the cross-modal retrieval is more accurate;
(4) Based on cross-media semantic consistency and diversity, visual language characteristic self-adaptive association and cross-modal semantic and relation fusion, a long-term memory model, a multi-modal cyclic neural network, a self-adaptive attention mechanism, a cross-modal aggregation network and a hierarchical association method model are provided, the cross-media semantic gap is reduced, and accurate understanding and efficient association of cross-media content are realized.
(5) The existing source domain knowledge is fully utilized through cross-domain knowledge migration, knowledge of target domain data is learned in an auxiliary mode, and a tie between the source domain and the target domain is established;
(6) Based on cross-media data unified characterization, association analysis and knowledge migration theory, combining public security, traffic and discipline detection subdivision scenes to construct intelligent public security, intelligent poison inhibition, intelligent traffic and intelligent discipline detection scene depth application; the collaborative mechanism among the multi-modal data is relied on, the public safety subdivision scene is energized based on cross-media retrieval, map link prediction based on a multi-modal knowledge base, analysis object behavior recognition and domain generalization pedestrian re-recognition algorithm model, iterative evolution is realized through algorithm training, actual combat application capability of a cross-media intelligent algorithm and a machine learning model is continuously improved, and intelligent processing capability of cross-media communication, coupling and collaboration is improved.
Further, in the step (6), the process of the intelligent public security scene depth application specifically includes: the analysis channel from low-value original video image data to valuable structured feature attribute data is provided, a standardized video image information data management mechanism is formed, video image information is created to be an important basic resource for supporting various public security works under dynamic and informationized conditions, and the video image information is an important support of a three-dimensional social security and protection system.
Further, in the step (6), the process of the intelligent poison-restricted scene depth application specifically includes:
based on multi-source heterogeneous big data fusion research and development, a data center is established to eliminate information island, a structured and unstructured multi-source big data center is established, data storage, transmission and calculation efficiency is improved, various data related to detoxification work are summarized and fused into different entities of human, matter, ground, things, organization and virtual identity, and a relation network is formed by the relation among the reconstructed data according to attribute relation, space-time relation, semantic relation, characteristic relation, position relation and the like.
Further, in the step (6), the process of the intelligent traffic scene depth application specifically includes:
automatically acquiring information and parameters of a vehicle number plate, a vehicle type and a driving direction, networking, comparing and alarming with a blacklist database, and automatically monitoring the driving route of the blacklist vehicle; the road traffic control system automatically acquires parameters of road junctions, road section traffic flows, saturation and occupancy, provides real-time traffic data for a traffic signal control system, participates in green signal ratio adjustment of a lamp-controlled road junction, provides real-time traffic data for a traffic intelligent guidance system, participates in regional traffic guidance, pushes text information or graphic information to intelligent terminals such as mobile phones or police service terminals, realizes dispatching and commanding of road police, organically combines road monitoring, public security bayonets and traffic parameter acquisition, and provides analysis and research of real-time video images and high-definition video recording data.
Further, in the step (6), the process of the intelligent discipline scene depth application specifically includes: analyzing large batch of data which are in butt joint with a plurality of units of public security, government affairs, banks and securities based on a large data technology, performing data arrangement and fusion on complex data and complex networks, extracting multidimensional data characteristics based on rules to form a relational network, and opening up a data island; and through analysis of the multipoint shared information and marking of the strong contact diagram, the risk analysis based on the mode is realized.
In the implementation process, based on theories in the aspects of unified characterization, association analysis, knowledge migration and the like of cross-media data, the method further combines with the field subdivision scenes of intelligent public security, intelligent traffic and the like to explore the collaboration mechanism among the multi-mode data, and breaks through the limitation of single media information processing; the method is characterized in that algorithm model indexes such as cross-media retrieval, map link prediction based on a multi-mode knowledge base, analysis object behavior recognition, domain generalized pedestrian re-recognition and the like are developed, verified and optimized in combination with actual services, energy is provided for public safety subdivision scenes, iterative evolution is realized through training of a new algorithm, actual combat application capacity of key technologies such as a cross-media intelligent algorithm and a machine learning model is continuously improved in project practice, and intelligent processing capacity of cross-media penetration, coupling and collaboration is improved.
Example 1
A cross-media intelligent sensing and analyzing processing method for public safety data comprises the following steps:
(1) Based on public safety data from different data sources, representing different media types, constructing an efficient data joint expression, similarity measurement and query algorithm, and based on a user-defined abstract semantic description and a search method, realizing fuzzy search by text description under the condition that an image or video sample of a target cannot be obtained; based on a cross-media hash retrieval model, unified joint expression and hash retrieval of multi-modal data of voice, video, image and text are realized, the unified expression is achieved to a common Hamming space, and the large-scale multi-modal data rapid query and retrieval facing to practical application is supported;
(2) After entity, relation and attribute knowledge elements are extracted from the original unstructured data through an information extraction map construction technology, the learned knowledge map is evaluated and filtered through knowledge reasoning, entity alignment and concept extraction subtasks, modeling of multi-modal entities and cross-modal relation is completed based on a heterogeneous graph neural network of the multi-modal knowledge map, and alignment of cross-modal information among the entities is completed;
(3) The uniform characterization of multi-modal coupling is realized by utilizing the semantic consistency among the multi-modal information so as to learn more comprehensive characteristic representation, and the cross-modal difference among the multi-modal data is reduced by improving the characteristic extraction and the public space mapping, so that the cross-modal retrieval is more accurate;
(4) Based on cross-media semantic consistency and diversity, visual language characteristic self-adaptive association and cross-modal semantic and relation fusion, a long-term memory model, a multi-modal cyclic neural network, a self-adaptive attention mechanism, a cross-modal aggregation network and a hierarchical association method model are provided, the cross-media semantic gap is reduced, and accurate understanding and efficient association of cross-media content are realized.
(5) The existing source domain knowledge is fully utilized through cross-domain knowledge migration, knowledge of target domain data is learned in an auxiliary mode, and a tie between the source domain and the target domain is established;
(6) Based on cross-media data unified characterization, association analysis and knowledge migration theory, combining public security, traffic and discipline detection subdivision scenes to construct intelligent public security, intelligent poison inhibition, intelligent traffic and intelligent discipline detection scene depth application; the collaborative mechanism among the multi-modal data is relied on, the public safety subdivision scene is energized based on cross-media retrieval, map link prediction based on a multi-modal knowledge base, analysis object behavior recognition and domain generalization pedestrian re-recognition algorithm model, iterative evolution is realized through algorithm training, actual combat application capability of a cross-media intelligent algorithm and a machine learning model is continuously improved, and intelligent processing capability of cross-media communication, coupling and collaboration is improved.
Example 2
Based on embodiment 1, in the step (6), the process of the intelligent public security scene depth application specifically includes: the analysis channel from low-value original video image data to valuable structured feature attribute data is provided, a standardized video image information data management mechanism is formed, video image information is created to be an important basic resource for supporting various public security works under dynamic and informationized conditions, and the video image information is an important support of a three-dimensional social security and protection system.
Example 3
Based on the above embodiment, in the step (6), the process of applying the intelligent poison-restricted scene depth specifically includes:
based on multi-source heterogeneous big data fusion research and development, a data center is established to eliminate information island, a structured and unstructured multi-source big data center is established, data storage, transmission and calculation efficiency is improved, various data related to detoxification work are summarized and fused into different entities of human, matter, ground, things, organization and virtual identity, and a relation network is formed by the relation among the reconstructed data according to attribute relation, space-time relation, semantic relation, characteristic relation, position relation and the like.
Example 4
Based on the above embodiment, in the step (6), the process of the intelligent traffic scene depth application specifically includes:
automatically acquiring information and parameters of a vehicle number plate, a vehicle type and a driving direction, networking, comparing and alarming with a blacklist database, and automatically monitoring the driving route of the blacklist vehicle; the road traffic control system automatically acquires parameters of road junctions, road section traffic flows, saturation and occupancy, provides real-time traffic data for a traffic signal control system, participates in green signal ratio adjustment of a lamp-controlled road junction, provides real-time traffic data for a traffic intelligent guidance system, participates in regional traffic guidance, pushes text information or graphic information to intelligent terminals such as mobile phones or police service terminals, realizes dispatching and commanding of road police, organically combines road monitoring, public security bayonets and traffic parameter acquisition, and provides analysis and research of real-time video images and high-definition video recording data.
Example 5
Based on the above embodiment, in the step (6), the process of applying the intelligent discipline scene depth specifically includes: analyzing large batch of data which are in butt joint with a plurality of units of public security, government affairs, banks and securities based on a large data technology, performing data arrangement and fusion on complex data and complex networks, extracting multidimensional data characteristics based on rules to form a relational network, and opening up a data island; and through analysis of the multipoint shared information and marking of the strong contact diagram, the risk analysis based on the mode is realized.
The above-described embodiments of the present invention. The foregoing description is illustrative of various preferred embodiments of the present invention, and the preferred embodiments of the various preferred embodiments may be used in any combination and stacked on the premise of a certain preferred embodiment, where the embodiments and specific parameters in the embodiments are only for clearly describing the verification process of the present invention, and are not intended to limit the scope of the present invention, and the scope of the present invention is still subject to the claims, and all equivalent structural changes made by applying the descriptions and the drawings of the present invention are included in the scope of the present invention.
Claims (5)
1. A cross-media intelligent sensing and analyzing processing method of public safety data is characterized by comprising the following steps:
(1) Based on public safety data from different data sources, representing different media types, constructing an efficient data joint expression, similarity measurement and query algorithm, and based on a user-defined abstract semantic description and a search method, realizing fuzzy search by text description under the condition that an image or video sample of a target cannot be obtained; based on a cross-media hash retrieval model, unified joint expression and hash retrieval of multi-modal data of voice, video, image and text are realized, the unified expression is achieved to a common Hamming space, and the large-scale multi-modal data rapid query and retrieval facing to practical application is supported;
(2) After entity, relation and attribute knowledge elements are extracted from the original unstructured data through an information extraction map construction technology, the learned knowledge map is evaluated and filtered through knowledge reasoning, entity alignment and concept extraction subtasks, modeling of multi-modal entities and cross-modal relation is completed based on a heterogeneous graph neural network of the multi-modal knowledge map, and alignment of cross-modal information among the entities is completed;
(3) The uniform characterization of multi-modal coupling is realized by utilizing the semantic consistency among the multi-modal information so as to learn more comprehensive characteristic representation, and the cross-modal difference among the multi-modal data is reduced by improving the characteristic extraction and the public space mapping, so that the cross-modal retrieval is more accurate;
(4) Based on cross-media semantic consistency and diversity, visual language characteristic self-adaptive association and cross-modal semantic and relation fusion, a long-term memory model, a multi-modal cyclic neural network, a self-adaptive attention mechanism, a cross-modal aggregation network and a hierarchical association method model are provided, the cross-media semantic gap is reduced, and accurate understanding and efficient association of cross-media content are realized.
(5) The existing source domain knowledge is fully utilized through cross-domain knowledge migration, knowledge of target domain data is learned in an auxiliary mode, and a tie between the source domain and the target domain is established;
(6) Based on cross-media data unified characterization, association analysis and knowledge migration theory, combining public security, traffic and discipline detection subdivision scenes to construct intelligent public security, intelligent poison inhibition, intelligent traffic and intelligent discipline detection scene depth application; the collaborative mechanism among the multi-modal data is relied on, the public safety subdivision scene is energized based on cross-media retrieval, map link prediction based on a multi-modal knowledge base, analysis object behavior recognition and domain generalization pedestrian re-recognition algorithm model, iterative evolution is realized through algorithm training, actual combat application capability of a cross-media intelligent algorithm and a machine learning model is continuously improved, and intelligent processing capability of cross-media communication, coupling and collaboration is improved.
2. The method for cross-media intelligent sensing and analyzing and processing of public safety data according to claim 1, wherein in the step (6), the process of intelligent public security scene depth application is specifically as follows: the analysis channel from low-value original video image data to valuable structured feature attribute data is provided, a standardized video image information data management mechanism is formed, video image information is created to be an important basic resource for supporting various public security works under dynamic and informationized conditions, and the video image information is an important support of a three-dimensional social security and protection system.
3. The method for cross-media intelligent sensing and analyzing and processing of public safety data according to claim 1, wherein in the step (6), the process of intelligent poison-restricted scene depth application is specifically as follows:
based on multi-source heterogeneous big data fusion research and development, a data center is established to eliminate information island, a structured and unstructured multi-source big data center is established, data storage, transmission and calculation efficiency is improved, various data related to detoxification work are summarized and fused into different entities of human, matter, ground, things, organization and virtual identity, and a relation network is formed by the relation among the reconstructed data according to attribute relation, space-time relation, semantic relation, characteristic relation, position relation and the like.
4. The method for cross-media intelligent sensing and analyzing and processing of public safety data according to claim 1, wherein in the step (6), the process of intelligent traffic scene depth application is specifically as follows:
automatically acquiring information and parameters of a vehicle number plate, a vehicle type and a driving direction, networking, comparing and alarming with a blacklist database, and automatically monitoring the driving route of the blacklist vehicle; the road traffic control system automatically acquires parameters of road junctions, road section traffic flows, saturation and occupancy, provides real-time traffic data for a traffic signal control system, participates in green signal ratio adjustment of a lamp-controlled road junction, provides real-time traffic data for a traffic intelligent guidance system, participates in regional traffic guidance, pushes text information or graphic information to intelligent terminals such as mobile phones or police service terminals, realizes dispatching and commanding of road police, organically combines road monitoring, public security bayonets and traffic parameter acquisition, and provides analysis and research of real-time video images and high-definition video recording data.
5. The method for cross-media intelligent sensing and analyzing and processing of public safety data according to claim 1, wherein in the step (6), the process of intelligent discipline scene depth application is specifically as follows: analyzing large batch of data which are in butt joint with a plurality of units of public security, government affairs, banks and securities based on a large data technology, performing data arrangement and fusion on complex data and complex networks, extracting multidimensional data characteristics based on rules to form a relational network, and opening up a data island; and through analysis of the multipoint shared information and marking of the strong contact diagram, the risk analysis based on the mode is realized.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310767292.0A CN116881482A (en) | 2023-06-27 | 2023-06-27 | Cross-media intelligent sensing and analyzing processing method for public safety data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310767292.0A CN116881482A (en) | 2023-06-27 | 2023-06-27 | Cross-media intelligent sensing and analyzing processing method for public safety data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116881482A true CN116881482A (en) | 2023-10-13 |
Family
ID=88257804
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310767292.0A Pending CN116881482A (en) | 2023-06-27 | 2023-06-27 | Cross-media intelligent sensing and analyzing processing method for public safety data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116881482A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118245616A (en) * | 2024-05-28 | 2024-06-25 | 北京警察学院 | Multi-mode knowledge graph integration method and system |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104346440A (en) * | 2014-10-10 | 2015-02-11 | 浙江大学 | Neural-network-based cross-media Hash indexing method |
| US20190205393A1 (en) * | 2016-07-11 | 2019-07-04 | Peking University Shenzhen Graduate School | A cross-media search method |
| CN111680173A (en) * | 2020-05-31 | 2020-09-18 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | CMR model for uniformly retrieving cross-media information |
| WO2022068196A1 (en) * | 2020-09-30 | 2022-04-07 | 三维通信股份有限公司 | Cross-modal data processing method and device, storage medium, and electronic device |
| CN114360731A (en) * | 2021-12-09 | 2022-04-15 | 武汉科技大学 | Method for constructing medical care and knowledge combined service system for old people care |
| CN115376045A (en) * | 2022-08-16 | 2022-11-22 | 四川九洲视讯科技有限责任公司 | Public safety command intelligent processing method based on multi-mode fusion deep learning |
| CN115905555A (en) * | 2022-10-27 | 2023-04-04 | 国网浙江省电力有限公司宁波供电公司 | Power data analysis method based on multimode data processing and fusion |
| CN116049454A (en) * | 2022-11-01 | 2023-05-02 | 齐鲁空天信息研究院 | Intelligent searching method and system based on multi-source heterogeneous data |
| CN116307757A (en) * | 2023-01-18 | 2023-06-23 | 辽宁荣科智维云科技有限公司 | Intelligent data interaction method, interaction system, computer equipment and application |
-
2023
- 2023-06-27 CN CN202310767292.0A patent/CN116881482A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104346440A (en) * | 2014-10-10 | 2015-02-11 | 浙江大学 | Neural-network-based cross-media Hash indexing method |
| US20190205393A1 (en) * | 2016-07-11 | 2019-07-04 | Peking University Shenzhen Graduate School | A cross-media search method |
| CN111680173A (en) * | 2020-05-31 | 2020-09-18 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | CMR model for uniformly retrieving cross-media information |
| WO2022068196A1 (en) * | 2020-09-30 | 2022-04-07 | 三维通信股份有限公司 | Cross-modal data processing method and device, storage medium, and electronic device |
| CN114360731A (en) * | 2021-12-09 | 2022-04-15 | 武汉科技大学 | Method for constructing medical care and knowledge combined service system for old people care |
| CN115376045A (en) * | 2022-08-16 | 2022-11-22 | 四川九洲视讯科技有限责任公司 | Public safety command intelligent processing method based on multi-mode fusion deep learning |
| CN115905555A (en) * | 2022-10-27 | 2023-04-04 | 国网浙江省电力有限公司宁波供电公司 | Power data analysis method based on multimode data processing and fusion |
| CN116049454A (en) * | 2022-11-01 | 2023-05-02 | 齐鲁空天信息研究院 | Intelligent searching method and system based on multi-source heterogeneous data |
| CN116307757A (en) * | 2023-01-18 | 2023-06-23 | 辽宁荣科智维云科技有限公司 | Intelligent data interaction method, interaction system, computer equipment and application |
Non-Patent Citations (8)
| Title |
|---|
| CHAO LI等: "CVPR 2018 | 自监督对抗哈希SSAH：当前最佳的跨模态检索框架", pages 1 - 5, Retrieved from the Internet <URL:https://zhuanlan.zhihu.com/p/35688497> * |
| GEORGII MIKRIUKOV 等: "Unsupervised Contrastive Hashing for Cross-Modal Retrieval in Remote Sensing", ICASSP 2022 - 2022 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP), 17 April 2022 (2022-04-17), pages 4463 - 4467 * |
| JUN YU 等: "Unsupervised Multi-modal Hashing for Cross-Modal Retrieval", pages 1 - 12, Retrieved from the Internet <URL:https://link.springer.com/article/10.1007/s12559-021-09847-4> * |
| 吴飞 等: "互联网跨媒体分析与检索:理论与算法", 计算机辅助设计与图形学学报, no. 01, 15 January 2010 (2010-01-15), pages 1 - 9 * |
| 姚涛 等: "基于映射字典学习的跨模态哈希检索", 自动化学报, no. 08, 11 December 2017 (2017-12-11), pages 1475 - 1485 * |
| 小王: "跨媒体检索：概念、方法、评测与挑战(一)：基础知识", pages 1 - 3, Retrieved from the Internet <URL:https://zhuanlan.zhihu.com/p/46284265> * |
| 小马: "《MMT》-同时利用字幕、音频、视频进行检索！", pages 1 - 5, Retrieved from the Internet <URL:https://zhuanlan.zhihu.com/p/434226505> * |
| 梁美玉 等: "基于多模态图和对抗哈希注意力网络的跨媒体细粒度表示学习", 模式识别与人工智能, 15 March 2022 (2022-03-15), pages 195 - 206 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118245616A (en) * | 2024-05-28 | 2024-06-25 | 北京警察学院 | Multi-mode knowledge graph integration method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110378824A (en) | A kind of public security traffic control data brain and construction method | |
| CN108964996A (en) | Urban-rural integration information grid system and information sharing method based on it | |
| CN116881482A (en) | Cross-media intelligent sensing and analyzing processing method for public safety data | |
| CN115376045A (en) | Public safety command intelligent processing method based on multi-mode fusion deep learning | |
| CN111552813A (en) | Power knowledge graph construction method based on power grid full-service data | |
| CN113886596A (en) | Method for constructing flexible city knowledge graph based on city element and multi-disaster fusion | |
| CN114328780B (en) | Hexagonal lattice-based smart city geographic information updating method, equipment and medium | |
| Shaw et al. | GIScience for human dynamics research in a changing world. | |
| CN115238774A (en) | Social network malicious account detection method and device based on heterogeneous graph neural network | |
| Tong et al. | Multimedia network public opinion supervision prediction algorithm based on big data | |
| Sivčević et al. | Possibilities of used intelligence based agents in instant messaging on e-government services | |
| CN116450730A (en) | Knowledge graph-based personnel relationship mining method and system | |
| CN116956212A (en) | Multi-source visual information feature recognition and extraction method | |
| CN116758604A (en) | Depth forgery detection method based on facial geometry relation reasoning | |
| CN112306985A (en) | Digital retina multi-modal feature combined accurate retrieval method | |
| Feng et al. | Embedded YOLO: A Real‐Time Object Detector for Small Intelligent Trajectory Cars | |
| CN113111691B (en) | Urban land function identification method based on remote sensing picture and interest point data | |
| Jain | A Survey on Railway Assets: A Potential Domain for Big Data | |
| Wang et al. | [Retracted] Attribute Reduction Algorithm on Concept Lattice and Application in Smart City Energy Consumption Analysis | |
| Lei | Emergency information communication structure by using multimodel fusion and artificial intelligence algorithm | |
| Luo | (Retracted) Role of video sensors in observing visual image design in the construction of smart cities | |
| CN110750678A (en) | Method and system for monitoring video data association description and storage management | |
| Tian | Embedded Real‐Time Detection of Vehicles and Pedestrians Based on Minimal Networks | |
| Liu et al. | Research on intelligent campus monitoring management system based on deep neural network algorithm | |
| Liu et al. | Research on Semantic Search Algorithm of Smart Grid Knowledge Map Based on Deep Learning |
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 |