CN119671794A - A smart community intelligent security dispatch management method based on deep learning - Google Patents
A smart community intelligent security dispatch management method based on deep learning Download PDFInfo
- Publication number
- CN119671794A CN119671794A CN202411559505.1A CN202411559505A CN119671794A CN 119671794 A CN119671794 A CN 119671794A CN 202411559505 A CN202411559505 A CN 202411559505A CN 119671794 A CN119671794 A CN 119671794A
- Authority
- CN
- China
- Prior art keywords
- data
- community
- vehicles
- personnel
- people
- 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
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of intelligent communities and discloses an intelligent security scheduling management method based on deep learning, which comprises the steps of obtaining video monitoring data and access control record data of a plurality of communities, cleaning the data, unifying data formats and contents by adopting a data standardization technology to obtain standardized community security data, mining personnel flow modes and vehicle track modes of different communities by adopting a graph algorithm analysis technology on the basis of a community relation network to obtain a cross-community personnel flow mode and a vehicle track mode, identifying potential safety hazards by adopting an anomaly detection algorithm according to analysis results of the cross-community personnel flow mode and the vehicle track mode, and generating early warning information if an anomaly condition is detected. The method and the system realize deep association analysis of the cross-community security data and effectively improve the intelligent level of community security management.
Description
Technical Field
The invention relates to the technical field of intelligent communities, in particular to an intelligent security dispatching management method for intelligent communities based on deep learning.
Background
In intelligent security scheduling management of intelligent communities, cross-domain association mining needs to be performed on massive heterogeneous security data from different communities so as to discover potential cross-regional potential safety hazards in time. However, because of the large differences in data format, data quality and data granularity of the various communities, directly performing data association analysis presents a number of technical challenges. First, the raw data from different sources needs to be cleaned and standardized, but the lack of uniform data specifications and standards results in a complex and time-consuming data preprocessing process. Secondly, the storage and management of massive heterogeneous data requires the construction of a proper data model, and the traditional relational database is difficult to effectively process highly-correlated complex data. Furthermore, cross-community data associative analysis involves time and space dimensions, and efficient spatio-temporal indexing needs to be established to support fast retrieval and computation, but indexing techniques for high-dimensional spatio-temporal data remain to be further optimized. Finally, cross-domain data association mining needs to analyze by means of a graph database and a graph algorithm to construct a relationship network of personnel and vehicles, however, the graph data has high calculation complexity and higher requirements on real-time performance and concurrency.
Therefore, there is a need for a smart community intelligent security scheduling management method based on deep learning, which automatically learns the association features of heterogeneous data by introducing a deep learning technology, optimizes the space-time index and the graph calculation efficiency, and thus realizes efficient, accurate and real-time cross-domain security data association mining.
Disclosure of Invention
The invention aims to overcome one or more of the prior technical problems and provides an intelligent security dispatching management method for intelligent communities based on deep learning.
In order to achieve the above purpose, the intelligent security and protection scheduling management method for intelligent communities based on deep learning provided by the invention comprises the following steps:
Acquiring video monitoring data and access control record data of a plurality of communities, cleaning the acquired data, removing noise and abnormal values, and unifying data formats and contents to obtain standardized community security data;
Constructing a relationship network model of personnel and vehicles among communities according to standardized community security data, wherein nodes in the relationship network model represent personnel or vehicles, represent association relations among the personnel or vehicles, and represent association relations among the personnel or vehicles;
Based on a relational network model, adopting a graph algorithm analysis technology to mine personnel flow modes and vehicle track modes of different communities to obtain a community-crossing personnel flow mode and a vehicle track mode;
organizing and managing community security data by adopting a space-time index technology, so as to perform space-time dimension cross-community data retrieval, associating each piece of data with time and space positions, constructing a multi-dimensional index structure, and retrieving personnel and vehicle information in a specific space-time range;
Identifying potential safety hazards by adopting an anomaly detection algorithm according to analysis results of the inter-community personnel flow mode and the vehicle track mode, and generating early warning information if an anomaly condition is detected;
aiming at the identified potential safety hazards, a cross-community relation graph and a space-time track graph are generated by adopting a visualization technology, the relation graph and the track graph show the flowing condition of personnel and vehicles among different communities, and abnormal personnel and vehicles are highlighted, so that security personnel can check and treat the security personnel.
According to one aspect of the invention, video monitoring data and access control record data of a plurality of communities are obtained, and the data are preprocessed to remove noise and abnormal values in the data;
unifying formats and contents of the preprocessed data by adopting a data standardization technology, and eliminating differences of different community data formats and contents to obtain standardized community security data;
Extracting features of standardized community security data, obtaining features such as key frames and target tracks in video monitoring data, and access time and personnel identity in access record data;
constructing feature vectors of community security data according to the extracted features, and performing numerical conversion on video monitoring data and access control record data;
Fusing video monitoring data and access control record data by adopting a data fusion technology, and comprehensively utilizing the characteristics of the two types of data to obtain community security information;
Mining implicit relations in community security data through a data association analysis technology, and finding valuable information such as suspicious personnel, abnormal events and the like;
if suspicious personnel or abnormal events are found, triggering an early warning mechanism, automatically generating early warning information and notifying related management personnel, and timely taking countermeasures;
according to the requirements of community security management, the fused community security data is visually displayed, and a visual and understandable chart and map are generated, so that management staff can grasp the community security condition conveniently;
And storing the processed community security data into a database to form a community security big data resource.
According to one aspect of the invention, a basic information table of community personnel and vehicles is constructed according to standardized community security data;
creating a relationship network model of personnel and vehicles among communities by adopting a graph database technology, wherein nodes represent personnel or vehicles, and edges represent association relations among the personnel or vehicles;
Acquiring association information between personnel and vehicles and between personnel and vehicles by analyzing community security data;
Adding corresponding edges in the relational network model according to the acquired association information, representing the association relation between people or vehicles, and endowing the edges with corresponding attributes;
inquiring and analyzing the relational network model to acquire personnel and vehicle related information crossing communities;
judging the association degree of people and vehicles in different communities according to the query result, and obtaining a relationship network diagram of the people and the vehicles across communities;
identifying important personnel and vehicles with frequent activities crossing communities through a relational network graph, and carrying out important attention and analysis;
according to the analysis result, determining abnormal behavior patterns of personnel and vehicles crossing communities, and establishing an abnormal behavior pattern library;
And a real-time updating mechanism of the graph database is adopted to update the association relation between the personnel and the vehicles in the relation network model, so that the relation network model is ensured to keep synchronous with the actual situation, and real-time and accurate data support is provided for the cross-community personnel and vehicle association analysis.
According to one aspect of the invention, an inter-community personnel flow graph and a vehicle track graph are constructed according to community relation network data, wherein nodes represent communities, and edges represent personnel flows or vehicle tracks;
clustering the inter-community personnel flow graphs by adopting a graph clustering algorithm to obtain personnel flow modes, wherein each cluster represents one personnel flow mode;
graph embedding is carried out on the inter-community vehicle track graphs, and the vehicle tracks are expressed as low-dimensional vectors;
Obtaining a vehicle track mode by carrying out cluster analysis on the vehicle track vector, wherein each cluster corresponds to one vehicle track mode;
According to the personnel flow mode and the vehicle track mode, constructing a correlation diagram of personnel and vehicles crossing communities, wherein nodes are the personnel flow mode and the vehicle track mode, and if the personnel flow mode is highly related to the vehicle track mode, an edge is established between the personnel flow mode and the vehicle track mode;
Sampling on a correlation diagram of personnel and vehicles crossing communities by adopting a random walk algorithm to obtain a track correlation sequence of personnel flow and vehicles;
The method comprises the steps that pattern mining is carried out on a correlation sequence, so that frequently-occurring track patterns of personnel flow and vehicles are obtained to serve as correlation patterns of personnel and vehicles across communities;
judging the significance of the association mode of the cross-community personnel and the vehicle, and if the support degree and the confidence degree of the association mode exceed the threshold value, taking the association mode as the association mode of the significant cross-community personnel and the vehicle;
according to the association modes of the cross-community personnel and the vehicles, the association rules of the personnel flows and the vehicle tracks among different communities are determined, and the internal connection of the cross-community personnel flows and the vehicle movements is revealed.
According to one aspect of the invention, a space-time index technology is adopted to organize and manage community security data, each piece of data is associated with time and space positions, and a multidimensional index structure is constructed;
according to the characteristics of community security data, a data structure of a space-time index is designed, wherein the data structure comprises index information of a time dimension and a space dimension;
organizing community security data according to a time sequence through a time dimension index to form a time dimension index tree, and dividing the community security data according to a space position through a space dimension index to form a space dimension index tree;
Acquiring time and space dimension information in a space-time index tree, constructing a query condition of a space-time index, searching index nodes meeting the condition in the time dimension index tree if the query condition comprises a time range, and searching index nodes meeting the condition in the space dimension index tree if the query condition comprises a space range;
according to the searched time dimension and space dimension index nodes, acquiring a corresponding community security data subset, and screening out personnel and vehicle information meeting space-time query conditions;
and returning the screened personnel and vehicle information as a result of space-time cross-community data retrieval, so as to realize quick retrieval within a specific space-time range.
According to one aspect of the invention, an anomaly detection model is constructed according to analysis results of a cross-community personnel flow mode and a vehicle track mode, wherein the model comprises a personnel flow anomaly detection sub-model and a vehicle track anomaly detection sub-model;
performing clustering analysis on the personnel flow data by adopting a clustering algorithm to obtain a normal mode of personnel flow;
judging whether the personnel flow is abnormal or not by comparing the deviation degree of the real-time personnel flow data and the normal mode;
if the personnel flow data deviate from the normal mode and exceed a preset threshold value, the personnel flow abnormality exists, and personnel flow abnormality early warning information is generated;
Acquiring historical data of a vehicle track, and excavating a frequent mode of the vehicle track by adopting a sequence mode excavation algorithm;
Judging whether the vehicle track is abnormal or not by calculating the similarity between the real-time vehicle track and the frequent pattern;
If the similarity between the real-time vehicle track and all the frequent modes is lower than a preset threshold value, the vehicle track is considered to be abnormal, and vehicle track abnormality early warning information is generated;
combining the personnel flow abnormality early warning information and the vehicle track abnormality early warning information, and determining a final abnormality early warning level by adopting an abnormality information fusion algorithm;
And according to the abnormal early warning level, adopting corresponding safety precaution measures to eliminate potential safety hazards.
According to one aspect of the invention, a data acquisition technology is adopted to acquire real-time position information of personnel and vehicles in a community and identity information of the personnel and the vehicles;
Constructing space-time track data of the flow of personnel and vehicles among different communities according to the acquired position information and identity information;
Analyzing the space-time track data of the personnel and the vehicles through a data analysis technology, and identifying the personnel and the vehicles which frequently flow in a plurality of social intervals;
Judging whether the current flow behavior is abnormal or not according to the historical flow modes of the personnel and the vehicle by adopting an abnormality detection algorithm to obtain information of the abnormal personnel and the vehicle;
According to the information of the abnormal personnel and the vehicle, combining the identity information of the abnormal personnel and the vehicle, and mining possible association relations between the abnormal personnel and the vehicle through an association analysis technology;
Adopting a social network analysis technology, and constructing a relationship graph of the cross-community personnel and the vehicle based on the association relationship between the abnormal personnel and the vehicle;
The space-time track data of abnormal personnel and vehicles are fused with the relation map by using a space-time data visualization technology, and an intuitive space-time track map is generated;
In the space-time track diagram, different colors or icons are adopted to highlight the identified abnormal personnel and vehicles, so that the security personnel can be conveniently and quickly positioned;
through visual interaction technology, provide the operation function on the picture for security personnel, inquire about the detailed information of unusual personnel and vehicle to realize accurate checking and quick processing.
In order to achieve the above purpose, the present invention provides an intelligent security dispatching management system for intelligent communities based on deep learning, comprising:
the community security data acquisition module is used for acquiring video monitoring data and access control record data of a plurality of communities, cleaning the acquired data, removing noise and abnormal values, and unifying data formats and contents to obtain standardized community security data;
The relation network model construction module is used for constructing a relation network model of personnel and vehicles among communities according to standardized community security data, wherein nodes in the relation network model represent personnel or vehicles, and represent association relations among the personnel or vehicles, so as to perform association representation on cross-community data;
The behavior mode acquisition module is used for mining personnel flow modes and vehicle track modes among different communities by adopting a graph algorithm analysis technology based on a relational network model to obtain a cross-community personnel flow mode and a vehicle track mode;
The behavior retrieval module is used for organizing and managing community security data by adopting a space-time index technology, so that space-time dimension cross-community data retrieval is performed, each piece of data is associated with time and space positions, a multidimensional index structure is constructed, and personnel and vehicle information in a specific space-time range are retrieved;
The behavior early warning module is used for identifying potential safety hazards by adopting an anomaly detection algorithm according to analysis results of the inter-community personnel flow mode and the vehicle track mode, and generating early warning information if an anomaly condition is detected;
And the early warning treatment module is used for generating a cross-community relation graph and a space-time track graph by adopting a visualization technology aiming at the identified potential safety hazards, displaying the flowing condition of personnel and vehicles among different communities by the relation graph and the track graph, and highlighting abnormal personnel and vehicles so as to enable security personnel to check and treat the abnormal personnel and vehicles.
In order to achieve the above purpose, the invention provides an electronic device, which comprises a processor, a memory and a computer program stored on the memory and capable of running on the processor, wherein the computer program realizes the intelligent security scheduling management method for intelligent communities based on deep learning when being executed by the processor.
In order to achieve the above object, the present invention provides a computer readable storage medium, on which a computer program is stored, which when executed by a processor implements the above intelligent security scheduling management method for intelligent communities based on deep learning.
Firstly, acquiring video monitoring and access control record data of a plurality of communities, cleaning and standardizing the video monitoring and access control record data, and then constructing a relationship network model of personnel and vehicles among communities by using a graph database technology to realize the associated representation of cross-community data;
adopting a graph algorithm analysis technology to mine personnel flow and vehicle track modes among different communities, supporting data retrieval of space-time dimensions, adopting a space-time index technology to organize and manage security data, identifying potential safety hazards through an anomaly detection algorithm and generating early warning information;
generating a cross-community relationship map and a space-time track map by using a visualization technology, and intuitively displaying abnormal conditions;
the method realizes the deep association analysis of the cross-community security data, and effectively improves the intelligent level of community security management.
Drawings
FIG. 1 is a flow chart of a smart community intelligent security dispatch management method based on deep learning, according to an example embodiment;
fig. 2 is a flow chart illustrating a smart community smart security dispatch management system based on deep learning according to an example embodiment.
Detailed Description
The present disclosure will now be discussed with reference to exemplary embodiments, it being understood that the embodiments discussed are merely for the purpose of enabling those of ordinary skill in the art to better understand and thus practice the present disclosure and do not imply any limitation to the scope of the present disclosure.
As used herein, the term "comprising" and variants thereof are to be interpreted as meaning "including but not limited to" open-ended terms. The terms "based on" and "based at least in part on" are to be construed as "at least one embodiment.
According to an embodiment of the present invention, fig. 1 is a flowchart of a smart security and protection scheduling management method for smart and intelligent communities based on deep learning according to an exemplary embodiment, and as shown in fig. 1, in order to achieve the above purpose, the smart security and protection scheduling management method for smart and intelligent communities based on deep learning provided by the present invention includes:
Acquiring video monitoring data and access control record data of a plurality of communities, cleaning the acquired data, removing noise and abnormal values, and unifying data formats and contents to obtain standardized community security data;
Constructing a relationship network model of personnel and vehicles among communities according to standardized community security data, wherein nodes in the relationship network model represent personnel or vehicles, represent association relations among the personnel or vehicles, and represent association relations among the personnel or vehicles;
Based on a relational network model, adopting a graph algorithm analysis technology to mine personnel flow modes and vehicle track modes of different communities to obtain a community-crossing personnel flow mode and a vehicle track mode;
organizing and managing community security data by adopting a space-time index technology, so as to perform space-time dimension cross-community data retrieval, associating each piece of data with time and space positions, constructing a multi-dimensional index structure, and retrieving personnel and vehicle information in a specific space-time range;
Identifying potential safety hazards by adopting an anomaly detection algorithm according to analysis results of the inter-community personnel flow mode and the vehicle track mode, and generating early warning information if an anomaly condition is detected;
aiming at the identified potential safety hazards, a cross-community relation graph and a space-time track graph are generated by adopting a visualization technology, the relation graph and the track graph show the flowing condition of personnel and vehicles among different communities, and abnormal personnel and vehicles are highlighted, so that security personnel can check and treat the security personnel.
According to one embodiment of the invention, video monitoring data and access control record data of a plurality of communities are obtained, and the data are preprocessed to remove noise and abnormal values in the data;
unifying formats and contents of the preprocessed data by adopting a data standardization technology, and eliminating differences of different community data formats and contents to obtain standardized community security data;
Extracting features of standardized community security data, obtaining features such as key frames and target tracks in video monitoring data, and access time and personnel identity in access record data;
constructing feature vectors of community security data according to the extracted features, and performing numerical conversion on video monitoring data and access control record data;
Fusing video monitoring data and access control record data by adopting a data fusion technology, and comprehensively utilizing the characteristics of the two types of data to obtain community security information;
Mining implicit relations in community security data through a data association analysis technology, and finding valuable information such as suspicious personnel, abnormal events and the like;
if suspicious personnel or abnormal events are found, triggering an early warning mechanism, automatically generating early warning information and notifying related management personnel, and timely taking countermeasures;
according to the requirements of community security management, the fused community security data is visually displayed, and a visual and understandable chart and map are generated, so that management staff can grasp the community security condition conveniently;
And storing the processed community security data into a database to form a community security big data resource.
According to one embodiment of the invention, a basic information table of community personnel and vehicles is constructed according to standardized community security data;
creating a relationship network model of personnel and vehicles among communities by adopting a graph database technology, wherein nodes represent personnel or vehicles, and edges represent association relations among the personnel or vehicles;
Acquiring association information between personnel and vehicles and between personnel and vehicles by analyzing community security data;
Adding corresponding edges in the relational network model according to the acquired association information, representing the association relation between people or vehicles, and endowing the edges with corresponding attributes;
inquiring and analyzing the relational network model to acquire personnel and vehicle related information crossing communities;
judging the association degree of people and vehicles in different communities according to the query result, and obtaining a relationship network diagram of the people and the vehicles across communities;
identifying important personnel and vehicles with frequent activities crossing communities through a relational network graph, and carrying out important attention and analysis;
according to the analysis result, determining abnormal behavior patterns of personnel and vehicles crossing communities, and establishing an abnormal behavior pattern library;
And a real-time updating mechanism of the graph database is adopted to update the association relation between the personnel and the vehicles in the relation network model, so that the relation network model is ensured to keep synchronous with the actual situation, and real-time and accurate data support is provided for the cross-community personnel and vehicle association analysis.
According to one embodiment of the invention, an inter-community personnel flow graph and a vehicle track graph are constructed according to community relation network data, wherein nodes represent communities, and edges represent personnel flows or vehicle tracks;
clustering the inter-community personnel flow graphs by adopting a graph clustering algorithm to obtain personnel flow modes, wherein each cluster represents one personnel flow mode;
graph embedding is carried out on the inter-community vehicle track graphs, and the vehicle tracks are expressed as low-dimensional vectors;
Obtaining a vehicle track mode by carrying out cluster analysis on the vehicle track vector, wherein each cluster corresponds to one vehicle track mode;
According to the personnel flow mode and the vehicle track mode, constructing a correlation diagram of personnel and vehicles crossing communities, wherein nodes are the personnel flow mode and the vehicle track mode, and if the personnel flow mode is highly related to the vehicle track mode, an edge is established between the personnel flow mode and the vehicle track mode;
Sampling on a correlation diagram of personnel and vehicles crossing communities by adopting a random walk algorithm to obtain a track correlation sequence of personnel flow and vehicles;
The method comprises the steps that pattern mining is carried out on a correlation sequence, so that frequently-occurring track patterns of personnel flow and vehicles are obtained to serve as correlation patterns of personnel and vehicles across communities;
judging the significance of the association mode of the cross-community personnel and the vehicle, and if the support degree and the confidence degree of the association mode exceed the threshold value, taking the association mode as the association mode of the significant cross-community personnel and the vehicle;
according to the association modes of the cross-community personnel and the vehicles, the association rules of the personnel flows and the vehicle tracks among different communities are determined, and the internal connection of the cross-community personnel flows and the vehicle movements is revealed.
According to one embodiment of the invention, a space-time index technology is adopted to organize and manage community security data, each piece of data is associated with time and space positions, and a multidimensional index structure is constructed;
according to the characteristics of community security data, a data structure of a space-time index is designed, wherein the data structure comprises index information of a time dimension and a space dimension;
organizing community security data according to a time sequence through a time dimension index to form a time dimension index tree, and dividing the community security data according to a space position through a space dimension index to form a space dimension index tree;
Acquiring time and space dimension information in a space-time index tree, constructing a query condition of a space-time index, searching index nodes meeting the condition in the time dimension index tree if the query condition comprises a time range, and searching index nodes meeting the condition in the space dimension index tree if the query condition comprises a space range;
according to the searched time dimension and space dimension index nodes, acquiring a corresponding community security data subset, and screening out personnel and vehicle information meeting space-time query conditions;
and returning the screened personnel and vehicle information as a result of space-time cross-community data retrieval, so as to realize quick retrieval within a specific space-time range.
According to one embodiment of the invention, an abnormality detection model is constructed according to analysis results of a cross-community personnel flow mode and a vehicle track mode, wherein the model comprises a personnel flow abnormality detection sub-model and a vehicle track abnormality detection sub-model;
performing clustering analysis on the personnel flow data by adopting a clustering algorithm to obtain a normal mode of personnel flow;
judging whether the personnel flow is abnormal or not by comparing the deviation degree of the real-time personnel flow data and the normal mode;
if the personnel flow data deviate from the normal mode and exceed a preset threshold value, the personnel flow abnormality exists, and personnel flow abnormality early warning information is generated;
Acquiring historical data of a vehicle track, and excavating a frequent mode of the vehicle track by adopting a sequence mode excavation algorithm;
Judging whether the vehicle track is abnormal or not by calculating the similarity between the real-time vehicle track and the frequent pattern;
If the similarity between the real-time vehicle track and all the frequent modes is lower than a preset threshold value, the vehicle track is considered to be abnormal, and vehicle track abnormality early warning information is generated;
combining the personnel flow abnormality early warning information and the vehicle track abnormality early warning information, and determining a final abnormality early warning level by adopting an abnormality information fusion algorithm;
And according to the abnormal early warning level, adopting corresponding safety precaution measures to eliminate potential safety hazards.
According to one embodiment of the invention, a data acquisition technology is adopted to acquire real-time position information of personnel and vehicles in a community and identity information of the personnel and the vehicles;
Constructing space-time track data of the flow of personnel and vehicles among different communities according to the acquired position information and identity information;
Analyzing the space-time track data of the personnel and the vehicles through a data analysis technology, and identifying the personnel and the vehicles which frequently flow in a plurality of social intervals;
Judging whether the current flow behavior is abnormal or not according to the historical flow modes of the personnel and the vehicle by adopting an abnormality detection algorithm to obtain information of the abnormal personnel and the vehicle;
According to the information of the abnormal personnel and the vehicle, combining the identity information of the abnormal personnel and the vehicle, and mining possible association relations between the abnormal personnel and the vehicle through an association analysis technology;
Adopting a social network analysis technology, and constructing a relationship graph of the cross-community personnel and the vehicle based on the association relationship between the abnormal personnel and the vehicle;
The space-time track data of abnormal personnel and vehicles are fused with the relation map by using a space-time data visualization technology, and an intuitive space-time track map is generated;
In the space-time track diagram, different colors or icons are adopted to highlight the identified abnormal personnel and vehicles, so that the security personnel can be conveniently and quickly positioned;
through visual interaction technology, provide the operation function on the picture for security personnel, inquire about the detailed information of unusual personnel and vehicle to realize accurate checking and quick processing.
According to one embodiment of the invention, the original data are firstly obtained from monitoring systems and access control systems of a plurality of communities through an API interface, the data are preprocessed, the noise is removed by using median filtering, and abnormal values are removed by adopting a 3 sigma criterion. And then, the JSON format is utilized to normalize the data and unify the formats of the data with different sources. And extracting key frames from the video data, identifying targets such as personnel, vehicles and the like by adopting a YOLOv target detection algorithm, and tracking target tracks by adopting a DeepSORT algorithm. And extracting the characteristics of time, ID and the like from the entrance guard record. And converting the extracted features into feature vectors, wherein the feature vectors comprise the attributes of position coordinates, speed, body types and the like. And fusing the video and the access control data by adopting a D-S evidence theory, and calculating the confidence coefficient after fusion. Data association rules are mined using Apriori algorithms, such as finding suspicious patterns of behavior that occur frequently. And setting an early warning threshold, automatically triggering early warning when abnormal behaviors are found, and informing a manager through a short message. A map of the thermodynamic, trajectory, etc. visualization is generated using ECharts. And finally, storing the processed data into a MongoDB database, and establishing index optimization query performance. The whole process forms a complete technical chain from data acquisition, preprocessing, feature extraction to data fusion, association analysis and visualization, and realizes intelligent processing and application of community security data.
According to one embodiment of the invention, when constructing community personnel and vehicle basic information tables, a MySQL database can be adopted to design a data table containing personnel ID, name, identification card number, vehicle ID, license plate number and other fields. For example, the personnel information table may contain 10000 records, and the vehicle information table may contain 5000 records. When the relationship network model of personnel and vehicles in the social space is created, a Neo4j graph database can be adopted, and the personnel and vehicles are used as nodes and the association relationship is used as an edge. By analyzing 1000 pieces of community security data within 1 month and utilizing a space-time correlation algorithm, if two persons or vehicles appear at the same place at the same time, adding a correlation edge in the graph model, and recording the properties of correlation time, place, frequency and the like. By using the Cypher query language, the activity track of a person in different communities can be queried, for example, zhang Sanu appears in A, B, C communities in a week. And visualizing the query result, generating a cross-community personnel and vehicle relation network diagram, and identifying key personnel and vehicles with higher centrality through a PageRank algorithm. For the key objects, an association rule mining algorithm is adopted to analyze the activity rule and the social network, for example, the phenomenon that Zhang san happens to be on the surface of the community A with Lisi in the morning every Monday and frequently goes to and from the community B, C on the weekend is found. Through cluster analysis, abnormal behavior patterns, such as frequent replacement of license plates of 5 vehicles in one month, can be found, and the abnormal behavior patterns occur in criminal high-incidence areas for multiple times. And adding the abnormal modes into an early warning rule base, and detecting abnormal behaviors in real time. And finally, updating the relational network model in real time by using an event-driven mechanism of the graph database, and ensuring the accuracy of analysis results.
According to one embodiment of the present invention, an inter-community personnel flow graph and a vehicle track graph are first constructed based on community relationship network data. Where nodes represent communities and edges represent personnel flows or vehicle trajectories. For example, each community can be abstracted into a node, and directed edges are established among the nodes of the corresponding communities according to the personnel flow records, wherein the edge weights are the personnel flow amounts. Similarly, vehicle track edges are established among communities according to the vehicle GPS track data. And then, carrying out community detection on the personnel flow chart by adopting a Louvain algorithm, and taking a community division result with the maximum modularity as a personnel flow mode. And then, learning 64-dimensional embedded vector representation of nodes in the vehicle track graph by using a Node2Vec algorithm, and clustering vectors by using K-Means to obtain 8 vehicle track modes. A cross-community personnel-vehicle association diagram is constructed on the basis, for example, when the Jaccard similarity of the personnel flow pattern i and the vehicle track pattern j exceeds 6, an edge is added between corresponding nodes of the personnel flow pattern i and the vehicle track pattern j. And then adopting Node2Vec to perform random walk sampling on the association graph, selecting 100 nodes each time, wherein the walk length is 10, and iterating for 50 rounds to obtain a personnel flow-vehicle track association sequence. And excavating frequent subsequence modes from the sequence by utilizing PrefixSpan algorithm, setting a support degree threshold value to be 05, setting a confidence degree threshold value to be 8, and finally obtaining 10 remarkable cross-community personnel-vehicle association modes. Analysis has found that the association patterns associated with business and traffic junctions are most pronounced, indicating that business and traffic activities are key factors in traction across community personnel flow and vehicle movement. In summary, the present study reveals the inherent correlation law of inter-community personnel flow and vehicle movement in a spatiotemporal pattern, which is of great value for understanding urban personnel-vehicle dynamics and optimizing traffic management.
According to one embodiment of the invention, in community security data management, a space-time index technology can be adopted to efficiently organize and retrieve data. Firstly, according to the characteristics of community security data, a multidimensional index structure comprising a time dimension and a space dimension is designed. The time dimension index may use a data structure such as a b+ tree to sort the data according to the time stamps to form an index tree of the time dimension. The space dimension index can adopt a space index structure such as an R tree, a quadtree and the like to divide the community into different space regions, and each region corresponds to one index node to form an index tree of the space dimension. When the query condition is constructed, if the time range is included, a binary search algorithm is used in the time dimension index tree to locate the index nodes meeting the time condition, and if the space range is included, a rectangular region intersection algorithm is used in the space dimension index tree to locate the index nodes falling in the target region. The community security data subset meeting the space-time condition can be rapidly obtained through the joint query of the space-time index tree. And finally, further screening personnel and vehicle information meeting specific query conditions from the data subset, and returning the personnel and vehicle information as a result of space-time cross-community data retrieval. Through a reasonably designed space-time index structure and an efficient query algorithm, the rapid retrieval of massive community security data can be realized, flexible space-time query conditions are supported, and the efficiency and performance of community security data management are improved.
According to one embodiment of the invention, an anomaly detection model is constructed according to the analysis results of the cross-community personnel flow pattern and the vehicle track pattern. The model comprises a personnel flow abnormality detection sub-model and a vehicle track abnormality detection sub-model. Firstly, carrying out cluster analysis on personnel flow data of 3 months by adopting a K-means clustering algorithm, classifying the personnel flow modes into 5 types, and respectively representing normal flow modes of working, learning, shopping, medical seeking and the like to obtain a normal mode of personnel flow. And then, calculating Euclidean distance between the real-time personnel flow data and each normal mode by comparing the deviation degree of the real-time personnel flow data and the normal mode, and if the distance from the nearest normal mode exceeds a preset threshold (e.g. 5), considering that personnel flow is abnormal, and generating personnel flow abnormality early warning information. Next, history data of the vehicle track for nearly 1 year is acquired, a PrefixSpan-sequence pattern mining algorithm is adopted to mine out a frequent pattern of the vehicle track, and the support degree threshold is set to be 20%. And judging whether the vehicle track is abnormal or not by calculating the similarity between the real-time vehicle track and the frequent pattern (such as by adopting an edit distance algorithm). If the similarity between the real-time vehicle track and all the frequent modes is lower than a preset threshold (e.g. 8), the vehicle track is considered to be abnormal, and vehicle track abnormality early warning information is generated. And finally, integrating the personnel flow abnormality early warning information and the vehicle track abnormality early warning information, respectively assigning weights 6 and 4 by adopting a weighted average algorithm, calculating an abnormality early warning comprehensive score, and determining a final abnormality early warning level (for example, 0-20 is divided into first-level early warning, 20-50 is divided into second-level early warning, and more than 50 is three-level early warning). And according to the abnormal early warning level, adopting corresponding safety precaution measures, such as strengthening the monitoring frequency of a key area during primary early warning, checking suspicious personnel and vehicles during secondary early warning, starting an emergency plan during tertiary early warning, and the like, so as to eliminate potential safety hazards.
Furthermore, in order to achieve the above object, the present invention provides a smart security and security management system for smart communities based on deep learning, and fig. 2 is a flowchart of a smart security and security management system for smart communities based on deep learning according to an exemplary embodiment, and as shown in fig. 2, the smart security and security management system for smart communities based on deep learning in the present invention includes:
the community security data acquisition module is used for acquiring video monitoring data and access control record data of a plurality of communities, cleaning the acquired data, removing noise and abnormal values, and unifying data formats and contents to obtain standardized community security data;
The relation network model construction module is used for constructing a relation network model of personnel and vehicles among communities according to standardized community security data, wherein nodes in the relation network model represent personnel or vehicles, and represent association relations among the personnel or vehicles, so as to perform association representation on cross-community data;
The behavior mode acquisition module is used for mining personnel flow modes and vehicle track modes among different communities by adopting a graph algorithm analysis technology based on a relational network model to obtain a cross-community personnel flow mode and a vehicle track mode;
The behavior retrieval module is used for organizing and managing community security data by adopting a space-time index technology, so that space-time dimension cross-community data retrieval is performed, each piece of data is associated with time and space positions, a multidimensional index structure is constructed, and personnel and vehicle information in a specific space-time range are retrieved;
The behavior early warning module is used for identifying potential safety hazards by adopting an anomaly detection algorithm according to analysis results of the inter-community personnel flow mode and the vehicle track mode, and generating early warning information if an anomaly condition is detected;
And the early warning treatment module is used for generating a cross-community relation graph and a space-time track graph by adopting a visualization technology aiming at the identified potential safety hazards, displaying the flowing condition of personnel and vehicles among different communities by the relation graph and the track graph, and highlighting abnormal personnel and vehicles so as to enable security personnel to check and treat the abnormal personnel and vehicles.
In order to achieve the aim of the invention, the invention also provides electronic equipment which comprises a processor, a memory and a computer program stored on the memory and capable of running on the processor, wherein the computer program is executed by the processor to achieve the intelligent security scheduling management method of the intelligent wisdom community based on deep learning.
In order to achieve the above object, the present invention further provides a computer readable storage medium, on which a computer program is stored, which when executed by a processor implements the intelligent security scheduling management method for intelligent communities based on deep learning.
Based on the method, the method has the advantages that video monitoring and access control record data of a plurality of communities are firstly obtained, cleaning and standardization processing are carried out on the video monitoring and access control record data, then a relationship network model of personnel and vehicles among communities is built by using a graph database technology, association representation of cross-community data is achieved, personnel flow and vehicle track modes among different communities are mined by adopting a graph algorithm analysis technology, space-time dimension data retrieval is supported, security data are organized and managed by adopting a space-time index technology, potential safety hazards are identified by an anomaly detection algorithm and early warning information is generated, a cross-community relationship map and a space-time track map are generated by using a visualization technology, anomaly conditions are visually displayed, depth association analysis of the cross-community security data is achieved, and intelligent level of community security management is effectively improved.
Those of ordinary skill in the art will appreciate that the modules and algorithm steps described in connection with the embodiments disclosed herein can be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
It will be clearly understood by those skilled in the art that, for convenience and brevity of description, specific working procedures of the apparatus and device described above may refer to corresponding procedures in the foregoing method embodiments, which are not described herein again.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, and for example, the division of the modules is merely a logical function division, and there may be additional divisions when actually implemented, for example, multiple modules or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or modules, which may be in electrical, mechanical, or other forms.
The modules described as separate components may or may not be physically separate, and components shown as modules may or may not be physical modules, i.e., may be located in one place, or may be distributed over a plurality of network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment of the invention.
In addition, each functional module in the embodiment of the present invention may be integrated in one processing module, or each module may exist alone physically, or two or more modules may be integrated in one module.
The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method for energy saving signal transmission/reception of the various embodiments of the present invention. The storage medium includes various media capable of storing program codes such as a U disk, a mobile hard disk, a ROM, a RAM, a magnetic disk or an optical disk.
The above description is only illustrative of the preferred embodiments of the present application and of the principles of the technology employed. It will be appreciated by persons skilled in the art that the scope of the application referred to in the present application is not limited to the specific combinations of the technical features described above, but also covers other technical features formed by any combination of the technical features described above or their equivalents without departing from the inventive concept. Such as the above-mentioned features and the technical features disclosed in the present application (but not limited to) having similar functions are replaced with each other.
It should be understood that, the sequence numbers of the steps in the summary and the embodiments of the present invention do not necessarily mean the order of execution, and the execution order of the processes should be determined by the functions and the internal logic, and should not be construed as limiting the implementation process of the embodiments of the present invention.
Claims (10)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411559505.1A CN119671794A (en) | 2024-11-04 | 2024-11-04 | A smart community intelligent security dispatch management method based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411559505.1A CN119671794A (en) | 2024-11-04 | 2024-11-04 | A smart community intelligent security dispatch management method based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN119671794A true CN119671794A (en) | 2025-03-21 |
Family
ID=94993689
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202411559505.1A Pending CN119671794A (en) | 2024-11-04 | 2024-11-04 | A smart community intelligent security dispatch management method based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119671794A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN120259984A (en) * | 2025-06-05 | 2025-07-04 | 北京升哲科技有限公司 | Security camera community relationship analysis method, device and system based on minimum entropy |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109918368A (en) * | 2019-03-27 | 2019-06-21 | 成都市公安科学技术研究所 | A kind of system and method that vehicle driver is identified by Track association degree |
| CN110766895A (en) * | 2019-09-17 | 2020-02-07 | 重庆特斯联智慧科技股份有限公司 | Intelligent community abnormity alarm system and method based on target trajectory analysis |
| CN111260525A (en) * | 2020-01-16 | 2020-06-09 | 深圳市广道高新技术股份有限公司 | Community security situation perception and early warning method, system and storage medium |
| WO2022047607A1 (en) * | 2020-09-01 | 2022-03-10 | 江苏星瀚智慧城市科技有限公司 | Community service system |
| CN118643465A (en) * | 2024-08-12 | 2024-09-13 | 常州市公安局金坛分局 | Case investigation auxiliary system based on multi-source data correlation analysis |
| CN118821019A (en) * | 2024-07-24 | 2024-10-22 | 厦门立林科技有限公司 | Smart community personnel track recognition method and system |
-
2024
- 2024-11-04 CN CN202411559505.1A patent/CN119671794A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109918368A (en) * | 2019-03-27 | 2019-06-21 | 成都市公安科学技术研究所 | A kind of system and method that vehicle driver is identified by Track association degree |
| CN110766895A (en) * | 2019-09-17 | 2020-02-07 | 重庆特斯联智慧科技股份有限公司 | Intelligent community abnormity alarm system and method based on target trajectory analysis |
| CN111260525A (en) * | 2020-01-16 | 2020-06-09 | 深圳市广道高新技术股份有限公司 | Community security situation perception and early warning method, system and storage medium |
| WO2022047607A1 (en) * | 2020-09-01 | 2022-03-10 | 江苏星瀚智慧城市科技有限公司 | Community service system |
| CN118821019A (en) * | 2024-07-24 | 2024-10-22 | 厦门立林科技有限公司 | Smart community personnel track recognition method and system |
| CN118643465A (en) * | 2024-08-12 | 2024-09-13 | 常州市公安局金坛分局 | Case investigation auxiliary system based on multi-source data correlation analysis |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN120259984A (en) * | 2025-06-05 | 2025-07-04 | 北京升哲科技有限公司 | Security camera community relationship analysis method, device and system based on minimum entropy |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8253564B2 (en) | Predicting a future location of a moving object observed by a surveillance device | |
| Thongsatapornwatana | A survey of data mining techniques for analyzing crime patterns | |
| US20080201277A1 (en) | System architecture and process for automating intelligent surveillance center operation | |
| CN103392187A (en) | Scene activity analysis using statistical and semantic feature learnt from object trajectory data | |
| CN103281341A (en) | Network event processing method and device | |
| CN119671794A (en) | A smart community intelligent security dispatch management method based on deep learning | |
| CN117540372B (en) | Database intrusion detection and response system for intelligent learning | |
| CN117436729A (en) | Government system based data management and data analysis method | |
| Yamini | A violent crime analysis using fuzzy c-means clustering approach | |
| CN119538149A (en) | A multi-object and multi-model control and early warning system | |
| Prathap et al. | Geospatial crime analysis to determine crime density using Kernel density estimation for the Indian context | |
| Borg et al. | Clustering residential burglaries using modus operandi and spatiotemporal information | |
| Yu et al. | Network security monitoring method based on deep learning | |
| CN118097198A (en) | Automatic dressing compliance management and control system and method based on artificial intelligence | |
| CN117874109A (en) | Operation method of IT asset data on data management platform | |
| Isafiade et al. | Citisafe: Adaptive spatial pattern knowledge using fp-growth algorithm for crime situation recognition | |
| Khatun et al. | Data mining technique to analyse and predict crime using crime categories and arrest records | |
| Al-Shalabi | Perceptions of Crime Behavior and Relationships: Rough Set Based Approach'' | |
| Das et al. | Crime pattern detection using data mining | |
| Adeyiga et al. | A comparative analysis of selected clustering algorithms for criminal profiling | |
| Al-Oraiqat et al. | Bridging machine learning and formal concept analysis for effective crowd detection | |
| CN118093666A (en) | System for managing target object | |
| Anand et al. | Anomaly Detection in Disaster Recovery: A Review, Current Trends and New Perspectives | |
| Gutchess et al. | Learning patterns of human activity for anomaly detection | |
| Dasgupta et al. | Survey and analysis of mining requirements in spatio-temporal database |
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 |