A method for realizing personnel activity graph based on four-dimensional travel model
Through the personnel activity map method based on the four-dimensional travel model, the problems of travel laws and abnormal warning in the existing technology are solved, and the travel trajectory tracking and accident warning of the care and attention personnel are realized, and the efficiency and safety of government management are improved. sex.
Patent Information
- Application Number
- CN202210019157.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-01-10
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2042-01-10
AI Technical Summary
The existing technology cannot effectively realize the excavation of travel laws and abnormal movement trajectories of caring and paying attention to people, and offline visits and online video surveillance have problems such as low efficiency and difficulty in ensuring personal safety.
The personnel activity map implementation method based on the four-dimensional travel model is used to build a four-dimensional travel model of travelers, obtain face image data, extract time, place, and equipment information, generate travel trajectory, and mine close contacts of abnormal people according to the set rules. High-frequency travel places and people appearing in the same frame form a map of personnel activity.
It has realized the awareness of travel laws and abnormal behaviors of caring and caring people, improved the efficiency of precise service in government management, saved management time, and enhanced the technical defense measures of civil defense.
Smart Images

Figure CN114066431B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of big data mining and application technology, and in particular to a method for realizing a personnel activity graph based on a four-dimensional travel model. Background Art
[0002] At present, the management of the caring and concerned persons under the jurisdiction of government management agencies at all levels mainly includes two methods: offline visits and online checking of Skynet and Snow Bright cameras. Among them, for the offline visit method, government staff need to regularly visit the caring and concerned persons in the jurisdiction to understand the latest situation of the caring and concerned persons, and implement one-to-one monitoring for those who need special attention. This method has the following problems: First, there are a large number of caring and concerned persons in the jurisdiction, and government staff generally manage multiple caring and concerned persons by one person, especially during special times, it takes a long time for multiple people to monitor, which is time-consuming and labor-intensive; second, it is necessary to manually actively obtain abnormal information of caring and concerned persons, which has problems such as low efficiency, omissions, strong subjectivity and weak reference; third, when staff visit more dangerous and aggressive persons in the jurisdiction, their personal safety is difficult to guarantee. As for the online viewing of Skynet and Snow Bright cameras, in order to assist government management agencies at all levels in the management of personnel and public security, the country has installed a large number of Skynet and Snow Bright cameras in urban areas, towns and villages, hoping to use the cameras to always know the travel conditions of caring and concerned personnel. When an accident occurs in the jurisdiction, government staff manually call and view relevant cameras for accident evidence collection according to the time and place of the accident. This method saves a lot of human resources and collects accident evidence through historical recordings, which further improves the personal safety of staff. However, there are also the following problems: First, there are a large number of cameras, and it is impossible to accurately locate specific cameras. Government staff need to spend a lot of time manually viewing relevant camera resources; second, only historical accident tracing can be performed, and accident warning cannot be performed; third, there are many video resources involved in post-event evidence tracing, and it is difficult to accurately extract relevant information, which is inefficient and has a high possibility of error; fourth, the video information of Skynet and Snow Bright cameras is retained for one to three months, and the video cannot be viewed beyond the retention date. It can be seen from this that the above two methods are unable to explore the travel patterns of caring and concerned persons within the jurisdiction, unable to track the movement trajectories of caring and concerned persons, and also unable to provide early warnings of abnormal movement trajectories of caring and concerned persons and early warnings of accidents within the jurisdiction. Summary of the invention
[0003] The purpose of the present invention is to overcome the deficiencies of the prior art and to provide a method for realizing a personnel activity map based on a four-dimensional travel model.
[0004] The objective of the present invention is achieved through the following technical solutions:
[0005] A method for realizing a personnel activity graph based on a four-dimensional travel model comprises the following steps:
[0006] S1: Construct a four-dimensional travel model of travelers;
[0007] S2: Acquire facial image data of travelers, and input the facial image data into a four-dimensional travel model, the four-dimensional travel model extracts time information, location information, device information and personnel information from the facial image data, and generates a travel trajectory of the travelers based on the time information, location information, device information and personnel information;
[0008] S3: According to the travel trajectory, the abnormal persons among the target persons are determined according to the set abnormal travel determination rules. The target persons are the persons whose travel behavior needs to be analyzed among the travel persons;
[0009] S4: mining close contacts A who have contact with the abnormal person within the first time period, where the close contacts A are persons whose number of contact times with the abnormal person is greater than or equal to a first threshold;
[0010] S5: mining high-frequency travel places A visited by the abnormal person in the second time period, where the high-frequency travel places A are places visited by the abnormal person a number of times greater than or equal to a second threshold;
[0011] S6: mining high-frequency visitors to the high-frequency travel place A within a third time period, where the high-frequency visitors are people who have visited the place a number of times greater than or equal to a third threshold;
[0012] S7, mining persons who appear in the same frame with the abnormal person in the fourth time period, where the persons who appear in the same frame with the abnormal person are persons who appear at the same place, nearby associated place, or collection area of the same device at the same time as the abnormal person, and the time interval between their appearances is less than or equal to the first preset time interval, and the nearby associated place is a place whose distance from the high-frequency travel place A is less than or equal to the fourth threshold;
[0013] S8, mining high-frequency travel places B of high-frequency visitors in the fifth time period, where the high-frequency travel places B are places where the number of visits by high-frequency visitors is greater than or equal to a fifth threshold;
[0014] S9, mining high-frequency travel places C of the persons who appear in the same frame within the sixth time period, where the high-frequency travel places C are places where the number of visits of the persons who appear in the same frame is greater than or equal to a sixth threshold;
[0015] S10: mining close contact person B of the person who appears in the same frame within the seventh time period, where the close contact person B is a person who has had contact with the person who appears in the same frame for a number of times greater than or equal to a seventh threshold;
[0016] S11, mining close contacts C of close contact A within an eighth time period, where close contact C is a person who has had contact with close contact A for a number of times greater than or equal to an eighth threshold;
[0017] S12, mining high-frequency travel places D of close contact person A in the ninth time period, where the high-frequency travel places D are places where the number of visits by close contact person A is greater than or equal to a ninth threshold;
[0018] S13: A personnel tracking chain is formed based on the obtained close contacts A, high-frequency travel places A, high-frequency visitors, high-frequency travel places B, people appearing in the same frame, high-frequency travel places C, close contacts B, close contacts C, and high-frequency travel places D, and a personnel activity map is generated with the abnormal persons as the center point.
[0019] In step S1, the specific steps of constructing the four-dimensional travel model of travelers include:
[0020] S101: Gather historical facial image data of travelers to obtain a flow of travel logs;
[0021] S102: Cleaning the personnel travel log flow according to a preset data cleaning strategy;
[0022] S103: Loading type tags to the personnel travel log flow, the type tags including personnel type tag X, location tag Z and equipment tag Y;
[0023] S104: Segment the personnel travel log flow according to the loaded personnel type label X, and then input the segmented personnel travel log flow into a preset analysis model for training. The analysis model includes four-dimensional analysis channels. According to the analysis strategy configured in the analysis model, the personnel travel log flow is input into one or more of the analysis channels. The analysis model completes the analysis and classification of the personnel travel log flow in parallel. The four dimensions are time dimension, place dimension, equipment dimension and personnel dimension.
[0024] S105, constructing a four-dimensional travel model of travelers based on the trained analysis model.
[0025] The step S101, before aggregating the historical facial image data of the travelers to obtain the travel log flow of the travelers, also includes the following steps:
[0026] S001: Input the image data of the traveler into the pre-trained face recognition model to screen the face images, and match the screened face images with the face pictures stored in the face recognition model; if the match is successful, mark the face image data of the traveler with an identity ID; if the match fails, proceed to step S002;
[0027] S002: Input the facial image data of the travelers into a pre-trained facial clustering model for personnel clustering, and after clustering, mark the facial image data of the travelers with an identity recognition ID.
[0028] In step S002, the specific steps of inputting the facial image data of the travelers into the pre-trained facial clustering model for clustering the travelers include: performing similarity analysis on the facial image data of the travelers and the facial images of the clustered people stored in the database of the facial clustering model; when the similarity analysis result is greater than or equal to the preset similarity threshold, determining that the person corresponding to the facial image data is a clustered person; when the similarity analysis result is less than the preset similarity threshold, defining the person corresponding to the facial image data as a new clustered person and adding his facial image to the database of the facial clustering model.
[0029] Further improved, the first time period is one week, one month or one quarter; the second time period is one week, one month or one quarter; the third time period is one week, one month or one quarter; the fourth time period is one week, one month or one quarter; the fifth time period is one week, one month or one quarter; the sixth time period is one week, one month or one quarter; the seventh time period is one week, one month or one quarter; the eighth time period is one week, one month or one quarter; the ninth time period is one week, one month or one quarter.
[0030] As a further improvement, the first preset time interval is 30 minutes.
[0031] In a further improvement, the facial image data is structured data.
[0032] As a further improvement, the data cleaning strategy preset in step S102 includes: discarding facial image data whose image clarity is lower than the tenth threshold; discarding facial image data of the same traveler that is repeatedly captured by the same device within one second.
[0033] As a further improvement, the analysis strategy in step S104 includes: a crowd analysis strategy and a travel period analysis strategy.
[0034] In a further improvement, the similarity threshold is 0.75.
[0035] The beneficial effects of the present invention are:
[0036] The present invention pre-constructs a four-dimensional travel model based on facial image data of travelers, and through the precipitation of massive data over a long period of time, in multiple places and on multiple devices, inputs the precipitated data into the constructed four-dimensional travel model to generate travel trajectories, finds out the travel trajectories of personnel (target personnel) whose travel behavior is to be analyzed, and determines the abnormal personnel among the target personnel according to pre-set abnormality determination rules. According to the travel trajectories, the close contacts A of the abnormal personnel, the high-frequency travel places A of the abnormal personnel, the high-frequency visitors to the high-frequency travel places A, the high-frequency travel places B of the high-frequency visitors, the personnel who appear in the same frame with the abnormal personnel, the close contacts B of the personnel who appear in the same frame, the high-frequency travel places C of the personnel who appear in the same frame, the close contacts C of the close contacts A, and the high-frequency travel places D of the close contacts A are mined. Finally, according to the mining results, a personnel activity map is generated with the abnormal personnel as the center point.
[0037] If the target person is selected from the caring and concerned persons, then based on the generated personnel activity map, the personnel of government management agencies at all levels can easily know and grasp the travel patterns and abnormal travel behaviors of the caring and concerned persons, as well as whether the caring and concerned persons in the jurisdiction are gang-committing crimes, etc., to achieve early warning of abnormal movement trajectories of caring and concerned persons and early warning of accidents in the jurisdiction, so that government staff can take supervision and preventive measures in advance. This method is intelligent and efficient, saving a lot of time spent by government management departments on the management of caring and concerned persons, and technical defense assists human defense to improve the efficiency of personnel precision management services. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] Figure 1 Flowchart of the method for implementing the personnel activity map. DETAILED DESCRIPTION
[0039] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.
[0040] like Figure 1 As shown, the present invention provides a method for realizing a personnel activity map based on a four-dimensional travel model. The steps of the method for realizing a personnel activity map include the following:
[0041] S1: Construct a four-dimensional travel model of travelers. The specific steps for constructing the four-dimensional travel model are as follows:
[0042] S101: Gathering facial image data of travelers into a travel log of travelers. Facial image data is structured data.
[0043] S102: Clean the personnel travel log flow according to the preset data cleaning strategy. The specific data cleaning strategy includes the following:
[0044] a. Discarding face image data whose image clarity is lower than the tenth threshold. Discarding face image data whose clarity is lower than the tenth threshold reduces invalid data.
[0045] b. Discard the data of the same person captured repeatedly by the same device within one second. You can also set other time intervals according to the specific situation, if the same device captures the person repeatedly, then the repeatedly captured data will be discarded. By discarding the repeatedly captured data, invalid data is reduced.
[0046] S103: Loading type tags to the personnel travel log, where the type tags include personnel type tag X, location tag Z, and equipment tag Y.
[0047] S104: According to the loaded personnel type label X, the personnel structured travel log flow is segmented, and then the segmented personnel structured travel log flow is input into the preset analysis model for training. The analysis model includes four-dimensional analysis channels. According to the analysis strategy configured in the analysis model, the personnel structured travel log flow is input into one or more of the analysis channels, and the analysis model completes the dynamic analysis and classification of the personnel structured travel log flow in parallel. The four dimensions are time dimension, place dimension, equipment dimension and personnel dimension. In this step, according to the personnel type label X loaded on the personnel travel flow log in step S103, the data loaded with different personnel type labels X are separated from each other, and then input into their respective analysis models respectively. The number of analysis models is set according to the number of types of personnel. Different types of personnel travel log flow are input into different analysis models. At the same time, different analysis strategies have been preset according to the characteristics of the personnel of this type in different analysis models, thereby avoiding the mutual interference of different analysis strategies, accelerating the efficiency of data analysis, and greatly reducing the waste of resources when analyzing unnecessary invalid data bodies. The analysis strategies include the following: crowd analysis strategy and travel period analysis strategy. Other analysis strategies can also be preset in the analysis model according to specific needs.
[0048] S105, constructing a four-dimensional travel model of travelers based on the trained analysis model.
[0049] Thus, the four-dimensional travel model is created. The four-dimensional travel model completes the extraction of time information, place information, equipment information and personnel information in the face image data through steps S101 to S105, and generates a four-dimensional travel trajectory according to the extracted time information, place information, equipment information and personnel information according to a differentiated analysis strategy.
[0050] S2: Obtain facial image data of travelers, and input the facial image data into the constructed four-dimensional travel model to generate the travel trajectory of travelers. The facial image data here is a massive amount of sedimentation data over a long period of time, multiple devices and multiple places.
[0051] S3: According to the travel trajectory, the abnormal persons among the target persons are determined according to the set abnormal travel determination rules. The target persons are the persons whose travel behavior needs to be determined among the travel persons. In this embodiment, the target persons correspond to the persons of care and concern. The abnormal travel determination rules include the following:
[0052] If the concerned person frequently appears in a sensitive place and the cumulative number of times exceeds the set threshold, it is determined to be abnormal travel;
[0053] The caring personnel deviate from the previous travel pattern (abnormal travel time period, abnormal travel places), and are determined to be abnormal travel;
[0054] Other abnormal travel determination rules can be set based on specific analysis needs.
[0055] S4: Mining close contacts A who have been in contact with the abnormal person within the first time period, where the close contacts A are persons whose number of contacts with the abnormal person is greater than or equal to a first threshold.
[0056] S5: mining high-frequency travel places A visited by abnormal persons in the second time period, where high-frequency travel places A are places visited by abnormal persons a number of times greater than or equal to a second threshold.
[0057] S6: mining high-frequency visitors to the high-frequency travel place A within the third time period, where the high-frequency visitors are people who visit the place a number of times greater than or equal to a third threshold.
[0058] S7, mining persons who appear in the same frame with the abnormal person in the fourth time period, the persons who appear in the same frame with the abnormal person are persons who appear at the same place, the nearest associated place or the collection area of the same device at the same time as the abnormal person, and the time interval between their appearances is less than or equal to the first preset time interval, and the nearest associated place is a place whose distance from the high-frequency travel place A is less than or equal to the fourth threshold. The first preset time interval is set to 30 minutes, and the fourth threshold is set to five meters or ten meters.
[0059] S8, mining high-frequency travel places B of high-frequency visitors in the fifth time period, where the high-frequency travel places B are places where the number of times the high-frequency visitors visit the places is greater than or equal to a fifth threshold.
[0060] S9, mining high-frequency travel places C of the persons appearing in the same frame within the sixth time period, the high-frequency travel places C being places visited by the persons appearing in the same frame for a number of times greater than or equal to a sixth threshold.
[0061] S10: mining close contact person B of the person who appears in the same frame within the seventh time period, where close contact person B is a person who has had contact with the person who appears in the same frame for a number of times greater than or equal to a seventh threshold.
[0062] S11, mining close contacts C of close contact A within the eighth time period, where close contact C is a person who has come into contact with close contact A for a number of times greater than or equal to an eighth threshold.
[0063] S12, mining the high-frequency travel places D of the close contact A in the ninth time period, the high-frequency travel places D being places where the number of visits by the close contact A is greater than or equal to the ninth threshold.
[0064] S13: A personnel tracking chain is formed based on the obtained close contacts A, high-frequency travel places A, high-frequency visitors, high-frequency travel places B, people appearing in the same frame, high-frequency travel places C, close contacts B, close contacts C, and high-frequency travel places D, and a personnel activity map is generated with the abnormal persons as the center point.
[0065] Before step S101, the following steps are also included:
[0066] S001: Input the image data of the traveler into the pre-trained face recognition model to screen the face images, and match the screened face images with the face pictures stored in the face recognition model; if the match is successful, mark the face image data of the traveler with an identity ID; if the match fails, proceed to step S002;
[0067] S002: Input the facial image data of the travelers into a pre-trained facial clustering model for personnel clustering, and after clustering, mark the facial image data of the travelers with an identity recognition ID.
[0068] The specific steps of inputting the facial image data of the travelers into the pre-trained facial clustering model for clustering the travelers in step S002 include: performing similarity analysis on the facial image data of the travelers and the facial images of clustered people stored in the database of the facial clustering model; when the similarity analysis result is greater than or equal to the preset similarity threshold, determining that the person corresponding to the facial image data is a clustered person; when the similarity analysis result is less than the preset similarity threshold, defining the person corresponding to the facial image data as a new clustered person and adding his facial image to the database of the facial clustering model. The similarity threshold is selected as 0.75.
[0069] Preferably, the first time period is one week, one month or one quarter, the second time period is one week, one month or one quarter, the third time period is one week, one month or one quarter, the fourth time period is one week, one month or one quarter, the fifth time period is one week, one month or one quarter, the sixth time period is one week, one month or one quarter, the seventh time period is one week, one month or one quarter, the eighth time period is one week, one month or one quarter, and the ninth time period is one week, one month or one quarter. These time periods can be changed according to specific research and judgment requirements.
[0070] The present invention has the following innovative features: knowing the travel patterns of people who are of concern within the jurisdiction, and mastering the travel tracking chain of the people.
[0071] The present invention is reasonable and efficient, which greatly saves the time spent by government management departments on the management of caring and concerned personnel. Technical defense assists human defense and improves the efficiency of precise personnel management services.
[0072] The above is only a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the form disclosed herein, and should not be regarded as excluding other embodiments, but can be used in various other combinations, modifications and environments, and can be modified within the scope of the concept described herein through the above teachings or the technology or knowledge of the relevant field. The changes and modifications made by those skilled in the art shall not deviate from the spirit and scope of the present invention, and shall be within the scope of protection of the claims attached to the present invention.
[0073] The processing of user data in the above embodiments should comply with relevant provisions of laws and regulations.
Claims
1. A method for realizing a personnel activity graph based on a four-dimensional travel model, characterized in that: The following steps are involved: S1: Construct a four-dimensional travel model of travelers; In step S1, the specific steps of constructing the four-dimensional travel model of travelers include: S101: Gather historical facial image data of travelers to obtain a flow of travel logs; S102: Cleaning the personnel travel log flow according to a preset data cleaning strategy; S103: Loading type tags to the personnel travel log flow, where the type tags include personnel type tag X, location tag Z, and equipment tag Y; S104: Segment the personnel travel log flow according to the loaded personnel type label X, and then input the segmented personnel travel log flow into a preset analysis model for training. The analysis model includes four-dimensional analysis channels. According to the analysis strategy configured in the analysis model, the personnel travel log flow is input into one or more of the analysis channels. The analysis model completes the analysis and classification of the personnel travel log flow in parallel. The four dimensions are time dimension, place dimension, equipment dimension and personnel dimension. S105, constructing a four-dimensional travel model of travelers according to the trained analysis model; The step S101, before aggregating the historical facial image data of the travelers to obtain the travel log flow of the travelers, also includes the following steps: S001: Input the image data of the traveler into the pre-trained face recognition model to screen the face images, and match the screened face images with the face pictures stored in the face recognition model; if the match is successful, mark the face image data of the traveler with an identity ID; if the match fails, proceed to step S002; S002: inputting the facial image data of the travelers into a pre-trained facial clustering model to cluster the travelers, and after clustering, marking the facial image data of the travelers with an identity ID; In the step S002, the specific steps of inputting the facial image data of the travelers into the pre-trained facial clustering model for clustering the travelers include: performing similarity analysis on the facial image data of the travelers and the facial images of clustered people stored in the database of the facial clustering model; when the similarity analysis result is greater than or equal to a preset similarity threshold, determining that the person corresponding to the facial image data is a clustered person; when the similarity analysis result is less than the preset similarity threshold, defining the person corresponding to the facial image data as a new clustered person and adding his facial image to the database of the facial clustering model; S2: Acquire facial image data of travelers, and input the facial image data into a four-dimensional travel model, the four-dimensional travel model extracts time information, location information, device information and personnel information from the facial image data, and generates a travel trajectory of the travelers based on the time information, location information, device information and personnel information; S3: According to the travel trajectory, the abnormal persons among the target persons are determined according to the set abnormal travel determination rules. The target persons are the persons whose travel behavior needs to be analyzed among the travel persons; S4: mining close contacts A who have contact with the abnormal person within the first time period, where the close contacts A are persons whose number of contact times with the abnormal person is greater than or equal to a first threshold; S5: mining high-frequency travel places A visited by the abnormal person in the second time period, where the high-frequency travel places A are places visited by the abnormal person a number of times greater than or equal to a second threshold; S6: mining high-frequency visitors to the high-frequency travel place A within a third time period, where the high-frequency visitors are people who have visited the place a number of times greater than or equal to a third threshold; S7, mining persons who appear in the same frame with the abnormal person in the fourth time period, where the persons who appear in the same frame with the abnormal person are persons who appear at the same place, nearby associated place, or collection area of the same device at the same time as the abnormal person, and the time interval between their appearances is less than or equal to the first preset time interval, and the nearby associated place is a place whose distance from the high-frequency travel place A is less than or equal to the fourth threshold; S8, mining high-frequency travel places B of high-frequency visitors in the fifth time period, where the high-frequency travel places B are places where the number of visits by high-frequency visitors is greater than or equal to a fifth threshold; S9, mining high-frequency travel places C of the persons who appear in the same frame within the sixth time period, where the high-frequency travel places C are places where the number of visits of the persons who appear in the same frame is greater than or equal to a sixth threshold; S10: mining close contact person B of the person who appears in the same frame within the seventh time period, where the close contact person B is a person who has had contact with the person who appears in the same frame for a number of times greater than or equal to a seventh threshold; S11, mining close contacts C of close contact A within an eighth time period, where close contact C is a person who has had contact with close contact A for a number of times greater than or equal to an eighth threshold; S12, mining high-frequency travel places D of close contact person A in the ninth time period, where the high-frequency travel places D are places where the number of visits by close contact person A is greater than or equal to a ninth threshold; S13: A personnel tracking chain is formed based on the obtained close contacts A, high-frequency travel places A, high-frequency visitors, high-frequency travel places B, people appearing in the same frame, high-frequency travel places C, close contacts B, close contacts C, and high-frequency travel places D, and a personnel activity map is generated with the abnormal persons as the center point.
2. The method for realizing a personnel activity graph based on a four-dimensional travel model according to claim 1, characterized in that: The first time period is one week, one month or one quarter; the second time period is one week, one month or one quarter; the third time period is one week, one month or one quarter; the fourth time period is one week, one month or one quarter; the fifth time period is one week, one month or one quarter; the sixth time period is one week, one month or one quarter; the seventh time period is one week, one month or one quarter; the eighth time period is one week, one month or one quarter; the ninth time period is one week, one month or one quarter.
3. The method for realizing a personnel activity graph based on a four-dimensional travel model according to claim 1, characterized in that: The first preset time interval is 30 minutes.
4. The method for realizing a personnel activity graph based on a four-dimensional travel model according to claim 1, characterized in that: The facial image data is structured data.
5. The method for realizing a personnel activity graph based on a four-dimensional travel model according to claim 1, characterized in that: The data cleaning strategy preset in step S102 includes: discarding facial image data whose image clarity is lower than the tenth threshold; discarding facial image data of the same traveler that is repeatedly captured by the same device within one second.
6. The method for realizing a personnel activity graph based on a four-dimensional travel model according to claim 1, characterized in that: The analysis strategies in step S104 include: crowd gathering analysis strategy and travel period analysis strategy.
7. The method for realizing a personnel activity graph based on a four-dimensional travel model according to claim 1, characterized in that: The similarity threshold is 0.75.
Citation Information
Patent Citations
Transportation transfer station recognition method based on wavelet analysis modulus maximum algorithm
CN104537052A
Identification method of spatial motion behavior semantic pattern
CN108052924A
Abnormal track analysis method based on an adjoint model
CN109657547A
Cited By
A method for constructing a multi-dimensional social group activity monitoring model
CN120278611B
