CN113919994A - Regional population classification management method based on image clustering data - Google Patents
Regional population classification management method based on image clustering data Download PDFInfo
- Publication number
- CN113919994A CN113919994A CN202111170830.5A CN202111170830A CN113919994A CN 113919994 A CN113919994 A CN 113919994A CN 202111170830 A CN202111170830 A CN 202111170830A CN 113919994 A CN113919994 A CN 113919994A
- Authority
- CN
- China
- Prior art keywords
- data
- face
- long
- time
- term
- 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
- 238000007726 management method Methods 0.000 title claims abstract description 22
- 238000000034 method Methods 0.000 claims abstract description 29
- 238000004458 analytical method Methods 0.000 claims abstract description 18
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 7
- 238000004364 calculation method Methods 0.000 claims abstract description 7
- 230000002776 aggregation Effects 0.000 claims abstract description 5
- 238000004220 aggregation Methods 0.000 claims abstract description 5
- 238000007635 classification algorithm Methods 0.000 claims abstract description 4
- 230000004927 fusion Effects 0.000 claims abstract description 4
- 230000007774 longterm Effects 0.000 claims description 29
- 238000012098 association analyses Methods 0.000 claims description 3
- 230000003068 static effect Effects 0.000 claims description 3
- 238000007621 cluster analysis Methods 0.000 claims description 2
- 238000007405 data analysis Methods 0.000 description 4
- 241001622623 Coeliadinae Species 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000012958 reprocessing Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Development Economics (AREA)
- Educational Administration (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a regional personnel classification management method based on image clustering data. The method comprises the following steps of accessing required video stream data, human face and human body snapshot picture data through a data access module; the video stream data, the face and human body snapshot picture data are respectively subjected to face and human body structuralization through a portrait video structuralization and picture structuralization calculation engine, and corresponding face and human body structuralization data are output and obtained; basic information data accessed to all face snapshot devices; calling a face clustering analysis algorithm, performing aggregation association on the same faces through fusion analysis of the location, time, structural characteristic value and structural attribute, outputting face clustering data, and completing the creation and endowing process of video Identity (ID); and constructing a face classification algorithm based on a Hadoop or Spark distributed computing frame. The method reprocesses the cluster collection data, analyzes and excavates the cluster collection data, finely classifies regional population and is convenient to use in the population management process.
Description
Technical Field
The invention relates to the technical field of computers, in particular to a regional population classification management method based on image clustering data.
Background
The prior art relates to the technical scheme of regional population classification management, which includes the following two schemes:
the utility model provides a management work load of current regional real population relies on the public policeman of community in a large number to carry out indefinite "go to" data acquisition, carries out the system through gathering APP and types, is limited to acquisition time, collection frequency and collection scope can't do to instant and full coverage, leads to data acquisition incomplete, data quality is poor, data authenticity is difficult to the scheduling problem, also further leads to public security population management work to be difficult to make the score, and its weak point lies in: 1. the data collection and aggregation seriously depend on manual work, the data is incomplete, and the error rate is high; 2. the data quality problem causes difficulty in classified management of regional personnel data; 3. a verification method for manually collecting and gathering data is lacked; 4. the area population is difficult to be finely managed based on manually collected and aggregated data.
And secondly, carrying out face snapshot on the entering and exiting of the regional personnel through a face recognition system, then carrying out query statistics on snapshot results, and analyzing to obtain the entering and exiting or occurrence frequency of the current snapshot target in the region. The disadvantages are that: 1. for the frequency analysis of regional personnel, the regional personnel can be obtained only by recalculating in massive snapshot data each time through a big data analysis model, and the face characteristic value comparison and calculation need to depend on GPU computing power, so that the GPU resource consumption of the model analysis and calculation is linearly increased and the face comparison precision is linearly reduced along with the gradual increase of the snapshot data quantity; 2. the first-label three-reality real population data collected by the community policemen are not combined for correlation analysis, abnormal data are found out, and the continuous improvement of the data accuracy is promoted.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provides a regional population classification management method based on image clustering data.
The technical scheme of the invention is as follows:
the regional population classification management method based on the image clustering data is characterized by comprising the following steps: comprises that
The method comprises the following steps of accessing required video stream data, human face and human body snapshot picture data through a data access module;
the video stream data, the face and human body snapshot picture data are respectively subjected to face and human body structuralization through a portrait video structuralization and picture structuralization calculation engine, and corresponding face and human body structuralization data are output and obtained;
basic information data of all face snapshot devices are docked;
calling a face clustering analysis algorithm, performing aggregation association on the same faces through fusion analysis of the location, time, structural characteristic value and structural attribute, outputting face clustering data, and completing the creation and endowing process of video Identity (ID);
and constructing a face classification algorithm based on a Hadoop or Spark distributed computing frame, and judging and analyzing by combining snapshot time and time rules.
The face cluster analysis algorithm comprises the following steps:
accessing processed human face and human body structural data, including characteristic values and structural attribute data;
loading basic information data of the accessed face snapshot equipment;
loading the existing full-scale clustering face data;
comparing the face characteristic values, and carrying out face structural attribute association analysis based on snapshot time;
performing space-time relation analysis based on the snapshot time and place of the associated equipment, and removing interference data from the primary clustering data;
and finishing the target clustering data updating.
Judging that the long-term-appearing personnel data appear in a certain area for a period of time and the appearance time is uniform, and obtaining the long-term-appearing personnel data, wherein the long-term-appearing personnel classification realization process comprises the following steps:
acquiring video identity data of a personnel target;
acquiring a video identity data associated snapshot record;
counting the area distribution with the most occurrence and determining the long-term occurrence area;
counting whether the occurrence times are larger than a threshold value or not, and analyzing whether the occurrence time periods are scattered or not;
it is determined whether a person is present for a long time.
Judging that people exist for a long time before but do not exist for a period of time, and obtaining long-term non-existence people data, wherein the long-term non-existence people classification implementation process comprises the following steps:
acquiring video identity data of a personnel target;
acquiring a video identity data associated snapshot record;
judging whether the difference between the latest recording time and the current time is greater than a threshold value;
it is determined whether there is a long-term absence of personnel.
Judging that the temporary personnel data do not appear in a certain area before, and appearing less times in a certain period of time to obtain temporary personnel data, wherein the long-term temporary personnel classification realization process comprises the following steps:
acquiring video identity data of a personnel target;
acquiring a video identity data associated snapshot record;
counting the occurrence times;
judging whether the occurrence frequency is less than a threshold value;
it is determined whether a person is temporarily present.
Judging whether the long-term presence exists in a certain area but the long-term presence does not register in the permanent population, the temporary population and the real population of the area, and obtaining the data of the unregistered persons, wherein the long-term unregistered person data classification implementation process comprises the following steps:
acquiring long-term occurrence personnel data;
performing static large-library comparison retrieval from the permanent/temporary population and the real population;
judging whether a record with the similarity larger than a threshold value exists;
it is determined whether it is an unregistered person.
The invention has the beneficial effects that: the invention carries out clustering collection on the face capture image data to form the face video identity data, and carries out refined classification on the face capture records which are not regular and difficult to apply.
1. Clustering and collecting target identity data which are originally scattered in a mass portrait snapshot record, and further reducing the operation quantity of mass data;
2. through real-time preprocessing analysis, GPU resource consumption during real-time data analysis is effectively reduced, and hardware cost is greatly reduced;
3. the efficiency of data analysis of the real-time clustering track data is far higher than the calculation efficiency in the prior art;
4. the problem of a regional population management data acquisition source is well solved;
5. reprocessing the clustering collected data, analyzing and mining, finely classifying regional population, and facilitating use in population management process;
6. the data acquisition efficiency of the real population in the area is greatly improved;
7. according to the data analysis result of the invention, the existing manual collected data can be reversely calibrated, the collection accuracy of the one-standard three-reality actual population data is supervised and promoted to be continuously improved, and a benign data control closed loop of collection, approval, complementary collection and reauthorization can be formed.
Drawings
FIG. 1 is a flow chart of a method for area population classification management based on image cluster data;
FIG. 2 is a flow chart of a long-term occurrence people classification implementation;
FIG. 3 is a flow chart of an implementation of long-term absence of people classification;
FIG. 4 is a flow chart of an implementation of temporal occurance people classification;
FIG. 5 is a flow chart of an implementation of unregistered people classification.
Detailed Description
For a better understanding of the invention, reference will now be made to the following examples and accompanying drawings.
The regional population classification management method based on the image clustering data is characterized by comprising the following steps: comprises that
S01: the method comprises the following steps of accessing required video stream data, human face and human body snapshot picture data through a data access module;
s02: video stream data, face and human body snapshot picture data are respectively transferred to a portrait video structuralization and picture structuralization calculation engine to carry out face and human body structuralization, and corresponding face and human body structuralization data are output and obtained;
s03: basic information data of all face snapshot devices are docked;
s04: calling a portrait clustering analysis algorithm, performing aggregation association on the same portraits through fusion analysis of the place, the time, the structural characteristic value and the structural attribute, outputting portrait clustering data, and completing one-time portrait clustering analysis;
s05: the creation and the assignment of the video identity ID are completed through a video identity ID management module;
s06: and constructing a portrait classification algorithm based on a Hadoop or Spark distributed computing framework, and carrying out judgment analysis by combining snapshot time and a time rule.
Specifically, the steps of the portrait clustering analysis algorithm are as follows
S51: accessing processed human face and human body structural data, including characteristic values and structural attribute data;
s52: loading basic information data of the accessed face snapshot equipment;
s53: loading the existing full-scale clustering face data;
s54: comparing the face characteristic values, and carrying out face structural attribute association analysis based on snapshot time;
s55: performing space-time relation analysis based on the snapshot time and place of the associated equipment, and removing interference data from the primary clustering data;
s56: and finishing the target clustering data updating.
Step S07 is to determine that the long-term-appearing person data is obtained when the long-term-appearing person data appears in a certain area for a certain period of time and the appearance time is relatively uniform, and the long-term-appearing person classification implementation process includes:
s11: acquiring video identity data of a personnel target;
s12: acquiring a video identity data associated snapshot record;
s13: counting the area distribution with the most occurrence and determining the long-term occurrence area;
s14: counting whether the occurrence times are larger than a threshold value or not, and analyzing whether the occurrence time periods are scattered or not;
s15: it is determined whether a person is present for a long time.
When the statistical occurrence frequency is larger than a threshold value, the analysis shows that the sections are not dispersed, and therefore the people are determined to be in long-term occurrence.
In one embodiment, the step S08, determining that the person is determined to be a long-term person but does not appear for a certain period of time, and obtaining long-term non-occurrence person data, wherein the long-term non-occurrence person classification implementation process includes:
s21: acquiring video identity data of a personnel target;
s22: acquiring a video identity data associated snapshot record;
s23: judging whether the difference between the latest recording time and the current time is greater than a threshold value;
s24: it is determined whether there is a long-term absence of personnel.
And when the difference between the latest recording time and the current time is greater than a threshold value, determining that no personnel exist for a long time.
In an embodiment, step S09 is to determine that the temporary person data does not appear in a certain area before, and appears a small number of times within a certain period of time, where the long-term temporary person classification implementation process includes:
s31: acquiring video identity data of a personnel target;
s32: acquiring a video identity data associated snapshot record;
s33: counting the occurrence times;
s34: judging whether the occurrence frequency is less than a threshold value;
s35: it is determined whether a person is temporarily present.
And counting the occurrence times, and judging that the occurrence times are smaller than a threshold value, thereby determining the temporary occurrence of the personnel.
In one embodiment, the step S10, the long-term unregistered person data classification implementation process includes:
s41: acquiring long-term occurrence personnel data;
s42: performing static large-library comparison retrieval from the permanent/temporary population and the real population;
s43: judging whether a record with the similarity larger than a threshold value exists;
s44: it is determined whether it is an unregistered person.
When the record with the similarity larger than the threshold value is judged, the record is determined as the unregistered person.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, so any modifications, equivalents or improvements made within the spirit and principles of the present invention should be included in the scope of the present invention.
Claims (6)
1. The regional population classification management method based on the image clustering data is characterized by comprising the following steps: comprises that
The method comprises the following steps of accessing required video stream data, human face and human body snapshot picture data through a data access module;
the video stream data, the face and human body snapshot picture data are respectively subjected to face and human body structuralization through a portrait video structuralization and picture structuralization calculation engine, and corresponding face and human body structuralization data are output and obtained;
basic information data of all face snapshot devices are docked;
calling a face clustering analysis algorithm, performing aggregation association on the same faces through fusion analysis of the location, time, structural characteristic value and structural attribute, outputting face clustering data, and completing the creation and endowing process of video Identity (ID);
and constructing a face classification algorithm based on a Hadoop or Spark distributed computing frame, and judging and analyzing by combining snapshot time and time rules.
2. The method for regional population classification management based on image clustering data as claimed in claim 1, wherein: the face cluster analysis algorithm comprises:
accessing processed human face and human body structural data, including characteristic values and structural attribute data;
loading basic information data of the accessed face snapshot equipment;
loading the existing full-scale clustering face data;
comparing the face characteristic values, and carrying out face structural attribute association analysis based on snapshot time;
performing space-time relation analysis based on the snapshot time and place of the associated equipment, and removing interference data from the primary clustering data;
and finishing the target clustering data updating.
3. The method for regional population classification management based on image clustering data as claimed in claim 1, wherein: judging that the long-term-appearing personnel data appear in a certain area for a period of time and the appearance time is uniform, and obtaining the long-term-appearing personnel data, wherein the long-term-appearing personnel classification realization process comprises the following steps:
acquiring video identity data of a personnel target;
acquiring a video identity data associated snapshot record;
counting the area distribution with the most occurrence and determining the long-term occurrence area;
counting whether the occurrence times are larger than a threshold value or not, and analyzing whether the occurrence time periods are scattered or not;
it is determined whether a person is present for a long time.
4. The method for regional population classification management based on image clustering data as claimed in claim 1, wherein: judging that people exist for a long time before but do not exist for a period of time, and obtaining long-term non-existence people data, wherein the long-term non-existence people classification implementation process comprises the following steps:
acquiring video identity data of a personnel target;
acquiring a video identity data associated snapshot record;
judging whether the difference between the latest recording time and the current time is greater than a threshold value;
it is determined whether there is a long-term absence of personnel.
5. The method for regional population classification management based on image clustering data as claimed in claim 1, wherein: judging that the temporary personnel data do not appear in a certain area before, and appearing less times in a certain period of time to obtain temporary personnel data, wherein the long-term temporary personnel classification realization process comprises the following steps:
acquiring video identity data of a personnel target;
acquiring a video identity data associated snapshot record;
counting the occurrence times;
judging whether the occurrence frequency is less than a threshold value;
it is determined whether a person is temporarily present.
6. The method for regional population classification management based on image clustering data as claimed in claim 1, wherein: judging whether the long-term presence exists in a certain area but the long-term presence does not register in the permanent population, the temporary population and the real population of the area, and obtaining the data of the unregistered persons, wherein the long-term unregistered person data classification implementation process comprises the following steps:
acquiring long-term occurrence personnel data;
performing static large-library comparison retrieval from the permanent/temporary population and the real population;
judging whether a record with the similarity larger than a threshold value exists;
it is determined whether it is an unregistered person.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111170830.5A CN113919994A (en) | 2021-10-08 | 2021-10-08 | Regional population classification management method based on image clustering data |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111170830.5A CN113919994A (en) | 2021-10-08 | 2021-10-08 | Regional population classification management method based on image clustering data |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113919994A true CN113919994A (en) | 2022-01-11 |
Family
ID=79238295
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111170830.5A Pending CN113919994A (en) | 2021-10-08 | 2021-10-08 | Regional population classification management method based on image clustering data |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113919994A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114333039A (en) * | 2022-03-03 | 2022-04-12 | 济南博观智能科技有限公司 | Portrait clustering method, device and medium |
-
2021
- 2021-10-08 CN CN202111170830.5A patent/CN113919994A/en active Pending
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114333039A (en) * | 2022-03-03 | 2022-04-12 | 济南博观智能科技有限公司 | Portrait clustering method, device and medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2019214557A1 (en) | Method and system for detecting face image generated by deep network | |
CN107230267B (en) | Intelligence In Baogang Kindergarten based on face recognition algorithms is registered method | |
CN105469105A (en) | Cigarette smoke detection method based on video monitoring | |
CN103246876A (en) | Image feature comparison based counterfeit vehicle registration plate identification method | |
CN112308001A (en) | Data analysis method and personnel tracking method and system for smart community | |
CN104036360A (en) | User data processing system and processing method based on magcard attendance behaviors | |
CN110334602B (en) | People flow statistical method based on convolutional neural network | |
CN109829721B (en) | Online transaction multi-subject behavior modeling method based on heterogeneous network characterization learning | |
CN107465691A (en) | Network attack detection system and detection method based on router log analysis | |
CN113919994A (en) | Regional population classification management method based on image clustering data | |
CN111353338A (en) | Energy efficiency improvement method based on business hall video monitoring | |
CN116384736A (en) | Smart city risk perception method and system | |
CN104077571A (en) | Method for detecting abnormal behavior of throng by adopting single-class serialization model | |
CN110211605A (en) | Smart machine speech sensitivity adjusting method, device, equipment and storage medium | |
CN109271859A (en) | It combines related cases method and apparatus, electronic equipment, computer storage medium | |
CN113343926A (en) | Driver fatigue detection method based on convolutional neural network | |
RU2723900C1 (en) | Method and system for analyzing static objects in a video stream | |
CN111798356A (en) | Rail transit passenger flow abnormal pattern recognition method based on big data | |
CN109194622B (en) | Encrypted flow analysis feature selection method based on feature efficiency | |
CN111800428A (en) | Real-time statistical method and system for digital conference participation | |
CN117011774A (en) | Cluster behavior understanding system based on scene knowledge element | |
CN110597993A (en) | Microblog hot topic data mining method | |
CN111369394B (en) | Scenic spot passenger flow volume statistical evaluation system and method based on big data | |
CN115842645A (en) | UMAP-RF-based network attack traffic detection method and device and readable storage medium | |
CN113919415A (en) | Abnormal group detection method based on unsupervised algorithm |
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 |