WO2021063037A1 - Person database partitioning method, and device - Google Patents

Person database partitioning method, and device Download PDF

Info

Publication number
WO2021063037A1
WO2021063037A1 PCT/CN2020/097668 CN2020097668W WO2021063037A1 WO 2021063037 A1 WO2021063037 A1 WO 2021063037A1 CN 2020097668 W CN2020097668 W CN 2020097668W WO 2021063037 A1 WO2021063037 A1 WO 2021063037A1
Authority
WO
WIPO (PCT)
Prior art keywords
person
database
camera
population
probability
Prior art date
Application number
PCT/CN2020/097668
Other languages
French (fr)
Chinese (zh)
Inventor
陈锦聪
钱苏敏
罗幼泉
Original Assignee
华为技术有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Publication of WO2021063037A1 publication Critical patent/WO2021063037A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/22—Indexing; Data structures therefor; Storage structures
    • G06F16/2282—Tablespace storage structures; Management thereof
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/22—Indexing; Data structures therefor; Storage structures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/29—Geographical information databases
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/172—Classification, e.g. identification

Definitions

  • This application relates to the field of population management, and more specifically, to a method and device for dividing a population database.
  • the city-level intelligent portrait tags the dynamic face in real time according to the sub-database list of the dynamic face and the real name, so as to implement the personal identity information of the dynamic face.
  • a 1:N comparison of dynamic face features and static face feature libraries is performed.
  • a dynamic face is a face that is collected in an unconstrained scene without the user's perception, for example, the face of a passerby captured by a face capture camera, etc.
  • a static face is a situation where the user has perception Bottom, the face collected in a specific constrained scene, for example, the face collected from the ID photo, etc.
  • the city-level static facial feature database is usually in the tens of millions.
  • the traditional method is to divide the static facial feature database according to the administrative division to which the static person belongs, that is, static sub-database. the way.
  • the permanent population is classified according to the administrative division where the household registration is located; for the floating population, the classification is performed according to the administrative division where the registered residence is located.
  • the particle size of the classification can be defined according to the level of administrative division.
  • the domestic administrative regions are divided into four levels: province, city, county (district), township and street.
  • the permanent population sometimes does not live and live in their registered household registration; for the floating population living in the jurisdiction, due to the lag in updating the data of the floating population database, it may not be in the population database sub-database. in.
  • the static sub-database method has a high probability that the actual population living and living in the jurisdiction will not be included in the sub-database of the population database, while the population who do not live and live in the jurisdiction are included in the population database.
  • the sub-Curi is a high probability that the actual population living and living in the jurisdiction will not be included in the sub-database of the population database, while the population who do not live and live in the jurisdiction are included in the population database.
  • the present application provides a method and device for sub-database of a population database, which can perform a more precise sub-database of the population database, thereby improving the matching hit rate of a 1:N comparison based on the sub-database of the population database.
  • the present application provides a method for dividing a population database.
  • the method includes: determining N first cameras in a first area, where N is a positive integer; and determining N according to the N first cameras.
  • Person sets where the probability of a person in the i-th person set in the N person sets appearing in the i-th camera is greater than zero, and the i-th person set in the N person sets is related to the corresponding to i cameras, the value of i is each value in [1, N]; determine the first population sub-database of the first area according to the N sets of people; store the first population sub-database The first population sub-database is used to determine the personal identity information corresponding to the first face data collected by the N first cameras.
  • a person whose appearance probability of the i-th camera is greater than a certain threshold is included in the set of persons corresponding to the i-th camera.
  • the threshold can be zero or its specific value.
  • different time periods may correspond to different population sub-databases.
  • a day can be divided into 1440/T time periods, where T is the duration of each time period, and 1440/T time periods can correspond to 1440/T personal database.
  • the set of persons likely to appear in the first camera is determined, and the population database sub-database corresponding to the area is determined according to the set of persons of each first camera. Since the occurrence probability of personnel can be continuously updated, it can be achieved that the population database sub-database always maintains the most fresh and actual active population in the area. Compared with the static sub-database, invalid personnel data can be greatly reduced.
  • it can increase the hit rate of the first comparison, greatly reduce the second comparison, and achieve the effect of less resource consumption and shorter time-consuming comparison.
  • the first population sub-database only includes persons whose occurrence probability is greater than zero, that is to say, the first population sub-database only includes persons who are likely to appear in the first camera. In this way, the number of the first population sub-database can be reduced. The amount of data can reduce the consumption of comparison resources.
  • the method before determining the first population sub-database of the first area, the method further includes: acquiring second face data captured by M second cameras, where the second camera is the The camera where the person in the i-th person set was located before the area migration; according to the second face data and the person migration probability, determine the probability that the person in the i-th person set appears in the i-th camera ,
  • the personnel migration probability is the probability of a personnel migration from the second camera to the i-th first camera.
  • the probability of occurrence of persons is updated according to the real-time face data from the cameras, that is, the list of persons ID corresponding to each camera can be dynamically updated.
  • the population database sub-database corresponding to each area can be determined, so that the population database sub-database can always maintain the most lively actual active population in the area.
  • determining the probability that a person in the i-th person set appears in the i-th camera according to the second face data and the person migration probability includes: comparing the The similarity between the two face data and the face data in the second population sub-database obtains the confidence level of the second face data, and the confidence level is used to indicate the second face data and the second face data.
  • the confidence probability that the face data in the population sub-database belongs to the same person, the second population sub-database is the population sub-database corresponding to the second camera; the first population sub-database is obtained according to the confidence level and the migration probability of the person The probability that a person in the i person set appears in the i-th camera.
  • the probability of occurrence of the person may be obtained by multiplying the confidence level corresponding to a certain face data and the person migration probability corresponding to the person.
  • the method further includes: determining the personnel according to the historical spatio-temporal trajectory data of each person in the resident population database and/or the floating population database in the first area within a preset time period. The probability of migrating from the second camera to the i-th first camera.
  • the method further includes: determining that a person in the i-th person set appears in the i-th person according to the permanent population database and/or the floating population database of the first region The initial probability of the first camera.
  • the present application provides a population database sub-database device, the device includes: a processing unit configured to determine N first cameras in a first area, where N is a positive integer; the processing unit further It is used to determine N sets of people according to the N first cameras, where the probability that a person in the i-th person set of the N sets of people appears in the i-th camera is greater than zero, and the N The i-th person set in the person set corresponds to the i-th camera, and the value of i is each value in [1, N]; the processing unit is configured to determine according to the N person sets The first population sub-database of the first area; a storage unit for storing the first population sub-database, and the first population sub-database is used to determine the first face collected by the N first cameras The personal identity information corresponding to the data.
  • a person whose appearance probability of the i-th camera is greater than a certain threshold is included in the set of persons corresponding to the i-th camera.
  • the threshold can be zero or its specific value.
  • different time periods may correspond to different population sub-databases.
  • a day can be divided into 1440/T time periods, where T is the duration of each time period, and 1440/T time periods can correspond to 1440/T personal database.
  • the set of persons likely to appear in the first camera is determined, and the population database sub-database corresponding to the area is determined according to the set of persons of each first camera. Since the occurrence probability of personnel can be continuously updated, it can be achieved that the population database sub-database always maintains the most fresh and actual active population in the area. Compared with the static sub-database, invalid personnel data can be greatly reduced.
  • it can increase the hit rate of the first comparison, greatly reduce the second comparison, and achieve the effect of less resource consumption and shorter time-consuming comparison.
  • the first population sub-database only includes persons whose occurrence probability is greater than zero, that is to say, the first population sub-database only includes persons who are likely to appear in the first camera. In this way, the number of the first population sub-database can be reduced. The amount of data can reduce the consumption of comparison resources.
  • the device further includes: an acquiring unit, configured to acquire second face data captured by M second cameras before determining the first population sub-database of the first area, and the first The second camera is the camera where the persons in the i-th person set were before the area migration; the processing unit is further configured to determine the i-th person set according to the second face data and the person migration probability The probability that a person in appears at the i-th camera, and the person migration probability is the probability that a person migrates from the second camera to the i-th first camera.
  • the probability of occurrence of persons is updated according to the real-time face data from the cameras, that is, the list of persons ID corresponding to each camera can be dynamically updated.
  • the population database sub-database corresponding to each area can be determined, so that the population database sub-database can always maintain the most lively actual active population in the area.
  • the processing unit is specifically configured to: compare the similarity between the second face data and the face data in the second population sub-database to obtain the second The confidence of the face data, where the confidence is used to indicate the confidence probability that the second face data and the face data in the second population sub-database belong to the same person, and the second population sub-database is the first The population sub-database corresponding to the two cameras; according to the confidence level and the probability of people migration, the probability that a person in the i-th person set appears at the i-th camera is obtained.
  • the probability of occurrence of the person may be obtained by multiplying the confidence level corresponding to a certain face data and the person migration probability corresponding to the person.
  • the processing unit is further configured to: according to the historical spatiotemporal trajectory data of each person in the resident population database and/or the floating population database in the first area within a preset time period, Determine the probability of a person moving from the second camera to the i-th first camera.
  • the processing unit is further configured to: according to the permanent population database and/or the floating population database in the first area, determine that a person in the i-th person set appears in the first area. The initial probability of the i first camera.
  • the present application provides a chip, which is connected to a memory, and is used to read and execute a software program stored in the memory to implement the first aspect or any one of the implementation manners of the first aspect Methods.
  • the present application provides a population database sub-database device, including a memory for storing programs; a processor for executing the programs stored in the memory, and when the programs stored in the memory are executed, the The processor is configured to execute the first aspect and the method in any one of the possible implementation manners of the first aspect.
  • the population database sub-database device further includes a transceiver.
  • the population database sub-database device is a chip that can be applied to network equipment.
  • the population database sub-database device is a server, a cloud host, or a container.
  • the present application provides a computer program product.
  • the computer program product includes computer instructions. When the computer instructions are executed, the foregoing first aspect or any possible implementation of the first aspect is executed.
  • the present application provides a computer-readable storage medium that stores computer instructions.
  • the computer instructions When the computer instructions are executed, the foregoing first aspect or any possible implementation of the first aspect The method is executed.
  • Figure 1 is a schematic diagram of the static sub-database of the population database.
  • Figure 2 is a comparison diagram of the population in the population sub-database obtained by static sub-database and the actual active population.
  • Fig. 3 is a schematic flowchart of a method for sub-database of a population database provided by an embodiment of the present application.
  • FIG. 4 is a schematic flowchart of a method for accurate population segmentation according to an embodiment of the present application.
  • Fig. 5 is a schematic diagram of a process flow for determining the initial probability of population appearance according to an embodiment of the present application.
  • Fig. 6 is an example of the initial appearance probability of a person in an embodiment of the present application.
  • FIG. 7 is a schematic diagram of a processing flow for determining a probability matrix of personnel spatiotemporal migration according to an embodiment of the present application.
  • FIG. 8 is an example of the probability distribution matrix of personnel transfer in an embodiment of the present application.
  • FIG. 9 is a schematic diagram of a processing flow for updating the occurrence probability of personnel in an embodiment of the present application.
  • FIG. 10 is an example of the appearance probability of update personnel in an embodiment of the present application.
  • Fig. 11 is a schematic diagram of a secondary population database sub-database according to an embodiment of the present application.
  • FIG. 12 is a schematic flowchart of personnel comparison based on the population sub-database of an embodiment of the present application.
  • FIG. 13 is a schematic flowchart of a search and comparison according to an embodiment of the present application.
  • Fig. 14 is an implementation form of the system of the embodiment of the present application.
  • Fig. 15 is a schematic structural diagram of a population database dividing device according to an embodiment of the present application.
  • FIG. 16 is a schematic structural diagram of a population database sub-database device provided by another embodiment of the present application.
  • the technical solutions of the embodiments of the present application can be applied to various scenarios, as long as the scenario requires the sub-database of the population database.
  • scenario requires the sub-database of the population database.
  • city-level smart portrait scenes public security population management scenes, intelligent traffic scenes, etc.
  • the following uses a city-level smart portrait scene as an example to describe the technical solution of the present application.
  • the city-level intelligent portrait tags the dynamic face in real time according to the sub-database list of the dynamic face and the real name, so as to implement the personal identity information of the dynamic face.
  • a 1:N comparison of dynamic face features and static face feature libraries is performed.
  • the face 1:N comparison refers to a face and a set comparison (including N comparison objects), and the query obtains a face with a high similarity to the specified face.
  • a dynamic face is a face that is collected in an unconstrained scene without the user's perception, for example, the face of a passerby collected by a camera that captures the face of the face.
  • a static face is a face collected in a specific constrained scene when the user is aware, for example, a face collected from a passport photo.
  • the face feature is a feature vector generated by mapping the pixels in the close-up image of the face.
  • a close-up face picture is a picture of a face close-up area that meets the pixel requirements of face recognition from a face scene picture, commonly known as a small face picture.
  • a face scene image is a captured image that contains at least one face and human elements, commonly known as a large face image.
  • the traditional method is to divide the population database according to the administrative division to which the static personnel belong, that is, the static database method.
  • Figure 1 is a schematic diagram of the static sub-database of the population database.
  • the permanent population is classified according to the administrative division where the household registration is located; for the floating population, the classification is performed according to the administrative division where the registered residence is located.
  • the particle size of the classification can be defined according to the level of administrative division.
  • the domestic administrative regions are divided into four levels: province, city, county (district), township and street.
  • the permanent population sometimes does not live and live in their registered household registration; for the floating population living in the jurisdiction, due to the lag in updating the data of the floating population database, it may not be in the population database sub-database. in. That is to say, the static sub-database method shown in Figure 2 has a high probability of causing the actual population living and living in the jurisdiction to not be included in the sub-database of the population database, but not the population living and living in the jurisdiction. , But included in the sub-library of the population bank, making the sub-library of the population bank lack fresh activity.
  • the embodiments of the present application provide a method and device for sub-database of a population database, which can perform a more accurate sub-database on the population database, thereby improving a 1:N matching hit rate based on the sub-database of the population database.
  • Fig. 3 is a schematic flowchart of a method for sub-database of a population database provided by an embodiment of the present application.
  • the method shown in FIG. 3 may be executed by a server, a cloud host, a container, etc., or may be executed by a chip or module included in the server, a cloud host, or a container.
  • the method shown in FIG. 3 includes at least part of the following content.
  • N determine N first cameras in the first area, and N is a positive integer.
  • the first area may be an area of any size, which is not specifically limited in the embodiment of the present application.
  • it may be administrative divisions such as provinces, cities, counties, districts, and streets.
  • it may be an area including a preset number of cameras.
  • the camera may also be another device or device with a photographing function, which is not specifically limited in the embodiment of the present application.
  • the N first cameras in the first area are all cameras deployed or set in the first area.
  • At least one camera list can be saved.
  • the N first cameras in the first area it can be determined according to the identity (ID) of the first area, etc.
  • the camera list may be determined according to the geographic location of the camera and the ID of the camera, and the geographic location of the camera may refer to the longitude and latitude where the camera is located.
  • the area where each first camera is located can be determined directly according to the geographic location of the first camera and the ID of the first camera, so as to determine the N first cameras in the first area.
  • the geographic location of the camera may refer to the longitude and latitude where the camera is located.
  • 320 determine N sets of people according to the N first cameras, wherein the probability that a person in the i-th person set of the N sets of people appears in the i-th camera is greater than zero (hereinafter referred to as Person appearance probability), the i-th person set in the N person sets corresponds to the i-th camera, and the value of i is each value in [1, N].
  • a first population sub-database of the first area is determined according to the N person sets.
  • the first population sub-database is stored.
  • the N person sets corresponding to the N first cameras are determined, and the first person set is further determined according to the N person sets.
  • the population sub-database of 375 is composed of some or all of the 375 people.
  • the population sub-database in this embodiment of the present application may be divided according to time slices. Specifically, according to the probability that a person in the N person sets in the first time period appears in the corresponding first camera, determine the N person sets corresponding to the N first cameras in the first time period, and further according to N A collection of persons determines the population sub-database of the first area in the first time period; according to the probability that the persons in the N person sets in the second time period appear in the corresponding first camera, determine the N in the second time period The N personnel sets corresponding to the first cameras are further determined according to the N personnel sets to determine the population sub-database of the first area in the second time period.
  • the first area may correspond to different population sub-databases in different time periods.
  • the time period in this embodiment of the present application may be a time slice divided according to a preset time length. For example, if the time slice length is T, 1440 minutes a day can be divided into 1440/T time slices, and T can be divisible by 1440. In this example, the division of the time slice is divided into a period of days and a unit of minutes. It is understandable that the division of the time slice in the embodiment of the present application may also be divided into periods of other granularities and/or units of other granularities. of.
  • the probability of occurrence of the above-mentioned personnel may also be determined first.
  • the M second cameras capture the face data of a person in the person set corresponding to the i-th first camera, it is triggered to update the occurrence probability of the person in the i-th person set.
  • the second camera is the camera where the personnel in the personnel set corresponding to the first camera are located before the area migration.
  • the probability that a person in the i-th person set appears in the i-th camera is determined according to the face data and the person migration probability, where The probability of personnel migration is the probability of personnel migration from the second camera to the i-th first camera.
  • the face data is compared in the second population sub-database to obtain the confidence of the face data, and the i-th person is obtained according to the obtained confidence and the probability of personnel migration The probability that a person in the set appears at the i-th camera.
  • the confidence is used to indicate the confidence probability that the second face data and the face data in the second population sub-database belong to the same person, for example, it may be similarity, etc.; the second population sub-database is the second The population sub-database corresponding to the camera.
  • the above-mentioned personnel migration probability may be applied to every personnel in the first area.
  • the above-mentioned personnel migration probability may also be stated for each personnel.
  • the probability of moving from the second camera to the first camera is zero, which means that person A will not move from the second camera to the first camera; for person B, it will move from the second camera to the first camera.
  • the probability of the first camera is not zero, which means that Person B may move from the second camera to the first camera.
  • the probability of persons in the set of persons corresponding to the third camera may also be updated to migrate to the third camera.
  • the specific update method please refer to the update method in the first camera above, which will not be repeated here.
  • the first population sub-database of the first area is determined.
  • people whose appearance probability is not zero can be included in the first population sub-database.
  • the probability of appearing at the second camera and moving from the second camera to the first camera is not Those who are zero are included in the first population sub-bank.
  • a real-time population accurate sub-database can be formed according to the occurrence probability of personnel, the area corresponding to the geographic location of the front-end camera, and the retrieval logic of personnel archives.
  • the second-level sub-library is divided according to the districts, counties and sub-district offices; the first-level sub-libraries are divided according to the prefectures and cities; the total amount of the library includes the full amount of permanent population and floating population, and is generally based on the province or city as the unit.
  • each level of the above-mentioned sub-libraries can be divided into one-level or multi-level precise sub-libraries according to accuracy.
  • the two-level accurate sub-database as an example.
  • the first accurate sub-database includes persons whose appearance probability is not zero
  • the second accurate sub-database includes persons who have never appeared in the area (that is, persons whose appearance probability is 0) and those whose appearance probability is not zero. Zero personnel.
  • the first accurate sub-database can be prioritized for comparison, which further improves the hit rate of a 1:N personnel comparison and reduces secondary comparisons.
  • the probability of personnel transfer may also be determined.
  • the probability of a person migrating from the second camera to the first camera may be determined according to the historical spatiotemporal trajectory data of each person in the N person sets within a preset time period.
  • the historical spatiotemporal trajectory data of all or part of the people in the permanent population database and/or the floating population database in the first area within a preset time period it can be determined that the personnel migrated from the second camera to the Probability of the first camera.
  • the preset time period can be any possible time length, for example, 1 month, 1 week, 3 months, 1 year, and so on.
  • the historical spatio-temporal trajectory data may be the activity trajectory data of each person every day, and the data may include the time captured, the captured camera, and the like.
  • the initial population sub-database can be generated through administrative divisions and real-time population distribution data. Specifically, according to the permanent population database and/or the floating population database in the first area, the initial probability of a person in the i-th person set appearing in the i-th first camera is determined.
  • FIG. 4 is a schematic flowchart of a method for accurately sub-database of population sensing the temporal and spatial patterns of personnel migration according to an embodiment of the present application.
  • the real-name permanent population database and floating population database data are imported. Based on the administrative divisions of the permanent population and the floating population, and the geographic location of each camera, the initial probability of occurrence of personnel is generated. And the initial population sub-database.
  • the input information of the system includes:
  • Camera ID and geographic location of the camera such as longitude and latitude.
  • Personnel information divided by province-city-county administrative region.
  • Personnel information includes information on permanent residents and information on floating population.
  • the processing flow of the initial probability of occurrence of personnel includes:
  • the initial population sub-database when the system was initially launched, because there was no real-time data of personnel, the initial population sub-database was generated through administrative divisions and real-time population distribution data, or the initial population sub-database of each division was generated based on the permanent population database and the floating population database. Library.
  • the initial population sub-database records information such as the person's ID, the person's face data, the spatial trajectory data, and the place of residence.
  • the registered residence of Person A is Division B, and Person A belongs to Division B's population sub-database.
  • a list of personnel IDs for each camera is established.
  • the list of person IDs includes the IDs of persons who may appear in the camera.
  • the list of person IDs may correspond to the person set above.
  • the registered residence of person A is zone B, and person A is included in the list of personnel IDs of cameras located in zone B.
  • the initial appearance probability is assigned to the personnel ID in the initial personnel sub-database, and the appearance probability includes the initial appearance probability in each time slice.
  • the initial appearance probabilities of permanent residents and mobile persons may be different, for example, p% for permanent residents and q% for mobile persons.
  • the values of p and q can also be the same.
  • the probability value in the table shown in Figure 5 represents the probability of a certain person appearing in the corresponding camera. For example, in the time slice (0:00, T), the occurrence probability of the person corresponding to the person ID1 appearing in the camera 1 is p%, the probability that the person corresponding to the person ID3 appears in the camera 1 is q%.
  • the time slice may be a period of time. For example, a day of 1440 minutes can be divided into 1440/T time periods according to the time length T, where each time period can be regarded as a time slice, and the value of T is not limited.
  • the setting method of the initial appearance probability described here may indicate that the permanent residents and floating people in the initial personnel sub-database have a higher probability of appearing in the corresponding area (or the area where the camera is located) than other people.
  • Fig. 6 is an example of the initial appearance probability of a person in an embodiment of the present application.
  • the initial probability data of a person takes a time slice as a period, and the initial probability data of a person in each time slice includes a person ID, a corresponding time slice, a camera ID, and the probability of a person appearing at the camera.
  • the appearance probability of the person corresponding to the person ID1 appearing on the camera 1 is 5%
  • the appearance probability of the person corresponding to the person ID3 appearing on the camera 1 is 10%.
  • the personnel migration probability distribution matrix in each time slice is determined.
  • the historical spatiotemporal trajectory data may be historical spatiotemporal trajectory data within a period of time, for example, historical spatiotemporal trajectory data within 1 month, historical spatiotemporal trajectory data within 1 week, historical spatiotemporal trajectory data within 3 months, Historical time-space estimation data within 1 year, etc.
  • the input information of the system includes historical spatio-temporal trajectory data of personnel.
  • the historical spatiotemporal trajectory data of the personnel may be "one person, one file” historical spatiotemporal trajectory data of the personnel.
  • "one person, one file” refers to the establishment of a personnel file for each person, and the personnel file may include the historical time-space trajectory data of the corresponding personnel and other personal information. For example, obtain the activity trajectory data in the personnel file of each personnel within 1 month.
  • extract information such as personnel ID, capture time, and capture camera ID.
  • the camera is associated to generate the activity track of each person every day.
  • the "one-step" track points of two adjacent cameras in each activity track are extracted.
  • the reason why they are called “one-step” is to describe that people step by step from the scope of one camera into the scope of subsequent cameras.
  • Each entry into the range of a new camera is equivalent to a person "taking a step.”
  • the two cameras are adjacent to each other.
  • the "one-step" track points of the trajectory are (camera 1, camera 2) and (camera 2, camera 3). That is to say, the person starts from the shooting range of the camera 1, the next one enters the shooting range of the camera 2, and then the shooting range of the camera 3 appears.
  • the probability of personnel migration at each trajectory point can be obtained by statistics.
  • a time slice is used as a unit to generate a probability distribution matrix of personnel migration within a day.
  • FIG. 8 shows an example of the probability distribution matrix of personnel transfer in an embodiment of the present application. Specifically, Figure 8 shows the probability distribution matrix of personnel migration in each time slice. Taking the time slice (1440-T, 1440) as an example, the probability of personnel migration from camera 3 to camera 1 is 1.2%. The probability of migrating to camera N is 3%.
  • the "one-step" personnel migration probability distribution matrix in the embodiment of the present application is static.
  • the probability of personnel migration to each camera is calculated, that is, the appearance of update personnel Probability.
  • the bayonet data includes travel industry data, travel data, etc.
  • the input information of the system includes:
  • Real-time captured face data for example, person ID, capture time, captured camera ID, matching confidence, etc.
  • real-time bayonet data for example, person ID, appearance time, appearance location, etc.
  • the confidence level may be the similarity obtained by the person performing a 1:N comparison in the population sub-database corresponding to the current camera. In other words, the probability that the person is in the population database corresponding to the current camera.
  • FIG. 10 shows an example of the occurrence probability of update personnel in an embodiment of the present application.
  • the personnel IDs in the personnel appearance probability list include real-name archive activity personnel and personnel who moved from the previous jump in the previous time slice to this time slice (for example, it can be determined according to the personnel migration probability distribution matrix ).
  • the probability of a person appearing in the other camera is set to 0.
  • the person ID1 and the person ID3 in FIG. 10 are confirmed to appear in the camera 3, and it is impossible to appear in the camera 1 within T time, so the probability of the person ID1 and the person ID3 appearing in the camera 1 is set to 0.
  • the personnel ID1 and the personnel ID2 are the permanent residents of the real-name archive personnel, so the personnel ID1 and the personnel ID2 have an initial appearance probability of 10%.
  • Personnel ID3 and Personnel ID4 are the floating population among real-name filing personnel, so Personnel ID3 and Personnel ID4 have an initial appearance probability of 15%.
  • the personnel ID list (ie, the personnel set) of the camera 1 in the time slice (mT, mT+1) includes personnel ID2, personnel ID4, personnel ID5, personnel ID6, personnel ID7, and personnel ID8.
  • the update of the person appearance probability in the embodiment of the present application is triggered by the face data or bayonet data captured in real time.
  • the corresponding person ID list of each camera should include the person ID and facial feature information that may be captured by the camera in each time slice. At the same time, it is also necessary to avoid including the entire population database personnel information in the ID list, which will degenerate into a traversal comparison of the entire population database.
  • a real-time population accurate sub-database can be formed according to the updated probability of occurrence of personnel in 430, the division corresponding to the geographic location of the front-end camera, and the retrieval logic of personnel archives. Specifically, the person ID that may appear in the current time slice of each camera is added to the corresponding sub-database.
  • the face base database of the current time slice sub-database may be formed based on the person IDs that may appear in the current time slice of each current camera, and feature values are extracted based on the sub-database face images to form an accurate sub-database feature database.
  • Fig. 11 is a schematic diagram of a secondary population database sub-database according to an embodiment of the present application. According to the geographical location of the camera combined with the administrative area to divide. For example, divide the secondary sub-library according to districts, counties and sub-district offices; divide the primary sub-library according to prefectures and cities; the total volume includes all permanent residents and floating population, generally based on provinces or prefectures and cities.
  • each level of sub-library can be divided into one-level or multi-level precise sub-libraries according to accuracy.
  • the two-level accurate sub-database includes persons whose appearance probability is not zero
  • the second accurate sub-database includes persons who have never appeared in the area (that is, persons whose appearance probability is 0) and those whose appearance probability is not zero. Zero personnel.
  • the first accurate sub-database can be prioritized for comparison, which further improves the hit rate of a 1:N personnel comparison and reduces secondary comparisons.
  • the update of the sub-library list in each time slice needs to be completed in a relatively short time. For example, it takes less than 0.1T.
  • a 1:N comparison of persons based on the population sub-database obtained in 440 may also be performed.
  • FIG. 12 is a schematic flowchart of personnel comparison based on the population sub-database of an embodiment of the present application.
  • the front-end camera After the front-end camera captures the face data, it associates the corresponding population sub-database according to the geographic location of the captured camera, and performs a 1:N comparison in the population sub-database to obtain the person with the TOP1 similarity to the captured face, and Obtain the real-name tag information of the person, that is, obtain the real-name tag information of the captured person.
  • the population sub-database is the accurate sub-database of the population obtained in 440.
  • FIG. 13 is a schematic flowchart of a search and comparison according to an embodiment of the present application.
  • Slice at the current time After the front-end camera captures the face data, it will first compare it according to the second-level sub-database corresponding to the area where the camera is located. If the comparison fails, it will further compare the first-level sub-database corresponding to the area where the camera is located. Yes, if it still misses, it will be further compared in the full library.
  • the retrieval logic shown in Fig. 13 adopts accurate sub-database grading comparison, which can improve the accuracy and efficiency of the comparison, and increase the hit rate of personnel in the sub-database.
  • Fig. 14 is an implementation form of the system of the embodiment of the present application.
  • Figure 14 takes the automatic archiving of real-time face capture data of the video surveillance intelligent analysis system as an example.
  • the active population database based on real-name archiving (corresponding to the historical temporal and spatial migration data described above) and personnel flow record data , Combined with the real-name permanent population database and floating population database of the administrative area to generate real-time regional personnel dynamic and accurate sub-database; after importing the video surveillance intelligent analysis system, it is processed by the portrait algorithm to form the population sub-database and face feature value database; front-end intelligence
  • the face capture data of the camera (for example, the small face image and the corresponding large image of the scene) are uploaded to the video surveillance intelligent analysis system.
  • the video surveillance intelligent analysis system calls the face algorithm for feature extraction and combines the specific location of the front-end smart camera. Map the real-time sub-database of the corresponding area, perform a 1:N face comparison, and push the comparison result to the personnel archive for archiving.
  • the technical solution described above is to dynamically update the list of personnel IDs corresponding to each camera based on the real-time face data from the cameras, and further determine the population database sub-database corresponding to each area according to the area where the camera is located, so as to realize the population database
  • the sub-database always maintains the most fresh and actual active population in the area. Compared with the full static database, it greatly reduces invalid personnel data.
  • it can improve the hits of a comparison Rate, greatly reduce the secondary comparison, and achieve the effect of less resource consumption and shorter time-consuming comparison.
  • the population database sub-database method provided by the embodiments of the present application is described in detail above with reference to FIGS. 1 to 14.
  • the device embodiments of the present application will be described in detail below with reference to FIGS. 15 and 16. It should be understood that the devices in the embodiments of the present application can execute the various methods of the foregoing embodiments of the present application, that is, for the specific working processes of the following various products, reference may be made to the corresponding processes in the foregoing method embodiments.
  • Fig. 15 is a schematic structural diagram of a population database dividing device according to an embodiment of the present application.
  • the device 1500 shown in FIG. 15 is only an example, and the device in the embodiment of the present application may further include other modules or units.
  • the device 1500 includes a processing unit 1520 and a storage unit 1530.
  • the processing unit 1520 is configured to determine N first cameras in the first area, where N is a positive integer.
  • the processing unit 1520 is further configured to determine N sets of people according to the N first cameras, where the probability that a person in the i-th person set of the N-th person sets appears in the i-th camera is greater than Zero, the i-th person set in the N person sets corresponds to the i-th camera, and the value of i is each value in [1, N].
  • the processing unit 1520 is further configured to determine the first population sub-database of the first area according to the N person sets.
  • the storage unit 1530 is configured to store the first population sub-database, and the first population sub-database is used to determine the personal identity information corresponding to the first face data collected by the N first cameras.
  • a person whose appearance probability of the i-th camera is greater than a certain threshold is included in the set of persons corresponding to the i-th camera.
  • the threshold can be zero or its specific value.
  • different time periods may correspond to different population sub-databases.
  • a day can be divided into 1440/T time periods, where T is the duration of each time period, and 1440/T time periods can correspond to 1440/T personal database.
  • the set of persons likely to appear in the first camera is determined, and the population database sub-database corresponding to the area is determined according to the set of persons of each first camera. Since the occurrence probability of personnel can be continuously updated, it can be achieved that the population database sub-database always maintains the most fresh and actual active population in the area. Compared with the static sub-database, invalid personnel data can be greatly reduced.
  • it can increase the hit rate of the first comparison, greatly reduce the second comparison, and achieve the effect of less resource consumption and shorter time-consuming comparison.
  • the first population sub-database only includes persons whose occurrence probability is greater than zero, that is to say, the first population sub-database only includes persons who are likely to appear in the first camera. In this way, the number of the first population sub-database can be reduced. The amount of data can reduce the consumption of comparison resources.
  • the device further includes: an acquiring unit 1510, configured to acquire second face data captured by M second cameras before determining the first population sub-database of the first area, where the second camera is Describe the camera where the person in the i-th person set was located before the area relocation.
  • the processing unit 1520 is further configured to determine the probability that a person in the i-th person set appears in the i-th camera according to the second face data and the person migration probability, and the person migration probability is The probability of a person moving from the second camera to the i-th first camera.
  • the probability of occurrence of persons is updated according to the real-time face data from the cameras, that is, the list of persons ID corresponding to each camera can be dynamically updated.
  • the population database sub-database corresponding to each area can be determined, so that the population database sub-database can always maintain the most lively actual active population in the area.
  • the processing unit 1520 is specifically configured to: compare the similarity between the second face data and the face data in the second population sub-database to obtain the confidence of the second face data,
  • the confidence level is used to indicate the confidence probability that the second face data and the face data in the second population sub-database belong to the same person, and the second population sub-database is the population sub-database corresponding to the second camera ; According to the confidence level and the personnel migration probability, the probability that a person in the i-th person set appears in the i-th camera is obtained.
  • the probability of occurrence of the person may be obtained by multiplying the confidence level corresponding to a certain face data and the person migration probability corresponding to the person.
  • the processing unit 1520 is further configured to: according to the historical spatio-temporal trajectory data of each person in the resident population database and/or the floating population database in the first area within a preset time period, determine the personnel's travel The probability that the second camera migrates to the i-th first camera.
  • the processing unit 1520 is further configured to: according to the permanent population database and/or the floating population database of the first region, determine that a person in the i-th person set appears in the i-th first person. The initial probability of the camera.
  • the acquiring unit 1510 may be implemented by a transceiver.
  • the processing unit 1520 may be implemented by a processor.
  • the storage unit 1530 may be realized by a memory.
  • FIG. 16 is a schematic structural diagram of a population database sub-database device provided by another embodiment of the present application.
  • the apparatus 1600 shown in FIG. 16 includes a memory 1601, a processor 1602, a communication interface 1603, and a bus 1604.
  • the memory 1601, the processor 1602, and the communication interface 1603 implement communication connections between each other through the bus 1604.
  • the processor 1602 is configured to determine N first cameras in the first area, where N is a positive integer; and determine N sets of people according to the N first cameras, where the first set of N persons The probability that a person in the i person set appears at the i-th camera is greater than zero, the i-th person set in the N person sets corresponds to the i-th camera, and the value of i is [1, N] Each value in; according to the N person sets, determine the first population sub-database of the first area.
  • the memory 1601 is configured to store the first population sub-database, and the first population sub-database is used to determine the personal identity information corresponding to the first face data collected by the N first cameras.
  • the memory 1601 may be a read only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM).
  • the memory 1601 may store a program.
  • the processor 1602 is configured to execute each step of the method for sub-database of the population database in the embodiment of the present application. For example, the method embodiment in FIG. 3 may be executed. The various steps.
  • the processor 1602 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or instructions in the form of software.
  • the processor described in each embodiment of the present application may be a general-purpose processor, a digital signal processor (digital signal processor, DSP), an application specific integrated circuit (ASIC), and a field programmable gate array (field programmable gate array). , FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components.
  • DSP digital signal processor
  • ASIC application specific integrated circuit
  • FPGA field programmable gate array
  • the methods, steps, and logical block diagrams disclosed in the embodiments of the present application can be implemented or executed.
  • the general-purpose processor may be a microprocessor or the processor may also be any conventional processor or the like.
  • the steps of the method disclosed in the embodiments of the present application can be directly embodied as being executed and completed by a hardware decoding processor, or executed and completed by a combination of hardware and software modules in the decoding processor.
  • the software module can be located in random access memory (RAM), flash memory, read-only memory (read-only memory, ROM), programmable read-only memory, or electrically erasable programmable memory, registers, etc. mature in the field Storage medium.
  • the storage medium is located in the memory, and the processor reads the instructions in the memory and completes the steps of the above method in combination with its hardware.
  • the steps of the method disclosed in the embodiments of the present application can be directly embodied as being executed and completed by a hardware decoding processor, or executed and completed by a combination of hardware and software modules in the decoding processor.
  • the software module can be located in a mature storage medium in the field, such as random access memory, flash memory, read-only memory, programmable read-only memory, or electrically erasable programmable memory, registers.
  • the storage medium is located in the memory 1601, and the processor 1602 reads the information in the memory 1601, combines its hardware to complete the functions required by the units included in the population database sub-library device, or executes the methods in the method embodiments of the present application.
  • the communication interface 1603 uses a transceiving device such as but not limited to a transceiver to implement communication between the device 1600 and other devices or a communication network.
  • a transceiving device such as but not limited to a transceiver to implement communication between the device 1600 and other devices or a communication network.
  • the face data captured dynamically by the front-end camera can be obtained through the communication interface 1603.
  • the bus 1604 may include a path for transferring information between various components of the device 1600 (for example, the memory 1601, the processor 1602, and the communication interface 1603).
  • the processor in the device 1600 in FIG. 16 may correspond to the processing unit 1520 in the device 1500 in FIG. 15, the communication interface 1603 may correspond to the acquiring unit 1510, and the memory 1601 may correspond to the storage unit 1530.
  • the population database sub-database device shown in the embodiment of the present application may be a server, for example, it may be a server in the cloud, or may also be a chip configured in a server in the cloud.
  • the population database sub-library device may be an electronic device, or may also be a chip configured in the electronic device.
  • the apparatus 1600 may also include other devices necessary for normal operation.
  • the apparatus 1600 may also include hardware devices that implement other additional functions.
  • the device 1600 may also include only the components necessary to implement the embodiments of the present application, and does not necessarily include all the components shown in FIG. 16.
  • computer-readable media may include, but are not limited to: magnetic storage devices (for example, hard disks, floppy disks, or tapes, etc.), optical disks (for example, compact discs (CD), digital versatile discs (DVD)) Etc.), smart cards and flash memory devices (for example, erasable programmable read-only memory (EPROM), cards, sticks or key drives, etc.).
  • various storage media described herein may represent one or more devices and/or other machine-readable media for storing information.
  • machine-readable medium may include, but is not limited to, wireless channels and various other media capable of storing, containing, and/or carrying instructions and/or data.
  • the size of the sequence number of each process does not mean the order of execution.
  • the execution order of each process should be determined by its function and internal logic, and should not constitute the implementation process of the embodiments of this application. Any restrictions.
  • the computer may be implemented in whole or in part by software, hardware, firmware or any other combination.
  • software it can be implemented in the form of a computer program product in whole or in part.
  • the computer program product includes one or more computer instructions.
  • the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.
  • the computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium.
  • the computer instructions may be transmitted from a website, computer, server, or data center.
  • the computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or a data center integrated with one or more available media.
  • the usable medium may be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a digital video disc (DVD)), or a semiconductor medium (for example, a solid state disk (SSD)), etc.
  • the disclosed system, device, and method may be implemented in other ways.
  • the device embodiments described above are merely illustrative, for example, the division of the units is only a logical function division, and there may be other divisions in actual implementation, for example, multiple units or components may be combined or It can be integrated into another system, or some features can be ignored or not implemented.
  • the displayed or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, devices or units, and may be in electrical, mechanical or other forms.
  • the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the objectives of the solutions of the embodiments.
  • the functional units in the various embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units may be integrated into one unit.
  • the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium.
  • the technical solution of the present application essentially or the part that contributes to the existing technology or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including Several instructions are used to make a computer device (which may be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods described in the various embodiments of the present application.
  • the aforementioned storage media include: U disk, mobile hard disk, read-only memory (read-only memory, ROM), random access memory (random access memory, RAM), magnetic disks or optical disks and other media that can store program codes. .

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Probability & Statistics with Applications (AREA)
  • Operations Research (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Biology (AREA)
  • Remote Sensing (AREA)
  • Algebra (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Health & Medical Sciences (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Human Computer Interaction (AREA)
  • General Health & Medical Sciences (AREA)
  • Multimedia (AREA)
  • Image Analysis (AREA)
  • Collating Specific Patterns (AREA)

Abstract

The present application provides a person database partitioning method and a device. The technical solution of the present application comprises: determining, according to a personnel capture probability of personnel being captured by a first camera, personnel sets that might be captured by the first camera, and determining, according to each of the personnel sets of the first camera, a personnel database partition corresponding to a region. The personnel capture probability can be constantly updated, enabling the people most recently active in the region to be maintained in the person database partitions, thereby significantly reducing inactive personnel data compared to static database partitioning. The invention improves the hit rate of first-time comparison when performing 1:N face identification and real-name labeling, and reduces the need to perform a second comparison operation, thereby achieving low resource consumption and reduced time for comparison.

Description

人口库分库方法和装置Population database dividing method and device 技术领域Technical field
本申请涉及人口管理领域，并且更具体地，涉及人口库分库方法和装置。This application relates to the field of population management, and more specifically, to a method and device for dividing a population database.
背景技术Background technique
城市级智能人像是根据动态人脸和实名的分库名单，对动态人脸进行实时打标签，从而落实动态人脸的人员身份信息。在对动态人脸进行实时打标签的过程中，会对动态人脸特征和静态人脸特征库进行1：N比对。其中，动态人脸是在用户无感知的情况下，在非约束场景中采集到的人脸，例如，人脸抓拍摄像机采集到的路人的人脸等；静态人脸是在用户有感知的情况下，在特定约束场景下采集到的人脸，例如，证件照采集的人脸等。The city-level intelligent portrait tags the dynamic face in real time according to the sub-database list of the dynamic face and the real name, so as to implement the personal identity information of the dynamic face. In the process of real-time tagging of dynamic faces, a 1:N comparison of dynamic face features and static face feature libraries is performed. Among them, a dynamic face is a face that is collected in an unconstrained scene without the user's perception, for example, the face of a passerby captured by a face capture camera, etc.; a static face is a situation where the user has perception Bottom, the face collected in a specific constrained scene, for example, the face collected from the ID photo, etc.
随着静态人脸特征库(或称人口库)规模N的不断增大，会导致1：N比对的精度下降，以及比对消耗的资源和比对耗时的增加。城市级静态人脸特征库通常在千万级规模，为了降低静态人脸特征库规模N的大小，传统做法是按照静态人员所属的行政区划对静态人脸特征库进行分库，即静态分库方式。As the scale N of the static face feature database (or population database) continues to increase, the accuracy of the 1:N comparison will decrease, and the resources consumed and the comparison time will increase. The city-level static facial feature database is usually in the tens of millions. In order to reduce the size of the static facial feature database N, the traditional method is to divide the static facial feature database according to the administrative division to which the static person belongs, that is, static sub-database. the way.
在人口库的静态分库方式中，对于常住人口，按照户籍地所在行政区划来进行分类；对于流动人口，按照登记的居住地所在行政区划进行分类。分类的颗粒粒度可以根据行政区划的层次大小来定义。国内的行政区划分为省、市、县(区)、乡镇街道四级。In the static sub-database method of the population database, the permanent population is classified according to the administrative division where the household registration is located; for the floating population, the classification is performed according to the administrative division where the registered residence is located. The particle size of the classification can be defined according to the level of administrative division. The domestic administrative regions are divided into four levels: province, city, county (district), township and street.
但是，由于人口的流动性，对于常住人口，有时并不在其登记的户籍地生活和居住；对于居住在辖区内的流动人口，由于流动人口库数据更新滞后性，有可能并不在人口库分库中。也就是说，静态分库方式有很大概率会导致实际生活和居住在本辖区的人口，不包含在人口库的分库里面，而没生活和居住在本辖区的人口，却包含在人口库的分库里。这样，会导致基于人口库的分库的一次1：N比对匹配命中率低，而且即便采用一定比对步骤和逻辑，例如，一次比对在本级辖区分库完成，二次比对在上级辖区的分库完成，也会带来比对资源消耗大、比对耗时长等问题。However, due to the mobility of the population, the permanent population sometimes does not live and live in their registered household registration; for the floating population living in the jurisdiction, due to the lag in updating the data of the floating population database, it may not be in the population database sub-database. in. In other words, the static sub-database method has a high probability that the actual population living and living in the jurisdiction will not be included in the sub-database of the population database, while the population who do not live and live in the jurisdiction are included in the population database. The sub-Curi. In this way, it will lead to a low hit rate of a 1:N comparison match based on the population database, and even if certain comparison steps and logic are adopted, for example, the first comparison is completed in the sub-base under the jurisdiction of the level, and the second comparison is The completion of the sub-database in the higher-level jurisdiction will also bring about problems such as high resource consumption and long comparison time.
发明内容Summary of the invention
本申请提供人口库分库方法和装置，能够对人口库进行更精准地分库，从而提高基于人口库的分库的一次1：N比对匹配命中率。The present application provides a method and device for sub-database of a population database, which can perform a more precise sub-database of the population database, thereby improving the matching hit rate of a 1:N comparison based on the sub-database of the population database.
第一方面，本申请提供了一种人口库分库方法，所述方法包括：确定在第一区域内的N个第一摄像机，N为正整数；根据所述N个第一摄像机，确定N个人员集合，其中，所述N个人员集合中的第i个人员集合中的人员出现在第i个摄像机的概率大于零，所述N个人员集合中的第i个人员集合与所述第i个摄像机对应，i的取值是[1，N]中的每一个值；根据所述N个人员集合，确定所述第一区域的第一人口分库；存储所述第一人口分库，所述第一人口分库用于确定所述N个第一摄像机采集到的第一人脸数据对应的人员身份信息。In the first aspect, the present application provides a method for dividing a population database. The method includes: determining N first cameras in a first area, where N is a positive integer; and determining N according to the N first cameras. Person sets, where the probability of a person in the i-th person set in the N person sets appearing in the i-th camera is greater than zero, and the i-th person set in the N person sets is related to the corresponding to i cameras, the value of i is each value in [1, N]; determine the first population sub-database of the first area according to the N sets of people; store the first population sub-database The first population sub-database is used to determine the personal identity information corresponding to the first face data collected by the N first cameras.
例如，将出现在第i个摄像机的出现概率大于某个阈值的人员，纳入第i个摄像机对应的人员集合。该阈值可以为零或者其特数值。For example, a person whose appearance probability of the i-th camera is greater than a certain threshold is included in the set of persons corresponding to the i-th camera. The threshold can be zero or its specific value.
可选地，在本申请实施例中，不同的时间段可以对应于不同的人口分库。例如，可以把一天分为1440/T个时间段，T为每个时间段的时长，1440/T个时间段可以对应于1440/T个人口分库。Optionally, in this embodiment of the application, different time periods may correspond to different population sub-databases. For example, a day can be divided into 1440/T time periods, where T is the duration of each time period, and 1440/T time periods can correspond to 1440/T personal database.
在上述技术方案中，根据人员出现在第一摄像机的人员出现概率，确定有可能出现在第一摄像机的人员集合，根据各个第一摄像机的人员集合确定该区域对应的人口库分库。 由于人员出现概率可以是不断更新的，因此可以实现人口库分库内始终保持该区域最鲜活的实际活动人口，相比静态分库，可以大大减少无效人员数据。当进行人脸1:N比对实名打标签流程时，能够提升一次比对的命中率，大幅减少二次比对，达到比对资源消耗少、耗时短的效果。In the above technical solution, according to the occurrence probability of the person appearing in the first camera, the set of persons likely to appear in the first camera is determined, and the population database sub-database corresponding to the area is determined according to the set of persons of each first camera. Since the occurrence probability of personnel can be continuously updated, it can be achieved that the population database sub-database always maintains the most fresh and actual active population in the area. Compared with the static sub-database, invalid personnel data can be greatly reduced. When performing a 1:N face comparison real-name tagging process, it can increase the hit rate of the first comparison, greatly reduce the second comparison, and achieve the effect of less resource consumption and shorter time-consuming comparison.
此外，在第一人口分库中仅包含人员出现概率大于零的人员，也就是说，第一人口分库仅包括有可能出现在第一摄像机的人员，这样，可以减少第一人口分库中的数据量，降低对比对资的源消耗。In addition, the first population sub-database only includes persons whose occurrence probability is greater than zero, that is to say, the first population sub-database only includes persons who are likely to appear in the first camera. In this way, the number of the first population sub-database can be reduced. The amount of data can reduce the consumption of comparison resources.
在一种可能的实现方式中，在确定第一区域的第一人口分库之前，所述方法还包括：获取M个第二摄像机拍摄的第二人脸数据，所述第二摄像机为所述第i个人员集合中的人员进行区域迁移之前所在的摄像机；根据所述第二人脸数据和人员迁移概率，确定所述第i个人员集合中的人员出现在所述第i个摄像机的概率，所述人员迁移概率为人员从所述第二摄像机迁移至所述第i个第一摄像机的概率。In a possible implementation manner, before determining the first population sub-database of the first area, the method further includes: acquiring second face data captured by M second cameras, where the second camera is the The camera where the person in the i-th person set was located before the area migration; according to the second face data and the person migration probability, determine the probability that the person in the i-th person set appears in the i-th camera , The personnel migration probability is the probability of a personnel migration from the second camera to the i-th first camera.
在上述技术方案中，根据来自摄像机的实时人脸数据，更新人员出现概率，也就是说，可以动态更新每个摄像机所对应的人员ID列表。这样，根据摄像机所在区域，确定每个区域对应的人口库分库，可以实现人口库分库内始终保持该区域最鲜活的实际活动人口，相比全量静态库，大大减少了无效人员数据，当进行人脸1:N比对实名打标签流程时，能够提升一次比对的命中率，大幅减少二次比对，达到比对资源消耗少、耗时短的效果。In the above technical solution, the probability of occurrence of persons is updated according to the real-time face data from the cameras, that is, the list of persons ID corresponding to each camera can be dynamically updated. In this way, according to the area where the camera is located, the population database sub-database corresponding to each area can be determined, so that the population database sub-database can always maintain the most lively actual active population in the area. Compared with the full static database, it greatly reduces invalid personnel data. When performing a 1:N face comparison real-name tagging process, it can increase the hit rate of the first comparison, greatly reduce the second comparison, and achieve the effect of less resource consumption and shorter time-consuming comparison.
在一种可能的实现方式中，根据所述第二人脸数据和人员迁移概率，确定所述第i个人员集合中的人员出现在所述第i个摄像机的概率，包括：比较所述第二人脸数据与第二人口分库中的人脸数据之间的相似性，得到所述第二人脸数据的置信度，所述置信度用于指示所述第二人脸数据与第二人口分库中的人脸数据属于同一人员的置信概率，所述第二人口分库为所述第二摄像机对应的人口分库；根据所述置信度和所述人员迁移概率，得到所述第i个人员集合中的人员出现在所述第i个摄像机的概率。In a possible implementation manner, determining the probability that a person in the i-th person set appears in the i-th camera according to the second face data and the person migration probability includes: comparing the The similarity between the two face data and the face data in the second population sub-database obtains the confidence level of the second face data, and the confidence level is used to indicate the second face data and the second face data. The confidence probability that the face data in the population sub-database belongs to the same person, the second population sub-database is the population sub-database corresponding to the second camera; the first population sub-database is obtained according to the confidence level and the migration probability of the person The probability that a person in the i person set appears in the i-th camera.
可选地，可以将某个人脸数据对应的置信度与该人员对应的人员迁移概率相乘得到所述人员出现概率。Optionally, the probability of occurrence of the person may be obtained by multiplying the confidence level corresponding to a certain face data and the person migration probability corresponding to the person.
在一种可能的实现方式中，所述方法还包括：根据所述第一区域的常住人口库和/或流动人口库中的每个人员在预设时间段内的历史时空轨迹数据，确定人员从所述第二摄像机迁移至所述第i个第一摄像机的概率。In a possible implementation manner, the method further includes: determining the personnel according to the historical spatio-temporal trajectory data of each person in the resident population database and/or the floating population database in the first area within a preset time period. The probability of migrating from the second camera to the i-th first camera.
在一种可能的实现方式中，所述方法还包括：根据所述第一区域的常住人口库和/或流动人口库，确定所述第i个人员集合中的人员出现在所述第i个第一摄像机的初始概率。In a possible implementation, the method further includes: determining that a person in the i-th person set appears in the i-th person according to the permanent population database and/or the floating population database of the first region The initial probability of the first camera.
第二方面，本申请提供了一种人口库分库装置，所述装置包括：处理单元，用于确定在第一区域内的N个第一摄像机，N为正整数；所述处理单元，还用于根据所述N个第一摄像机，确定N个人员集合，其中，所述N个人员集合中的第i个人员集合中的人员出现在第i个摄像机的概率大于零，所述N个人员集合中的第i个人员集合与所述第i个摄像机对应，i的取值是[1，N]中的每一个值；所述处理单元，用于根据所述N个人员集合，确定所述第一区域的第一人口分库；存储单元，用于存储所述第一人口分库，所述第一人口分库用于确定所述N个第一摄像机采集到的第一人脸数据对应的人员身份信息。In a second aspect, the present application provides a population database sub-database device, the device includes: a processing unit configured to determine N first cameras in a first area, where N is a positive integer; the processing unit further It is used to determine N sets of people according to the N first cameras, where the probability that a person in the i-th person set of the N sets of people appears in the i-th camera is greater than zero, and the N The i-th person set in the person set corresponds to the i-th camera, and the value of i is each value in [1, N]; the processing unit is configured to determine according to the N person sets The first population sub-database of the first area; a storage unit for storing the first population sub-database, and the first population sub-database is used to determine the first face collected by the N first cameras The personal identity information corresponding to the data.
例如，将出现在第i个摄像机的出现概率大于某个阈值的人员，纳入第i个摄像机对应的人员集合。该阈值可以为零或者其特数值。For example, a person whose appearance probability of the i-th camera is greater than a certain threshold is included in the set of persons corresponding to the i-th camera. The threshold can be zero or its specific value.
可选地，在本申请实施例中，不同的时间段可以对应于不同的人口分库。例如，可以把一天分为1440/T个时间段，T为每个时间段的时长，1440/T个时间段可以对应于1440/T个人口分库。Optionally, in this embodiment of the application, different time periods may correspond to different population sub-databases. For example, a day can be divided into 1440/T time periods, where T is the duration of each time period, and 1440/T time periods can correspond to 1440/T personal database.
在上述技术方案中，根据人员出现在第一摄像机的人员出现概率，确定有可能出现在第一摄像机的人员集合，根据各个第一摄像机的人员集合确定该区域对应的人口库分库。由于人员出现概率可以是不断更新的，因此可以实现人口库分库内始终保持该区域最鲜活的实际活动人口，相比静态分库，可以大大减少无效人员数据。当进行人脸1:N比对实名打标签流程时，能够提升一次比对的命中率，大幅减少二次比对，达到比对资源消耗少、耗时短的效果。In the above technical solution, according to the occurrence probability of the person appearing in the first camera, the set of persons likely to appear in the first camera is determined, and the population database sub-database corresponding to the area is determined according to the set of persons of each first camera. Since the occurrence probability of personnel can be continuously updated, it can be achieved that the population database sub-database always maintains the most fresh and actual active population in the area. Compared with the static sub-database, invalid personnel data can be greatly reduced. When performing a 1:N face comparison real-name tagging process, it can increase the hit rate of the first comparison, greatly reduce the second comparison, and achieve the effect of less resource consumption and shorter time-consuming comparison.
此外，在第一人口分库中仅包含人员出现概率大于零的人员，也就是说，第一人口分库仅包括有可能出现在第一摄像机的人员，这样，可以减少第一人口分库中的数据量，降低对比对资的源消耗。In addition, the first population sub-database only includes persons whose occurrence probability is greater than zero, that is to say, the first population sub-database only includes persons who are likely to appear in the first camera. In this way, the number of the first population sub-database can be reduced. The amount of data can reduce the consumption of comparison resources.
在一种可能的实现方式中，所述装置还包括：获取单元，用于在确定第一区域的第一人口分库之前，获取M个第二摄像机拍摄的第二人脸数据，所述第二摄像机为所述第i个人员集合中的人员进行区域迁移之前所在的摄像机；所述处理单元，还用于根据所述第二人脸数据和人员迁移概率，确定所述第i个人员集合中的人员出现在所述第i个摄像机的概率，所述人员迁移概率为人员从所述第二摄像机迁移至所述第i个第一摄像机的概率。In a possible implementation, the device further includes: an acquiring unit, configured to acquire second face data captured by M second cameras before determining the first population sub-database of the first area, and the first The second camera is the camera where the persons in the i-th person set were before the area migration; the processing unit is further configured to determine the i-th person set according to the second face data and the person migration probability The probability that a person in appears at the i-th camera, and the person migration probability is the probability that a person migrates from the second camera to the i-th first camera.
在上述技术方案中，根据来自摄像机的实时人脸数据，更新人员出现概率，也就是说，可以动态更新每个摄像机所对应的人员ID列表。这样，根据摄像机所在区域，确定每个区域对应的人口库分库，可以实现人口库分库内始终保持该区域最鲜活的实际活动人口，相比全量静态库，大大减少了无效人员数据，当进行人脸1:N比对实名打标签流程时，能够提升一次比对的命中率，大幅减少二次比对，达到比对资源消耗少、耗时短的效果。In the above technical solution, the probability of occurrence of persons is updated according to the real-time face data from the cameras, that is, the list of persons ID corresponding to each camera can be dynamically updated. In this way, according to the area where the camera is located, the population database sub-database corresponding to each area can be determined, so that the population database sub-database can always maintain the most lively actual active population in the area. Compared with the full static database, it greatly reduces invalid personnel data. When performing a 1:N face comparison real-name tagging process, it can increase the hit rate of the first comparison, greatly reduce the second comparison, and achieve the effect of less resource consumption and shorter time-consuming comparison.
在一种可能的实现方式中，所述处理单元具体用于：比较所述第二人脸数据与第二人口分库中的人脸数据之间的相似性进行比对，得到所述第二人脸数据的置信度，所述置信度用于指示所述第二人脸数据与第二人口分库中的人脸数据属于同一人员的置信概率，所述第二人口分库为所述第二摄像机对应的人口分库；根据所述置信度和所述人员迁移概率，得到所述第i个人员集合中的人员出现在所述第i个摄像机的概率。In a possible implementation, the processing unit is specifically configured to: compare the similarity between the second face data and the face data in the second population sub-database to obtain the second The confidence of the face data, where the confidence is used to indicate the confidence probability that the second face data and the face data in the second population sub-database belong to the same person, and the second population sub-database is the first The population sub-database corresponding to the two cameras; according to the confidence level and the probability of people migration, the probability that a person in the i-th person set appears at the i-th camera is obtained.
可选地，可以将某个人脸数据对应的置信度与该人员对应的人员迁移概率相乘得到所述人员出现概率。Optionally, the probability of occurrence of the person may be obtained by multiplying the confidence level corresponding to a certain face data and the person migration probability corresponding to the person.
在一种可能的实现方式中，所述处理单元还用于：根据所述第一区域的常住人口库和/或流动人口库中的每个人员在预设时间段内的历史时空轨迹数据，确定人员从所述第二摄像机迁移至所述第i个第一摄像机的概率。In a possible implementation, the processing unit is further configured to: according to the historical spatiotemporal trajectory data of each person in the resident population database and/or the floating population database in the first area within a preset time period, Determine the probability of a person moving from the second camera to the i-th first camera.
在一种可能的实现方式中，所述处理单元还用于：根据所述第一区域的常住人口库和/或流动人口库，确定所述第i个人员集合中的人员出现在所述第i个第一摄像机的初始概率。In a possible implementation manner, the processing unit is further configured to: according to the permanent population database and/or the floating population database in the first area, determine that a person in the i-th person set appears in the first area. The initial probability of the i first camera.
第三方面，本申请提供了一种芯片，所述芯片与存储器相连，用于读取并执行所述存储器中存储的软件程序，以实现第一方面或第一方面任意一种实现方式所述的方法。In a third aspect, the present application provides a chip, which is connected to a memory, and is used to read and execute a software program stored in the memory to implement the first aspect or any one of the implementation manners of the first aspect Methods.
第四方面，本申请提供了一种人口库分库装置，包括存储器，用于存储程序；处理器，用于执行所述存储器存储的程序，当所述存储器存储的程序被执行时，所述处理器用于执行第一方面以及第一方面中任一种可能实现方式中的方法。In a fourth aspect, the present application provides a population database sub-database device, including a memory for storing programs; a processor for executing the programs stored in the memory, and when the programs stored in the memory are executed, the The processor is configured to execute the first aspect and the method in any one of the possible implementation manners of the first aspect.
在一种可能的实现方式中，该人口库分库装置还包括收发器。In a possible implementation manner, the population database sub-database device further includes a transceiver.
在一种可能的实现方式中，该人口库分库装置为可以应用于网络设备的芯片。In a possible implementation manner, the population database sub-database device is a chip that can be applied to network equipment.
在一种可能的实现方式中，该人口库分库装置为服务器、云主机或者容器。In a possible implementation manner, the population database sub-database device is a server, a cloud host, or a container.
第五方面，本申请提供一种计算机程序产品，该计算机程序产品包括计算机指令，当该计算机指令被执行时，使得前述第一方面或第一方面的任意可能的实现方式中的方法被执行。In a fifth aspect, the present application provides a computer program product. The computer program product includes computer instructions. When the computer instructions are executed, the foregoing first aspect or any possible implementation of the first aspect is executed.
第六方面，本申请提供了一种计算机可读存储介质，该存储介质存储有计算机指令，当所述计算机指令被执行时，使得前述第一方面或第一方面的任意可能的实现方式中的方法被执行。In a sixth aspect, the present application provides a computer-readable storage medium that stores computer instructions. When the computer instructions are executed, the foregoing first aspect or any possible implementation of the first aspect The method is executed.
附图说明Description of the drawings
图1是人口库的静态分库的示意图。Figure 1 is a schematic diagram of the static sub-database of the population database.
图2是静态分库得到的人口分库中的人口与实际活动人口的对比图。Figure 2 is a comparison diagram of the population in the population sub-database obtained by static sub-database and the actual active population.
图3是本申请实施例提供的人口库分库方法的示意性流程图。Fig. 3 is a schematic flowchart of a method for sub-database of a population database provided by an embodiment of the present application.
图4是本申请实施例的人口精准分库的方法的示意性流程图。FIG. 4 is a schematic flowchart of a method for accurate population segmentation according to an embodiment of the present application.
图5是本申请实施例的确定人口初始出现概率处理流程的示意图。Fig. 5 is a schematic diagram of a process flow for determining the initial probability of population appearance according to an embodiment of the present application.
图6是本申请实施例的人员初始出现概率的一个示例。Fig. 6 is an example of the initial appearance probability of a person in an embodiment of the present application.
图7是本申请实施例的确定人员时空迁移概率矩阵处理流程的示意图。FIG. 7 is a schematic diagram of a processing flow for determining a probability matrix of personnel spatiotemporal migration according to an embodiment of the present application.
图8是本申请实施例的人员迁移概率分布矩阵的一个示例。FIG. 8 is an example of the probability distribution matrix of personnel transfer in an embodiment of the present application.
图9是本申请实施例的更新人员出现概率处理流程的示意图。FIG. 9 is a schematic diagram of a processing flow for updating the occurrence probability of personnel in an embodiment of the present application.
图10是本申请实施例的更新人员出现概率的一个示例。FIG. 10 is an example of the appearance probability of update personnel in an embodiment of the present application.
图11是本申请实施例的二级人口库分库的示意图。Fig. 11 is a schematic diagram of a secondary population database sub-database according to an embodiment of the present application.
图12是基于本申请实施例的人口分库进行人员比对的示意性流程图。FIG. 12 is a schematic flowchart of personnel comparison based on the population sub-database of an embodiment of the present application.
图13是本申请实施例的一种检索比对的示意性流程图。FIG. 13 is a schematic flowchart of a search and comparison according to an embodiment of the present application.
图14是本申请实施例的***的一种实现形态。Fig. 14 is an implementation form of the system of the embodiment of the present application.
图15是本申请实施例的人口库分库装置的示意性结构图。Fig. 15 is a schematic structural diagram of a population database dividing device according to an embodiment of the present application.
图16是本申请另一实施例提供的人口库分库装置的示意性结构图。FIG. 16 is a schematic structural diagram of a population database sub-database device provided by another embodiment of the present application.
具体实施方式Detailed ways
下面将结合附图，对本申请中的技术方案进行描述。The technical solution in this application will be described below in conjunction with the accompanying drawings.
本申请实施例的技术方案可以应用各种场景中，只要该场景需要进行人口库的分库。例如，城市级智能人像场景、公安人口管理场景、智能交通场景等。下面以城市级智能人像场景为例，对本申请的技术方案进行描述。The technical solutions of the embodiments of the present application can be applied to various scenarios, as long as the scenario requires the sub-database of the population database. For example, city-level smart portrait scenes, public security population management scenes, intelligent traffic scenes, etc. The following uses a city-level smart portrait scene as an example to describe the technical solution of the present application.
城市级智能人像是根据动态人脸和实名的分库名单，对动态人脸进行实时打标签，从而落实动态人脸的人员身份信息。在对动态人脸进行实时打标签的过程中，会对动态人脸特征和静态人脸特征库进行1：N比对。The city-level intelligent portrait tags the dynamic face in real time according to the sub-database list of the dynamic face and the real name, so as to implement the personal identity information of the dynamic face. In the process of real-time tagging of dynamic faces, a 1:N comparison of dynamic face features and static face feature libraries is performed.
其中，人脸1：N比对是指一张人脸和一个集合比对(含N个比对对象)，查询得到和指定人脸相似度高的人脸。动态人脸是在用户无感知的情况下，在非约束场景中采集到的人脸，例如，人脸抓拍摄像机采集到的路人的人脸等。静态人脸是在用户有感知的情况下，在特定约束场景下采集到的人脸，例如，证件照采集的人脸等。人脸特征是由人脸特写图内的像素点，映射生成的特征向量。人脸特写图是从人脸场景图中扣出满足人脸识别像素 要求的人脸特写区域的图片，俗称人脸小图。人脸场景图是至少包含一个人脸和人体要素的抓拍图片，俗称人脸大图。Among them, the face 1:N comparison refers to a face and a set comparison (including N comparison objects), and the query obtains a face with a high similarity to the specified face. A dynamic face is a face that is collected in an unconstrained scene without the user's perception, for example, the face of a passerby collected by a camera that captures the face of the face. A static face is a face collected in a specific constrained scene when the user is aware, for example, a face collected from a passport photo. The face feature is a feature vector generated by mapping the pixels in the close-up image of the face. A close-up face picture is a picture of a face close-up area that meets the pixel requirements of face recognition from a face scene picture, commonly known as a small face picture. A face scene image is a captured image that contains at least one face and human elements, commonly known as a large face image.
随着静态人脸特征库(或称人口库)规模N的不断增大，会导致1：N比对的精度下降，以及比对消耗的资源和比对耗时的增加。城市级人口库通常在千万级规模，为了降低人口库规模N的大小，传统做法是按照静态人员所属的行政区划对人口库进行分库，即静态分库方式。As the scale N of the static face feature database (or population database) continues to increase, the accuracy of the 1:N comparison will decrease, and the resources consumed and the comparison time will increase. The city-level population database is usually in the tens of millions. In order to reduce the size of the population database N, the traditional method is to divide the population database according to the administrative division to which the static personnel belong, that is, the static database method.
图1是人口库的静态分库的示意图。如图1所示，在人口库的静态分库方式中，对于常住人口，按照户籍地所在行政区划来进行分类；对于流动人口，按照登记的居住地所在行政区划进行分类。分类的颗粒粒度可以根据行政区划的层次大小来定义。国内的行政区划分为省、市、县(区)、乡镇街道四级。Figure 1 is a schematic diagram of the static sub-database of the population database. As shown in Figure 1, in the static sub-database method of the population database, the permanent population is classified according to the administrative division where the household registration is located; for the floating population, the classification is performed according to the administrative division where the registered residence is located. The particle size of the classification can be defined according to the level of administrative division. The domestic administrative regions are divided into four levels: province, city, county (district), township and street.
但是，由于人口的流动性，对于常住人口，有时并不在其登记的户籍地生活和居住；对于居住在辖区内的流动人口，由于流动人口库数据更新滞后性，有可能并不在人口库分库中。也就是说，如图2所示的静态分库方式有很大概率会导致实际生活和居住在本辖区的人口，不包含在人口库的分库里面，而没生活和居住在本辖区的人口，却包含在人口库的分库里，使得人口库的分库缺乏鲜活性。这样，会导致基于人口库的分库的一次1：N比对匹配命中率低，而且即便采用一定比对步骤和逻辑，例如，一次比对在本级辖区分库完成，二次比对在上级辖区的分库完成，也会带来比对资源消耗大、比对耗时长等问题。However, due to the mobility of the population, the permanent population sometimes does not live and live in their registered household registration; for the floating population living in the jurisdiction, due to the lag in updating the data of the floating population database, it may not be in the population database sub-database. in. That is to say, the static sub-database method shown in Figure 2 has a high probability of causing the actual population living and living in the jurisdiction to not be included in the sub-database of the population database, but not the population living and living in the jurisdiction. , But included in the sub-library of the population bank, making the sub-library of the population bank lack fresh activity. In this way, it will lead to a low hit rate of a 1:N comparison match based on the population database, and even if certain comparison steps and logic are adopted, for example, the first comparison is completed in the sub-base under the jurisdiction of the level, and the second comparison is The completion of the sub-database in the higher-level jurisdiction will also bring about problems such as high resource consumption and long comparison time.
针对上述问题，本申请实施例提供了人口库分库方法和装置，能够对人口库进行更精准地分库，从而提高基于人口库的分库的一次1：N比对匹配命中率。In response to the above-mentioned problems, the embodiments of the present application provide a method and device for sub-database of a population database, which can perform a more accurate sub-database on the population database, thereby improving a 1:N matching hit rate based on the sub-database of the population database.
图3是本申请实施例提供的人口库分库方法的示意性流程图。图3所示的方法可以由服务器、云主机、容器等执行，也可以由服务器、云主机、容器等包含的芯片或者模块执行。图3所示的方法包括以下内容的至少部分内容。Fig. 3 is a schematic flowchart of a method for sub-database of a population database provided by an embodiment of the present application. The method shown in FIG. 3 may be executed by a server, a cloud host, a container, etc., or may be executed by a chip or module included in the server, a cloud host, or a container. The method shown in FIG. 3 includes at least part of the following content.
在310中，确定在第一区域内的N个第一摄像机，N为正整数。In 310, determine N first cameras in the first area, and N is a positive integer.
第一区域可以是任意大小的区域，本申请实施例不作具体限定。例如，可以是行政上的划分的省、市、县、区、街道等区划。又例如，可以是以包括预设数量的摄像机的区域。The first area may be an area of any size, which is not specifically limited in the embodiment of the present application. For example, it may be administrative divisions such as provinces, cities, counties, districts, and streets. For another example, it may be an area including a preset number of cameras.
摄像机还可以是其他具有拍照功能的装置或者设备，本申请实施例不作具体限定。The camera may also be another device or device with a photographing function, which is not specifically limited in the embodiment of the present application.
在第一区域内的N个第一摄像机为部署或者设置在第一区域内的全部摄像机。The N first cameras in the first area are all cameras deployed or set in the first area.
在一种可能的实现方式中，针对每个区域，可以保存至少一个摄像机列表，当确定第一区域内的N个第一摄像机时，可以根据第一区域的标识(identity，ID)等，确定与第一区域对应的摄像机列表。其中，摄像机列表可以是根据摄像机的地理位置和摄像机的ID确定，摄像机的地理位置可以指摄像机所在的经度和纬度度等。In a possible implementation manner, for each area, at least one camera list can be saved. When determining the N first cameras in the first area, it can be determined according to the identity (ID) of the first area, etc. A list of cameras corresponding to the first area. The camera list may be determined according to the geographic location of the camera and the ID of the camera, and the geographic location of the camera may refer to the longitude and latitude where the camera is located.
在另一种可能的实现方式中，可以直接根据第一摄像机的地理位置以及第一摄像机的ID，确定每个第一摄像机所在的区域，从而确定第一区域内的N个第一摄像机。其中，摄像机的地理位置可以指摄像机所在的经度和纬度等。In another possible implementation manner, the area where each first camera is located can be determined directly according to the geographic location of the first camera and the ID of the first camera, so as to determine the N first cameras in the first area. Among them, the geographic location of the camera may refer to the longitude and latitude where the camera is located.
在320中，根据所述N个第一摄像机，确定N个人员集合，其中，所述N个人员集合中的第i个人员集合中的人员出现在第i个摄像机的概率大于零(下文简称人员出现概率)，所述N个人员集合中的第i个人员集合与所述第i个摄像机对应，i的取值是[1，N]中的每一个值。In 320, determine N sets of people according to the N first cameras, wherein the probability that a person in the i-th person set of the N sets of people appears in the i-th camera is greater than zero (hereinafter referred to as Person appearance probability), the i-th person set in the N person sets corresponds to the i-th camera, and the value of i is each value in [1, N].
在330中，根据所述N个人员集合，确定所述第一区域的第一人口分库。In 330, a first population sub-database of the first area is determined according to the N person sets.
在340中，存储第一人口分库。In 340, the first population sub-database is stored.
在一种可能的实现方式中，根据N个人员集合中的人员出现在对应的第一摄像机的概率，确定N个第一摄像机所对应的N个人员集合，进一步根据N个人员集合确定第一区域的人口分库。其中，人员集合中人员出现在第一摄像机的概率大于0。例如，第一区域内设置有摄像机1-3，有300个人员可能会出现摄像机1，有60个人员可能会出现在摄像机2，有15个人员有可能会出现在摄像机3，那么第一区域的人口分库由375个人员中的部分或者全部组成。In a possible implementation manner, according to the probability that a person in the N person sets appears in the corresponding first camera, the N person sets corresponding to the N first cameras are determined, and the first person set is further determined according to the N person sets. The population sub-database of the region. Among them, the probability of a person in the group of persons appearing in the first camera is greater than zero. For example, there are cameras 1-3 in the first area, 300 people may appear at camera 1, 60 people may appear at camera 2, 15 people may appear at camera 3, then the first area The population sub-database of 375 is composed of some or all of the 375 people.
在另一种可能的实现方式中，本申请实施例的人口分库可以是按时间切片的划分的。具体地，根据第一时间段内的N个人员集合中的人员出现在对应的第一摄像机的概率，确定第一时间段内的N个第一摄像机所对应的N个人员集合，进一步根据N个人员集合确定第一时间段内的第一区域的人口分库；根据第二时间段内的N个人员集合中的人员出现在对应的第一摄像机的概率，确定第二时间段内的N个第一摄像机所对应的N个人员集合，进一步根据N个人员集合确定第二时间段内的第一区域的人口分库。也就是说，第一区域在不同的时间段内可以对应于不同的人口分库。本申请实施例的时间段可以是按预设时长进行划分的时间切片。例如，时间切片长度为T，一天1440分钟可以被划分为1440/T个时间切片，T可以被1440整除。该示例中时间切片的划分是以天为周期，分钟为单位进行划分的，可以理解地，本申请实施例的时间切片的划分还可以是以其他粒度的周期和/或其他粒度的单位进行划分的。In another possible implementation manner, the population sub-database in this embodiment of the present application may be divided according to time slices. Specifically, according to the probability that a person in the N person sets in the first time period appears in the corresponding first camera, determine the N person sets corresponding to the N first cameras in the first time period, and further according to N A collection of persons determines the population sub-database of the first area in the first time period; according to the probability that the persons in the N person sets in the second time period appear in the corresponding first camera, determine the N in the second time period The N personnel sets corresponding to the first cameras are further determined according to the N personnel sets to determine the population sub-database of the first area in the second time period. In other words, the first area may correspond to different population sub-databases in different time periods. The time period in this embodiment of the present application may be a time slice divided according to a preset time length. For example, if the time slice length is T, 1440 minutes a day can be divided into 1440/T time slices, and T can be divisible by 1440. In this example, the division of the time slice is divided into a period of days and a unit of minutes. It is understandable that the division of the time slice in the embodiment of the present application may also be divided into periods of other granularities and/or units of other granularities. of.
在执行320之前，还可以先确定上文所述的人员出现概率。Before executing 320, the probability of occurrence of the above-mentioned personnel may also be determined first.
在一种可能的实现方式中，当M个第二摄像机抓拍到第i个第一摄像机对应的人员集合中的人员的人脸数据时，触发更新第i个人员集合的人员出现概率。其中，第二摄像机为第一摄像机对应的人员集合中的人员进行区域迁移之前所在的摄像机。具体地，当M个第二摄像机获取到人脸数据时，根据所述人脸数据和人员迁移概率，确定所述第i个人员集合中的人员出现在所述第i个摄像机的概率，其中人员迁移概率为人员从第二摄像机迁移至第i个第一摄像机的概率。In a possible implementation manner, when the M second cameras capture the face data of a person in the person set corresponding to the i-th first camera, it is triggered to update the occurrence probability of the person in the i-th person set. Wherein, the second camera is the camera where the personnel in the personnel set corresponding to the first camera are located before the area migration. Specifically, when the face data is acquired by the M second cameras, the probability that a person in the i-th person set appears in the i-th camera is determined according to the face data and the person migration probability, where The probability of personnel migration is the probability of personnel migration from the second camera to the i-th first camera.
更具体地，作为一个示例，将所述人脸数据在第二人口分库中进行比对，得到所述人脸数据的置信度，根据得到的置信度和人员迁移概率，得到第i个人员集合中的人员出现在所述第i个摄像机的概率。其中，所述置信度用于指示所述第二人脸数据与第二人口分库中的人脸数据属于同一人员的置信概率，例如，可以是相似度等；第二人口分库为第二摄像机对应的人口分库。More specifically, as an example, the face data is compared in the second population sub-database to obtain the confidence of the face data, and the i-th person is obtained according to the obtained confidence and the probability of personnel migration The probability that a person in the set appears at the i-th camera. Wherein, the confidence is used to indicate the confidence probability that the second face data and the face data in the second population sub-database belong to the same person, for example, it may be similarity, etc.; the second population sub-database is the second The population sub-database corresponding to the camera.
可选地，上述人员迁移概率可以适用于第一区域的每一个人员。Optionally, the above-mentioned personnel migration probability may be applied to every personnel in the first area.
可选地，上述人员迁移概率也可以是针对每个人员所说的。例如，对于人员A，其从第二摄像机迁移至第一摄像机的概率为零，则意味着人员A则不会从第二摄像机迁移至第一摄像机；对于人员B，其从第二摄像机迁移至第一摄像机的概率不为零，则意味着人员B则有可能会从第二摄像机迁移至第一摄像机。Optionally, the above-mentioned personnel migration probability may also be stated for each personnel. For example, for person A, the probability of moving from the second camera to the first camera is zero, which means that person A will not move from the second camera to the first camera; for person B, it will move from the second camera to the first camera. The probability of the first camera is not zero, which means that Person B may move from the second camera to the first camera.
还应理解，当将第一摄像机抓拍到的人脸数据在第一人口分库进行比对时，还可以更新第三摄像机所对应的人员集合中的人员的迁移至第三摄像机的概率。具体地更新方法，可参见上文第一摄像机中的更新方式，在此不再赘述。It should also be understood that when the face data captured by the first camera is compared in the first population sub-database, the probability of persons in the set of persons corresponding to the third camera may also be updated to migrate to the third camera. For the specific update method, please refer to the update method in the first camera above, which will not be repeated here.
根据所述N个人员集合，确定所述第一区域的第一人口分库。在一种可能的实现方式 中，可以将出现在第一摄像机的出现概率不为零的人员纳入第一人口分库，即将出现在第二摄像机且从第二摄像机迁移至第一摄像机的概率不为零的人员纳入第一人口分库。According to the collection of the N persons, the first population sub-database of the first area is determined. In a possible implementation, people whose appearance probability is not zero can be included in the first population sub-database. The probability of appearing at the second camera and moving from the second camera to the first camera is not Those who are zero are included in the first population sub-bank.
在另一种可能的实现方式中，可以根据人员出现概率、前端摄像机的地理位置对应的区域以及人员归档的检索逻辑，形成实时人口精准分库。具体地，按照区县、街道办为单位划分二级分库；按照地市为单位划分一级分库；全库量包括全量常住人口和流动人口，一般是以省或者地市为单位。In another possible implementation manner, a real-time population accurate sub-database can be formed according to the occurrence probability of personnel, the area corresponding to the geographic location of the front-end camera, and the retrieval logic of personnel archives. Specifically, the second-level sub-library is divided according to the districts, counties and sub-district offices; the first-level sub-libraries are divided according to the prefectures and cities; the total amount of the library includes the full amount of permanent population and floating population, and is generally based on the province or city as the unit.
进一步地，上述每级分库按照精准度又可以分为一级或者多级精准分库。以两级精准分库为例，第一精准分库包括人员出现概率不为零的人员，第二精准分库包括从未出现在该区域的人员(即人员出现概率为0)和出现概率不为零的人员。这样，在进行人员比对时，可以优先在第一精准分库中进行比对，进一步提高一次人员1：N比对的命中率，减少二次对比。Furthermore, each level of the above-mentioned sub-libraries can be divided into one-level or multi-level precise sub-libraries according to accuracy. Take the two-level accurate sub-database as an example. The first accurate sub-database includes persons whose appearance probability is not zero, and the second accurate sub-database includes persons who have never appeared in the area (that is, persons whose appearance probability is 0) and those whose appearance probability is not zero. Zero personnel. In this way, when performing personnel comparisons, the first accurate sub-database can be prioritized for comparison, which further improves the hit rate of a 1:N personnel comparison and reduces secondary comparisons.
在执行320之前，还可以先确定人员迁移概率。Before executing 320, the probability of personnel transfer may also be determined.
可选地，可以根据N个人员集合中每个人员在预设时间段内的历史时空轨迹数据，确定人员从所述第二摄像机迁移至所述第一摄像机的概率。Optionally, the probability of a person migrating from the second camera to the first camera may be determined according to the historical spatiotemporal trajectory data of each person in the N person sets within a preset time period.
可选地，可以根据第一区域的常住人口库和/或流动人口库中的全部人员或者部分人员在预设时间段内的历史时空轨迹数据，确定人员从所述第二摄像机迁移至所述第一摄像机的概率。Optionally, according to the historical spatiotemporal trajectory data of all or part of the people in the permanent population database and/or the floating population database in the first area within a preset time period, it can be determined that the personnel migrated from the second camera to the Probability of the first camera.
可选地，预设时间段可以为任意可能的时间长度，例如，1个月、1周、3个月、1年等。其中，历史时空轨迹数据可以是每个人每一天的活动轨迹数据，该数据可以包括被抓拍的时间、抓拍的摄像机等。Optionally, the preset time period can be any possible time length, for example, 1 month, 1 week, 3 months, 1 year, and so on. Among them, the historical spatio-temporal trajectory data may be the activity trajectory data of each person every day, and the data may include the time captured, the captured camera, and the like.
可以理解地，在***进行初始化时，由于没有人员的实时数据，因此可以通过行政区划、实时人口分布数据生成初始人口分库。具体地，根据所述第一区域的常住人口库和/或流动人口库，确定所述第i个人员集合中的人员出现在所述第i个第一摄像机的初始概率。Understandably, when the system is initialized, since there is no real-time data of personnel, the initial population sub-database can be generated through administrative divisions and real-time population distribution data. Specifically, according to the permanent population database and/or the floating population database in the first area, the initial probability of a person in the i-th person set appearing in the i-th first camera is determined.
下面结合具体地例子，对本申请实施例的技术方案进行更详细地描述。The technical solutions of the embodiments of the present application will be described in more detail below in conjunction with specific examples.
图4是本申请实施例的感知人员迁移时空规律的人口精准分库的方法的示意性流程图。FIG. 4 is a schematic flowchart of a method for accurately sub-database of population sensing the temporal and spatial patterns of personnel migration according to an embodiment of the present application.
如图4所示，在410中，***初始化启动后，导入区划实名常住人口库和流动人口库数据，基于常住人口和流动人口的行政区划，以及各摄像机的地理位置，生成人员的初始出现概率和初始人口分库。As shown in Figure 4, in 410, after the system is initialized, the real-name permanent population database and floating population database data are imported. Based on the administrative divisions of the permanent population and the floating population, and the geographic location of each camera, the initial probability of occurrence of personnel is generated. And the initial population sub-database.
具体地，如图5所示，***的输入信息包括：Specifically, as shown in Figure 5, the input information of the system includes:
(1)摄像机ID和摄像机的地理位置，例如，经度和纬度等。(1) Camera ID and geographic location of the camera, such as longitude and latitude.
(2)按省-市-县行政区域划分的人员信息，人员信息包括常住人口信息和流动人口信息。(2) Personnel information divided by province-city-county administrative region. Personnel information includes information on permanent residents and information on floating population.
人员的初始出现概率的处理流程包括：The processing flow of the initial probability of occurrence of personnel includes:
在4101中，***初始上线时，由于没有人员的实时数据，因此通过行政区划、实时人口分布数据生成初始人口分库，或者说根据常住人口库和流动人口库，生成每个区划的初始人口分库。初始人口分库中记录有人员的ID、人员的人脸数据、空间轨迹数据、居住地等信息。In 4101, when the system was initially launched, because there was no real-time data of personnel, the initial population sub-database was generated through administrative divisions and real-time population distribution data, or the initial population sub-database of each division was generated based on the permanent population database and the floating population database. Library. The initial population sub-database records information such as the person's ID, the person's face data, the spatial trajectory data, and the place of residence.
例如，人员A登记的居住地为区划B，人员A则属于划区划B人口分库。For example, the registered residence of Person A is Division B, and Person A belongs to Division B's population sub-database.
在4102中，基于4101中得到的初始人口分库、摄像机ID和摄像机的地理位置，建立 每个摄像机的人员ID列表。该人员ID列表包括可能出现在该摄像机的人员的ID。人员ID列表可以对应于上文的人员集合。In 4102, based on the initial population sub-database, camera ID, and geographic location of the camera obtained in 4101, a list of personnel IDs for each camera is established. The list of person IDs includes the IDs of persons who may appear in the camera. The list of person IDs may correspond to the person set above.
例如，人员A登记的居住地为区划B，人员A则划入地理位置处于区划B的摄像机的人员ID列表。For example, the registered residence of person A is zone B, and person A is included in the list of personnel IDs of cameras located in zone B.
除了居住地之外，还可以用出生地、工作所在地、消费所在地等信息建立人员ID与摄像机的对应关系。In addition to the place of residence, you can also use information such as the place of birth, the place of work, and the place of consumption to establish the correspondence between the personnel ID and the camera.
在4103中，***初始启动时，为初始人员分库中的人员ID赋予初始出现概率，该出现概率包括每个时间切片内的初始出现概率。可选地，常住人员和流动人员的初始出现概率可以不同，例如，常住人员为p％，流动人员为q％。p和q可以为任意可能的取值，本申请实施例不作具体限定，例如，p＝5，q＝10；p＝15，q＝10；p＝25，q＝40；p＝50，q＝20；p＝5.5，q＝1.3等。当然p和q的取值也可以相同。In 4103, when the system is initially started, the initial appearance probability is assigned to the personnel ID in the initial personnel sub-database, and the appearance probability includes the initial appearance probability in each time slice. Optionally, the initial appearance probabilities of permanent residents and mobile persons may be different, for example, p% for permanent residents and q% for mobile persons. p and q can be any possible values, which are not specifically limited in the embodiment of the present application, for example, p=5, q=10; p=15, q=10; p=25, q=40; p=50, q =20; p=5.5, q=1.3, etc. Of course, the values of p and q can also be the same.
图5中所示的表格中的概率值表示某个人员出现在相应的摄像机的概率，例如，在时间切片(0:00，T)内，人员ID1对应的人员出现在摄像机1的出现概率为p％，人员ID3对应的人员出现在摄像机1的出现概率为q％。The probability value in the table shown in Figure 5 represents the probability of a certain person appearing in the corresponding camera. For example, in the time slice (0:00, T), the occurrence probability of the person corresponding to the person ID1 appearing in the camera 1 is p%, the probability that the person corresponding to the person ID3 appears in the camera 1 is q%.
其中，时间切片可以为一段时间，例如，1天1440分钟按照时长T，可以被划分为1440/T个时间段，其中每个时间段可以看作一个时间切片，T的取值不限定。The time slice may be a period of time. For example, a day of 1440 minutes can be divided into 1440/T time periods according to the time length T, where each time period can be regarded as a time slice, and the value of T is not limited.
此处所述的初始出现概率的设置方式可以表示初始人员分库中常住人员和流动人员，相对其他人员出现在相应区划的(或者摄像机所在区域)概率更高。图6是本申请实施例的人员初始出现概率的一个示例。如图6所示，人员初始出现概率数据以时间切片为周期，每个时间切片内的人员初始概率数据包括人员ID、对应的时间切片、摄像机ID以及人员出现在摄像机的概率。例如，在时间切片(0:00，T)内，人员ID1对应的人员出现在摄像机1的出现概率为5％，人员ID3对应的人员出现在摄像机1的出现概率为10％。The setting method of the initial appearance probability described here may indicate that the permanent residents and floating people in the initial personnel sub-database have a higher probability of appearing in the corresponding area (or the area where the camera is located) than other people. Fig. 6 is an example of the initial appearance probability of a person in an embodiment of the present application. As shown in Figure 6, the initial probability data of a person takes a time slice as a period, and the initial probability data of a person in each time slice includes a person ID, a corresponding time slice, a camera ID, and the probability of a person appearing at the camera. For example, in the time slice (0:00, T), the appearance probability of the person corresponding to the person ID1 appearing on the camera 1 is 5%, and the appearance probability of the person corresponding to the person ID3 appearing on the camera 1 is 10%.
应理解，除非后续触发了某个人员的实时人脸抓拍匹配或者实时卡口检测，否则该初始出现概率不随时间变化。It should be understood that unless the real-time face capture matching or real-time bayonet detection of a certain person is subsequently triggered, the initial probability of occurrence does not change with time.
在420中，结合人员的历史时空轨迹数据，确定每个时间切片内的人员迁移概率分布矩阵。可选地，历史时空轨迹数据可以是一段时间内的历史时空轨迹数据，例如，1个月内的历史时空轨迹数据，1周内的历史时空轨迹数据，3个月内的历史时空轨迹数据，1年内的历史时空估计数据等。In 420, combined with the historical time-space trajectory data of the personnel, the personnel migration probability distribution matrix in each time slice is determined. Optionally, the historical spatiotemporal trajectory data may be historical spatiotemporal trajectory data within a period of time, for example, historical spatiotemporal trajectory data within 1 month, historical spatiotemporal trajectory data within 1 week, historical spatiotemporal trajectory data within 3 months, Historical time-space estimation data within 1 year, etc.
具体地，如图7所示，***的输入信息包括人员的历史时空轨迹数据。人员的历史时空轨迹数据可以是“一人一档”的人员历史时空轨迹数据。其中，“一人一档”为针对每个人建立人员档案，该人员档案中可以包括对应的人员的历史时空轨迹数据以及其他个人信息。例如，获取每个人员的人员档案中1个月内的活动轨迹数据。Specifically, as shown in Fig. 7, the input information of the system includes historical spatio-temporal trajectory data of personnel. The historical spatiotemporal trajectory data of the personnel may be "one person, one file" historical spatiotemporal trajectory data of the personnel. Among them, "one person, one file" refers to the establishment of a personnel file for each person, and the personnel file may include the historical time-space trajectory data of the corresponding personnel and other personal information. For example, obtain the activity trajectory data in the personnel file of each personnel within 1 month.
在4201中，抽取人员ID、抓拍时间、抓拍摄像机ID等信息。In 4201, extract information such as personnel ID, capture time, and capture camera ID.
在4202中，按照人员ID，关联摄像机，生成每个人每一天的活动轨迹。In 4202, according to the person ID, the camera is associated to generate the activity track of each person every day.
例如，描述人员所经过的摄像机的清单：(摄像机1，摄像机2，…，摄像机k)。For example, describe the list of cameras passed by the person: (camera 1, camera 2, ..., camera k).
又例如，使用(人员ID，时间，摄像机ID)的方式，描述人员在各个时间所经过的摄像机的清单：For another example, use (person ID, time, camera ID) to describe the list of cameras passed by the person at each time:
{PersonID_1,Date_1,weekday_1,(Time_i,Camera_k),(Time_j,Camera_m),…}；{PersonID_1,Date_1,weekday_1,(Time_i,Camera_k),(Time_j,Camera_m),…};
{PersonID_1,Date_2,weekday_2,(Time_n,Camera_l),(Time_s,Camera_r),…}；{PersonID_1,Date_2,weekday_2,(Time_n,Camera_l),(Time_s,Camera_r),…};
……
{PersonID_x,Date_y,weekday_y,(Time_t,Camera_c),(Time_e,Camera_f),…}。{PersonID_x,Date_y,weekday_y,(Time_t,Camera_c),(Time_e,Camera_f),...}.
在4203中，抽取每条活动轨迹中两两相邻的摄像机的“一步”轨迹点，之所以称为“一步”，是为了描述，人员从一个摄像机的范围一步一步的进入后续摄像机的范围，每进入一个新的摄像机的范围就相当于人员“走了一步”。在人员的活动轨迹中，这两个摄像机之间是相邻的关系。In 4203, the "one-step" track points of two adjacent cameras in each activity track are extracted. The reason why they are called "one-step" is to describe that people step by step from the scope of one camera into the scope of subsequent cameras. Each entry into the range of a new camera is equivalent to a person "taking a step." In the trajectory of people, the two cameras are adjacent to each other.
例如，轨迹(摄像机1，摄像机2，摄像机3)的“一步”轨迹点为(摄像机1，摄像机2)和(摄像机2，摄像机3)。也就是说人员从摄像机1的拍摄范围出发，下一个进入的是摄像机2的拍摄范围，接着出现的是摄像机3的拍摄范围。For example, the "one-step" track points of the trajectory (camera 1, camera 2, camera 3) are (camera 1, camera 2) and (camera 2, camera 3). That is to say, the person starts from the shooting range of the camera 1, the next one enters the shooting range of the camera 2, and then the shooting range of the camera 3 appears.
在4204中，计算各个“一步”轨迹点的人员迁移概率。In 4204, calculate the probability of personnel transfer at each "one-step" track point.
可选地，可以根据某个区划的常住人口库和流动人口库所包含的全部人员的活动轨迹，统计得到各个轨迹点的人员迁移概率。Optionally, according to the activity trajectories of all the people included in the permanent population database and the floating population database of a certain district, the probability of personnel migration at each trajectory point can be obtained by statistics.
在4205中，以时间切片为单位，生成一天内的人员迁移概率分布矩阵。In 4205, a time slice is used as a unit to generate a probability distribution matrix of personnel migration within a day.
图8示出了本申请实施例的人员迁移概率分布矩阵的一个示例。具体地，图8示出了在各个时间切片下的人员迁移概率分布矩阵，以时间切片(1440-T，1440)为例，人员从摄像机3迁移至摄像机1的概率为1.2％，从摄像机3迁移至摄像机N的概率为3％。FIG. 8 shows an example of the probability distribution matrix of personnel transfer in an embodiment of the present application. Specifically, Figure 8 shows the probability distribution matrix of personnel migration in each time slice. Taking the time slice (1440-T, 1440) as an example, the probability of personnel migration from camera 3 to camera 1 is 1.2%. The probability of migrating to camera N is 3%.
可选地，本申请实施例的“一步”人员迁移概率分布矩阵是静态的。Optionally, the "one-step" personnel migration probability distribution matrix in the embodiment of the present application is static.
在430中，基于前端摄像机抓拍的人脸数据和卡口数据，以及410中生成的人员初始出现概率和420中生成的人员迁移概率分布矩阵，计算人员向各个摄像机迁移的概率，即更新人员出现概率。其中，卡口数据包括旅业数据、出行数据等。In 430, based on the face data and bayonet data captured by the front-end camera, as well as the initial appearance probability of personnel generated in 410 and the personnel migration probability distribution matrix generated in 420, the probability of personnel migration to each camera is calculated, that is, the appearance of update personnel Probability. Among them, the bayonet data includes travel industry data, travel data, etc.
具体地，如图9所示，***的输入信息包括：Specifically, as shown in Figure 9, the input information of the system includes:
(1)410中生成的常住人员和流动人员初始出现概率；(1) The initial appearance probability of permanent residents and floating persons generated in 410;
(2)420中生成的人员迁移概率分布矩阵；(2) The probability distribution matrix of personnel migration generated in 420;
(3)实时抓拍的人脸数据(例如，人员ID、抓拍时间、抓拍摄像机ID、匹配置信度等)，和/或实时卡口数据(例如，人员ID、出现时间、出现位置的等)。(3) Real-time captured face data (for example, person ID, capture time, captured camera ID, matching confidence, etc.), and/or real-time bayonet data (for example, person ID, appearance time, appearance location, etc.).
在4301中，在[mT，mT+T]时间切片内，人员抓拍记录进入缓冲队列，按照先后顺序遍历处理人员抓拍记录。In 4301, within the time slice of [mT, mT+T], the personnel capture records enter the buffer queue, and the personnel capture records are traversed and processed in sequence.
在4302中，判断人员的置信度是否高于阈值。若检测到某个人员的置信度高于或者等于阈值时，执行4303；若检测到某个人员的置信度低于阈值，则跳回4301，继续遍历。In 4302, it is determined whether the confidence of the person is higher than the threshold. If it is detected that the confidence of a certain person is higher than or equal to the threshold, execute 4303; if it is detected that the confidence of a certain person is lower than the threshold, then jump back to 4301 and continue the traversal.
其中，置信度可以为该人员在当前摄像机所对应的人口分库中进行1：N比对得到的相似度。也就是说，该人员在当前摄像机所对应的人口分库中的概率。Among them, the confidence level may be the similarity obtained by the person performing a 1:N comparison in the population sub-database corresponding to the current camera. In other words, the probability that the person is in the population database corresponding to the current camera.
在4303中，从该人员在[mT，mT+T]时间切片内的人员迁移概率分布矩阵中，获取该人员从当前摄像机向周边摄像机迁移的概率。In 4303, from the personnel migration probability distribution matrix of the personnel in the [mT, mT+T] time slice, the probability of the personnel migration from the current camera to the surrounding cameras is obtained.
在4304中，判断该人员从当前摄像机所在区域向周边摄像机迁移的概率是否大于0。若该概率大于0，执行4305；若该概率等于0，则跳回4304，继续遍历。In 4304, it is determined whether the probability of the person moving from the area where the current camera is located to the surrounding cameras is greater than 0. If the probability is greater than 0, execute 4305; if the probability is equal to 0, jump back to 4304 and continue to traverse.
在4305中，更新周边摄像机的人员出现概率。In 4305, update the probability of people appearing in surrounding cameras.
在4307中，加载新的摄像机区域人员时序分库列表，即更新各摄像机相对应的人员集合。In 4307, load the new camera area personnel sequence sub-library list, that is, update the personnel set corresponding to each camera.
图10示出了本申请实施例的更新人员出现概率的一个示例。如图10所示，该人员出 现概率列表中的人员ID包括实名归档活动人员，以及从上一个时间切片中的上一跳移动到本时间切片的人员(例如，可以根据人员迁移概率分布矩阵确定)。FIG. 10 shows an example of the occurrence probability of update personnel in an embodiment of the present application. As shown in Figure 10, the personnel IDs in the personnel appearance probability list include real-name archive activity personnel and personnel who moved from the previous jump in the previous time slice to this time slice (for example, it can be determined according to the personnel migration probability distribution matrix ).
当某个人员出现本时间切片的上一个时间切片的上一跳时，即该人员出现在图10所示的摄像机列表分别对应的前一个摄像机时，更新该人员出现在图10所示的摄像机列表中的摄像机的概率。例如，由于人员ID5-人员ID8出现在摄像机1的前一跳摄像机，人员ID5-人员ID8出现在摄像机1的概率被更新为30％、25％、28％和26％。又例如，人员ID2和人员ID4出现在摄像机1的前一跳摄像机，人员ID2和人员ID4出现在摄像机1的概率被更新为15％和6％。更新后的出现概率值可以是根据人员ID5-人员ID8在前一跳摄像机对应的人口分库中比对得到的置信度，以及人员从前一跳摄像机迁移到摄像机1的人员迁移概率得到的。When a person appears in the previous jump of the previous time slice in this time slice, that is, when the person appears in the previous camera corresponding to the camera list shown in Figure 10, update the person’s appearance in the camera shown in Figure 10 The probability of the cameras in the list. For example, since the person ID5-person ID8 appeared on the previous hop camera of the camera 1, the probability of the person ID5-person ID8 appearing on the camera 1 is updated to 30%, 25%, 28%, and 26%. For another example, the personnel ID2 and the personnel ID4 appear in the previous jump camera of the camera 1, and the probability of the personnel ID2 and the personnel ID4 appearing in the camera 1 is updated to 15% and 6%. The updated appearance probability value may be obtained based on the confidence level obtained by comparing the person ID5 to the person ID8 in the population sub-database corresponding to the previous hop camera, and the person migration probability of the person migrating from the previous hop camera to the camera 1.
当某个人员在(mT-T，mT)时间切片(即当前时间切片的上一个时间切片)内被证实出现在某个摄像机，并且在时长为T的时间段内不可能出现在另一个摄像机时，即在(mT，mT+1)时间切片内不可能出现在另一个摄像机时，将人员出现在该另一个摄像机的概率设置为0。例如，图10中的人员ID1和人员ID3被证实在出现在摄像机3，且在T时间内不可能出现在摄像机1，因此将人员ID1和人员ID3出现在摄像机1的概率设置为0。When a person is confirmed to appear in a camera in the (mT-T, mT) time slice (that is, the previous time slice of the current time slice), and it is impossible to appear in another camera during the time period of T When it is impossible to appear in another camera in the (mT, mT+1) time slice, the probability of a person appearing in the other camera is set to 0. For example, the person ID1 and the person ID3 in FIG. 10 are confirmed to appear in the camera 3, and it is impossible to appear in the camera 1 within T time, so the probability of the person ID1 and the person ID3 appearing in the camera 1 is set to 0.
其中，人员ID1和人员ID2为实名归档人员中的常住人口，因此人员ID1和人员ID2具有初始出现概率10％。人员ID3和人员ID4为实名归档人员中的流动人口，因此人员ID3和人员ID4具有初始出现概率15％。Among them, the personnel ID1 and the personnel ID2 are the permanent residents of the real-name archive personnel, so the personnel ID1 and the personnel ID2 have an initial appearance probability of 10%. Personnel ID3 and Personnel ID4 are the floating population among real-name filing personnel, so Personnel ID3 and Personnel ID4 have an initial appearance probability of 15%.
在图10中，摄像机1在时间切片(mT，mT+1)内的人员ID列表(即人员集合)包括人员ID2、人员ID4、人员ID5、人员ID6、人员ID7、人员ID8。In FIG. 10, the personnel ID list (ie, the personnel set) of the camera 1 in the time slice (mT, mT+1) includes personnel ID2, personnel ID4, personnel ID5, personnel ID6, personnel ID7, and personnel ID8.
可以理解地，本申请实施例的人员出现概率的更新由实时抓拍的人脸数据或者卡口数据触发。Understandably, the update of the person appearance probability in the embodiment of the present application is triggered by the face data or bayonet data captured in real time.
还应理解，各摄像机的对应的人员ID列表，应包含每一个时间切片内，可能被该摄像机抓拍到的人员ID和人脸特征信息。同时，还需避免将整个人口库人员信息全部纳入ID列表中，导致退化为全人口库的遍历比对。It should also be understood that the corresponding person ID list of each camera should include the person ID and facial feature information that may be captured by the camera in each time slice. At the same time, it is also necessary to avoid including the entire population database personnel information in the ID list, which will degenerate into a traversal comparison of the entire population database.
在440中，根据430中更新的人员出现概率，按照摄像机地理位置对应的区划，形成对应的按照时间切片划分的实时人口精准分库。In 440, according to the person appearance probability updated in 430, according to the division corresponding to the geographic location of the camera, a corresponding real-time population accurate sub-database divided by time slice is formed.
进一步地，可以根据430中更新的人员出现概率、前端摄像机的地理位置对应的区划以及人员归档的检索逻辑，形成实时人口精准分库。具体地，将每一个摄像机当前时间切片可能出现的人员ID加入到对应的分库中。可选地，可以基于当前每一个摄像机当前时间切片可能出现的人员ID，形成当前时间切片分库的人脸底库，基于分库人脸图片提取特征值，形成精准分库特征库。Further, a real-time population accurate sub-database can be formed according to the updated probability of occurrence of personnel in 430, the division corresponding to the geographic location of the front-end camera, and the retrieval logic of personnel archives. Specifically, the person ID that may appear in the current time slice of each camera is added to the corresponding sub-database. Optionally, the face base database of the current time slice sub-database may be formed based on the person IDs that may appear in the current time slice of each current camera, and feature values are extracted based on the sub-database face images to form an accurate sub-database feature database.
图11是本申请实施例的二级人口库分库的示意图。按照摄像机的地理位置结合行政区域进行划分。例如，按照区县、街道办为单位划分二级分库；按照地市为单位划分一级分库；全库量包括全量常住人口和流动人口，一般是以省或者地市为单位。Fig. 11 is a schematic diagram of a secondary population database sub-database according to an embodiment of the present application. According to the geographical location of the camera combined with the administrative area to divide. For example, divide the secondary sub-library according to districts, counties and sub-district offices; divide the primary sub-library according to prefectures and cities; the total volume includes all permanent residents and floating population, generally based on provinces or prefectures and cities.
可选地，每级分库按照精准度又可以分为一级或者多级精准分库。以两级精准分库为例，第一精准分库包括人员出现概率不为零的人员，第二精准分库包括从未出现在该区域的人员(即人员出现概率为0)和出现概率不为零的人员。这样，在进行人员比对时，可以优先在第一精准分库中进行比对，进一步提高一次人员1：N比对的命中率，减少二次对比。Optionally, each level of sub-library can be divided into one-level or multi-level precise sub-libraries according to accuracy. Take the two-level accurate sub-database as an example. The first accurate sub-database includes persons whose appearance probability is not zero, and the second accurate sub-database includes persons who have never appeared in the area (that is, persons whose appearance probability is 0) and those whose appearance probability is not zero. Zero personnel. In this way, when performing personnel comparisons, the first accurate sub-database can be prioritized for comparison, which further improves the hit rate of a 1:N personnel comparison and reduces secondary comparisons.
应理解，针对每个时间切片内的分库列表的更新，需在较短时间内完成。例如，耗时小于0.1T。It should be understood that the update of the sub-library list in each time slice needs to be completed in a relatively short time. For example, it takes less than 0.1T.
可选地，当在当前时间切片内有人脸数据或者卡口数据进入缓存时，在440之后，还可以基于440中得到的人口分库进行人员的1：N比对。Optionally, when the human face data or bayonet data enters the cache in the current time slice, after 440, a 1:N comparison of persons based on the population sub-database obtained in 440 may also be performed.
图12是基于本申请实施例的人口分库进行人员比对的示意性流程图。当前端摄像机抓拍到人脸数据后，根据抓拍摄像机的地理位置，关联对应的人口分库，并在该人口分库中进行1：N比对，得到与抓拍人脸相似度TOP1的人员，并获取该人员的实名标签信息，即获得了被抓拍的人员的实名标签信息。在本流程中，人口分库为在440中获得的人口精准分库。FIG. 12 is a schematic flowchart of personnel comparison based on the population sub-database of an embodiment of the present application. After the front-end camera captures the face data, it associates the corresponding population sub-database according to the geographic location of the captured camera, and performs a 1:N comparison in the population sub-database to obtain the person with the TOP1 similarity to the captured face, and Obtain the real-name tag information of the person, that is, obtain the real-name tag information of the captured person. In this process, the population sub-database is the accurate sub-database of the population obtained in 440.
进一步地，图13是本申请实施例的一种检索比对的示意性流程图。在当前时间切片，前端摄像机抓拍到人脸数据后，会首先根据摄像机所在区域对应的二级分库进行比对，如果比对未命中，则进一步在摄像机所在区域对应的一级分库进行比对，如果仍未命中，再进一步在全量库进行比对。图13所示的检索逻辑采用精准分库分级比对，可以提高比对准确度和效率，提高分库内人员命中率。Further, FIG. 13 is a schematic flowchart of a search and comparison according to an embodiment of the present application. Slice at the current time. After the front-end camera captures the face data, it will first compare it according to the second-level sub-database corresponding to the area where the camera is located. If the comparison fails, it will further compare the first-level sub-database corresponding to the area where the camera is located. Yes, if it still misses, it will be further compared in the full library. The retrieval logic shown in Fig. 13 adopts accurate sub-database grading comparison, which can improve the accuracy and efficiency of the comparison, and increase the hit rate of personnel in the sub-database.
图14是本申请实施例的***的一种实现形态。图14以视频监控智能分析***的人脸实时抓拍数据自动化归档为例，如图14所示，基于实名归档活动人口库(可以对应于上文所述的历史时空迁移数据)和人员流动记录数据，再结合行政区域的实名常住人口库以及流动人口库，生成实时的区域人员动态精准分库；导入视频监控智能分析***后，经过人像算法处理，形成人口分库人脸特征值库；前端智能摄像机的人脸抓拍数据(例如，人脸小图及对应场景大图)并上传到视频监控智能分析***，视频监控智能分析***调用人脸算法进行特征提取，并结合前端智能摄像机的具***置，映射对应区域的实时分库，进行人脸1：N比对，并将比对结果推送到人员档案库进行归档。Fig. 14 is an implementation form of the system of the embodiment of the present application. Figure 14 takes the automatic archiving of real-time face capture data of the video surveillance intelligent analysis system as an example. As shown in Figure 14, the active population database based on real-name archiving (corresponding to the historical temporal and spatial migration data described above) and personnel flow record data , Combined with the real-name permanent population database and floating population database of the administrative area to generate real-time regional personnel dynamic and accurate sub-database; after importing the video surveillance intelligent analysis system, it is processed by the portrait algorithm to form the population sub-database and face feature value database; front-end intelligence The face capture data of the camera (for example, the small face image and the corresponding large image of the scene) are uploaded to the video surveillance intelligent analysis system. The video surveillance intelligent analysis system calls the face algorithm for feature extraction and combines the specific location of the front-end smart camera. Map the real-time sub-database of the corresponding area, perform a 1:N face comparison, and push the comparison result to the personnel archive for archiving.
上文所述的技术方案是根据来自摄像机的实时人脸数据，动态更新每个摄像机所对应的人员ID列表，进一步根据摄像机所在区域，确定每个区域对应的人口库分库，可以实现人口库分库内始终保持该区域最鲜活的实际活动人口，相比全量静态库，大大减少了无效人员数据，当进行人脸1:N比对实名打标签流程时，能够提升一次比对的命中率，大幅减少二次比对，达到比对资源消耗少、耗时短的效果。The technical solution described above is to dynamically update the list of personnel IDs corresponding to each camera based on the real-time face data from the cameras, and further determine the population database sub-database corresponding to each area according to the area where the camera is located, so as to realize the population database The sub-database always maintains the most fresh and actual active population in the area. Compared with the full static database, it greatly reduces invalid personnel data. When performing the 1:N face comparison real-name tagging process, it can improve the hits of a comparison Rate, greatly reduce the secondary comparison, and achieve the effect of less resource consumption and shorter time-consuming comparison.
上文结合图1至图14，详细描述了本申请实施例提供的人口库分库方法，下面将结合图15和图16，详细描述本申请的装置实施例。应理解，本申请实施例中的装置可以执行前述本申请实施例的各种方法，即以下各种产品的具体工作过程，可以参考前述方法实施例中的对应过程。The population database sub-database method provided by the embodiments of the present application is described in detail above with reference to FIGS. 1 to 14. The device embodiments of the present application will be described in detail below with reference to FIGS. 15 and 16. It should be understood that the devices in the embodiments of the present application can execute the various methods of the foregoing embodiments of the present application, that is, for the specific working processes of the following various products, reference may be made to the corresponding processes in the foregoing method embodiments.
图15是本申请实施例的人口库分库装置的示意性结构图。图15示出的装置1500仅是示例，本申请实施例的装置还可包括其他模块或单元。如图15所示，装置1500包括处理单元1520和存储单元1530。Fig. 15 is a schematic structural diagram of a population database dividing device according to an embodiment of the present application. The device 1500 shown in FIG. 15 is only an example, and the device in the embodiment of the present application may further include other modules or units. As shown in FIG. 15, the device 1500 includes a processing unit 1520 and a storage unit 1530.
处理单元1520，用于确定在第一区域内的N个第一摄像机，N为正整数。The processing unit 1520 is configured to determine N first cameras in the first area, where N is a positive integer.
处理单元1520，，还用于根据所述N个第一摄像机，确定N个人员集合，其中，所述N个人员集合中的第i个人员集合中的人员出现在第i个摄像机的概率大于零，所述N个人员集合中的第i个人员集合与所述第i个摄像机对应，i的取值是[1，N]中的每一个值。The processing unit 1520 is further configured to determine N sets of people according to the N first cameras, where the probability that a person in the i-th person set of the N-th person sets appears in the i-th camera is greater than Zero, the i-th person set in the N person sets corresponds to the i-th camera, and the value of i is each value in [1, N].
处理单元1520，还用于根据所述N个人员集合，确定所述第一区域的第一人口分库。The processing unit 1520 is further configured to determine the first population sub-database of the first area according to the N person sets.
存储单元1530，用于存储所述第一人口分库，所述第一人口分库用于确定所述N个第一摄像机采集到的第一人脸数据对应的人员身份信息。The storage unit 1530 is configured to store the first population sub-database, and the first population sub-database is used to determine the personal identity information corresponding to the first face data collected by the N first cameras.
例如，将出现在第i个摄像机的出现概率大于某个阈值的人员，纳入第i个摄像机对应的人员集合。该阈值可以为零或者其特数值。For example, a person whose appearance probability of the i-th camera is greater than a certain threshold is included in the set of persons corresponding to the i-th camera. The threshold can be zero or its specific value.
可选地，在本申请实施例中，不同的时间段可以对应于不同的人口分库。例如，可以把一天分为1440/T个时间段，T为每个时间段的时长，1440/T个时间段可以对应于1440/T个人口分库。Optionally, in this embodiment of the application, different time periods may correspond to different population sub-databases. For example, a day can be divided into 1440/T time periods, where T is the duration of each time period, and 1440/T time periods can correspond to 1440/T personal database.
在上述技术方案中，根据人员出现在第一摄像机的人员出现概率，确定有可能出现在第一摄像机的人员集合，根据各个第一摄像机的人员集合确定该区域对应的人口库分库。由于人员出现概率可以是不断更新的，因此可以实现人口库分库内始终保持该区域最鲜活的实际活动人口，相比静态分库，可以大大减少无效人员数据。当进行人脸1:N比对实名打标签流程时，能够提升一次比对的命中率，大幅减少二次比对，达到比对资源消耗少、耗时短的效果。In the above technical solution, according to the occurrence probability of the person appearing in the first camera, the set of persons likely to appear in the first camera is determined, and the population database sub-database corresponding to the area is determined according to the set of persons of each first camera. Since the occurrence probability of personnel can be continuously updated, it can be achieved that the population database sub-database always maintains the most fresh and actual active population in the area. Compared with the static sub-database, invalid personnel data can be greatly reduced. When performing a 1:N face comparison real-name tagging process, it can increase the hit rate of the first comparison, greatly reduce the second comparison, and achieve the effect of less resource consumption and shorter time-consuming comparison.
此外，在第一人口分库中仅包含人员出现概率大于零的人员，也就是说，第一人口分库仅包括有可能出现在第一摄像机的人员，这样，可以减少第一人口分库中的数据量，降低对比对资的源消耗。In addition, the first population sub-database only includes persons whose occurrence probability is greater than zero, that is to say, the first population sub-database only includes persons who are likely to appear in the first camera. In this way, the number of the first population sub-database can be reduced. The amount of data can reduce the consumption of comparison resources.
可选地，所述装置还包括：获取单元1510，用于在确定第一区域的第一人口分库之前，获取M个第二摄像机拍摄的第二人脸数据，所述第二摄像机为所述第i个人员集合中的人员进行区域迁移之前所在的摄像机。所述处理单元1520，还用于根据所述第二人脸数据和人员迁移概率，确定所述第i个人员集合中的人员出现在所述第i个摄像机的概率，所述人员迁移概率为人员从所述第二摄像机迁移至所述第i个第一摄像机的概率。Optionally, the device further includes: an acquiring unit 1510, configured to acquire second face data captured by M second cameras before determining the first population sub-database of the first area, where the second camera is Describe the camera where the person in the i-th person set was located before the area relocation. The processing unit 1520 is further configured to determine the probability that a person in the i-th person set appears in the i-th camera according to the second face data and the person migration probability, and the person migration probability is The probability of a person moving from the second camera to the i-th first camera.
在上述技术方案中，根据来自摄像机的实时人脸数据，更新人员出现概率，也就是说，可以动态更新每个摄像机所对应的人员ID列表。这样，根据摄像机所在区域，确定每个区域对应的人口库分库，可以实现人口库分库内始终保持该区域最鲜活的实际活动人口，相比全量静态库，大大减少了无效人员数据，当进行人脸1:N比对实名打标签流程时，能够提升一次比对的命中率，大幅减少二次比对，达到比对资源消耗少、耗时短的效果。In the above technical solution, the probability of occurrence of persons is updated according to the real-time face data from the cameras, that is, the list of persons ID corresponding to each camera can be dynamically updated. In this way, according to the area where the camera is located, the population database sub-database corresponding to each area can be determined, so that the population database sub-database can always maintain the most lively actual active population in the area. Compared with the full static database, it greatly reduces invalid personnel data. When performing a 1:N face comparison real-name tagging process, it can increase the hit rate of the first comparison, greatly reduce the second comparison, and achieve the effect of less resource consumption and shorter time-consuming comparison.
可选地，所述处理单元1520具体用于：比较所述第二人脸数据与第二人口分库中的人脸数据之间的相似性，得到所述第二人脸数据的置信度，所述置信度用于指示所述第二人脸数据与第二人口分库中的人脸数据属于同一人员的置信概率，所述第二人口分库为所述第二摄像机对应的人口分库；根据所述置信度和所述人员迁移概率，得到所述第i个人员集合中的人员出现在所述第i个摄像机的概率。Optionally, the processing unit 1520 is specifically configured to: compare the similarity between the second face data and the face data in the second population sub-database to obtain the confidence of the second face data, The confidence level is used to indicate the confidence probability that the second face data and the face data in the second population sub-database belong to the same person, and the second population sub-database is the population sub-database corresponding to the second camera ; According to the confidence level and the personnel migration probability, the probability that a person in the i-th person set appears in the i-th camera is obtained.
可选地，可以将某个人脸数据对应的置信度与该人员对应的人员迁移概率相乘得到所述人员出现概率。Optionally, the probability of occurrence of the person may be obtained by multiplying the confidence level corresponding to a certain face data and the person migration probability corresponding to the person.
可选地，所述处理单元1520还用于：根据所述第一区域的常住人口库和/或流动人口库中的每个人员在预设时间段内的历史时空轨迹数据，确定人员从所述第二摄像机迁移至所述第i个第一摄像机的概率。Optionally, the processing unit 1520 is further configured to: according to the historical spatio-temporal trajectory data of each person in the resident population database and/or the floating population database in the first area within a preset time period, determine the personnel's travel The probability that the second camera migrates to the i-th first camera.
可选地，所述处理单元1520还用于：根据所述第一区域的常住人口库和/或流动人口库，确定所述第i个人员集合中的人员出现在所述第i个第一摄像机的初始概率。Optionally, the processing unit 1520 is further configured to: according to the permanent population database and/or the floating population database of the first region, determine that a person in the i-th person set appears in the i-th first person. The initial probability of the camera.
获取单元1510可以由收发器实现。处理单元1520可以由处理器实现。存储单元1530 可以由存储器实现。获取单元1510、处理单元1520和存储单元1530的具体功能和有益效果可以参见方法实施例的相关描述，在此就不再赘述。The acquiring unit 1510 may be implemented by a transceiver. The processing unit 1520 may be implemented by a processor. The storage unit 1530 may be realized by a memory. For the specific functions and beneficial effects of the acquiring unit 1510, the processing unit 1520, and the storage unit 1530, reference may be made to the relevant description of the method embodiment, which will not be repeated here.
图16是本申请另一实施例提供的人口库分库装置的示意性结构图。图16所示的装置1600(该装置1600具体可以是一种计算机设备)包括存储器1601、处理器1602、通信接口1603以及总线1604。其中，存储器1601、处理器1602、通信接口1603通过总线1604实现彼此之间的通信连接。FIG. 16 is a schematic structural diagram of a population database sub-database device provided by another embodiment of the present application. The apparatus 1600 shown in FIG. 16 (the apparatus 1600 may specifically be a computer device) includes a memory 1601, a processor 1602, a communication interface 1603, and a bus 1604. Among them, the memory 1601, the processor 1602, and the communication interface 1603 implement communication connections between each other through the bus 1604.
处理器1602，用于确定在第一区域内的N个第一摄像机，N为正整数；根据所述N个第一摄像机，确定N个人员集合，其中，所述N个人员集合中的第i个人员集合中的人员出现在第i个摄像机的概率大于零，所述N个人员集合中的第i个人员集合与所述第i个摄像机对应，i的取值是[1，N]中的每一个值；根据所述N个人员集合，确定所述第一区域的第一人口分库。The processor 1602 is configured to determine N first cameras in the first area, where N is a positive integer; and determine N sets of people according to the N first cameras, where the first set of N persons The probability that a person in the i person set appears at the i-th camera is greater than zero, the i-th person set in the N person sets corresponds to the i-th camera, and the value of i is [1, N] Each value in; according to the N person sets, determine the first population sub-database of the first area.
存储器1601，用于存储所述第一人口分库，所述第一人口分库用于确定所述N个第一摄像机采集到的第一人脸数据对应的人员身份信息。The memory 1601 is configured to store the first population sub-database, and the first population sub-database is used to determine the personal identity information corresponding to the first face data collected by the N first cameras.
其中，存储器1601可以是只读存储器(read only memory，ROM)，静态存储设备，动态存储设备或者随机存取存储器(random access memory，RAM)。存储器1601可以存储程序，当存储器1601中存储的程序被处理器1602执行时，处理器1602用于执行本申请实施例的人口库分库的方法的各个步骤，例如，可以执行图3方法实施例的各个步骤。The memory 1601 may be a read only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 1601 may store a program. When the program stored in the memory 1601 is executed by the processor 1602, the processor 1602 is configured to execute each step of the method for sub-database of the population database in the embodiment of the present application. For example, the method embodiment in FIG. 3 may be executed. The various steps.
处理器1602可能是一种集成电路芯片，具有信号的处理能力。在实现过程中，上述方法的各步骤可以通过处理器中的硬件的集成逻辑电路或者软件形式的指令完成。本申请各实施例所述的处理器可以是通用处理器、数字信号处理器(digital signal processor，DSP)、专用集成电路(application specific integrated circuit，ASIC)、现成可编程门阵列(field programmable gate array，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。可以实现或者执行本申请实施例中的公开的各方法、步骤及逻辑框图。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。结合本申请实施例所公开的方法的步骤可以直接体现为硬件译码处理器执行完成，或者用译码处理器中的硬件及软件模块组合执行完成。软件模块可以位于随机存取存储器(random access memory，RAM)、闪存、只读存储器(read-only memory，ROM)、可编程只读存储器或者电可擦写可编程存储器、寄存器等本领域成熟的存储介质中。该存储介质位于存储器，处理器读取存储器中的指令，结合其硬件完成上述方法的步骤。The processor 1602 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or instructions in the form of software. The processor described in each embodiment of the present application may be a general-purpose processor, a digital signal processor (digital signal processor, DSP), an application specific integrated circuit (ASIC), and a field programmable gate array (field programmable gate array). , FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps, and logical block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor or the like. The steps of the method disclosed in the embodiments of the present application can be directly embodied as being executed and completed by a hardware decoding processor, or executed and completed by a combination of hardware and software modules in the decoding processor. The software module can be located in random access memory (RAM), flash memory, read-only memory (read-only memory, ROM), programmable read-only memory, or electrically erasable programmable memory, registers, etc. mature in the field Storage medium. The storage medium is located in the memory, and the processor reads the instructions in the memory and completes the steps of the above method in combination with its hardware.
结合本申请实施例所公开的方法的步骤可以直接体现为硬件译码处理器执行完成，或者用译码处理器中的硬件及软件模块组合执行完成。软件模块可以位于随机存储器，闪存、只读存储器，可编程只读存储器或者电可擦写可编程存储器、寄存器等本领域成熟的存储介质中。该存储介质位于存储器1601，处理器1602读取存储器1601中的信息，结合其硬件完成人口库分库装置中包括的单元所需执行的功能，或者，执行本申请方法实施例各方法。The steps of the method disclosed in the embodiments of the present application can be directly embodied as being executed and completed by a hardware decoding processor, or executed and completed by a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the field, such as random access memory, flash memory, read-only memory, programmable read-only memory, or electrically erasable programmable memory, registers. The storage medium is located in the memory 1601, and the processor 1602 reads the information in the memory 1601, combines its hardware to complete the functions required by the units included in the population database sub-library device, or executes the methods in the method embodiments of the present application.
通信接口1603使用例如但不限于收发器一类的收发装置，来实现装置1600与其他设备或通信网络之间的通信。例如，可以通过通信接口1603获取前端摄像机动态抓拍的人脸数据。The communication interface 1603 uses a transceiving device such as but not limited to a transceiver to implement communication between the device 1600 and other devices or a communication network. For example, the face data captured dynamically by the front-end camera can be obtained through the communication interface 1603.
总线1604可包括在装置1600各个部件(例如，存储器1601、处理器1602、通信接口 1603)之间传送信息的通路。The bus 1604 may include a path for transferring information between various components of the device 1600 (for example, the memory 1601, the processor 1602, and the communication interface 1603).
需要说明的是，图16中装置1600中的处理器可以对应于图15中装置1500中的处理单元1520，通信接口1603可以对应于获取单元1510，存储器1601可以对应于存储单元1530。It should be noted that the processor in the device 1600 in FIG. 16 may correspond to the processing unit 1520 in the device 1500 in FIG. 15, the communication interface 1603 may correspond to the acquiring unit 1510, and the memory 1601 may correspond to the storage unit 1530.
应理解，本申请实施例所示的人口库分库装置可以是服务器，例如，可以是云端的服务器，或者，也可以是配置于云端的服务器中的芯片。此外，人口库分库装置可以是电子设备，或者，也可以是配置于电子设备中的芯片。It should be understood that the population database sub-database device shown in the embodiment of the present application may be a server, for example, it may be a server in the cloud, or may also be a chip configured in a server in the cloud. In addition, the population database sub-library device may be an electronic device, or may also be a chip configured in the electronic device.
应注意，尽管上述装置1600仅仅示出了存储器、处理器、通信接口，但是在具体实现过程中，本领域的技术人员应当理解，装置1600还可以包括实现正常运行所必须的其他器件。同时，根据具体需要，本领域的技术人员应当理解，装置1600还可包括实现其他附加功能的硬件器件。此外，本领域的技术人员应当理解，装置1600也可仅仅包括实现本申请实施例所必须的器件，而不必包括图16中所示的全部器件。It should be noted that although the foregoing apparatus 1600 only shows a memory, a processor, and a communication interface, in a specific implementation process, those skilled in the art should understand that the apparatus 1600 may also include other devices necessary for normal operation. At the same time, according to specific needs, those skilled in the art should understand that the apparatus 1600 may also include hardware devices that implement other additional functions. In addition, those skilled in the art should understand that the device 1600 may also include only the components necessary to implement the embodiments of the present application, and does not necessarily include all the components shown in FIG. 16.
装置1600的具体工作过程和有益效果可以参见上文方法实施例中的相关描述，在此不再赘述。For the specific working process and beneficial effects of the device 1600, reference may be made to the relevant description in the above method embodiments, and details are not described herein again.
应理解，本申请的各个方面或特征可以实现成方法、装置或使用标准编程和/或工程技术的制品。本申请中使用的术语“制品”涵盖可从任何计算机可读器件、载体或介质访问的计算机程序。例如，计算机可读介质可以包括，但不限于:磁存储器件(例如，硬盘、软盘或磁带等)，光盘(例如，压缩盘(compact disc，CD)、数字通用盘(digital versatile disc，DVD)等)，智能卡和闪存器件(例如，可擦写可编程只读存储器(erasable programmable read-only memory，EPROM)、卡、棒或钥匙驱动器等)。另外，本文描述的各种存储介质可代表用于存储信息的一个或多个设备和/或其它机器可读介质。术语“机器可读介质”可包括但不限于，无线信道和能够存储、包含和/或承载指令和/或数据的各种其它介质。It should be understood that various aspects or features of the present application can be implemented as methods, devices, or products using standard programming and/or engineering techniques. The term "article of manufacture" used in this application encompasses a computer program accessible from any computer-readable device, carrier, or medium. For example, computer-readable media may include, but are not limited to: magnetic storage devices (for example, hard disks, floppy disks, or tapes, etc.), optical disks (for example, compact discs (CD), digital versatile discs (DVD)) Etc.), smart cards and flash memory devices (for example, erasable programmable read-only memory (EPROM), cards, sticks or key drives, etc.). In addition, various storage media described herein may represent one or more devices and/or other machine-readable media for storing information. The term "machine-readable medium" may include, but is not limited to, wireless channels and various other media capable of storing, containing, and/or carrying instructions and/or data.
还应理解，还应当理解，在本申请以下各实施例中，“至少一个”、“一个或多个”是指一个、两个或两个以上。本申请中术语“和/或”，仅仅是一种描述关联对象的关联关系，表示可以存在三种关系，例如，A和/或B，可以表示：单独存在A，同时存在A和B，单独存在B这三种情况。另外，本文中字符“/”，一般表示前后关联对象是一种“或”的关系。It should also be understood that in the following embodiments of the present application, "at least one" and "one or more" refer to one, two, or more than two. The term "and/or" in this application is merely an association relationship describing associated objects, which means that there can be three types of relationships. For example, A and/or B can mean that there is A alone, and both A and B exist. There are three cases of B. In addition, the character "/" in this text generally indicates that the associated objects before and after are in an "or" relationship.
在本申请的各种实施例中，各过程的序号的大小并不意味着执行顺序的先后，各过程的执行顺序应以其功能和内在逻辑确定，而不应对本申请实施例的实施过程构成任何限定。In the various embodiments of this application, the size of the sequence number of each process does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute the implementation process of the embodiments of this application. Any restrictions.
在上述实施例中，可以全部或部分地通过软件、硬件、固件或者其他任意组合来实现。当使用软件实现时，可以全部或部分地以计算机程序产品的形式实现。所述计算机程序产品包括一个或多个计算机指令。在计算机上加载和执行所述计算机程序指令时，全部或部分地产生按照本申请实施例所述的流程或功能。所述计算机可以是通用计算机、专用计算机、计算机网络、或者其他可编程装置。所述计算机指令可以存储在计算机可读存储介质中，或者从一个计算机可读存储介质向另一个计算机可读存储介质传输，例如，所述计算机指令可以从一个网站站点、计算机、服务器或数据中心通过有线(例如同轴电缆、光纤、数字用户线(digital subscriber line，DSL))或无线(例如红外、无线、微波等)方式向另一个网站站点、计算机、服务器或数据中心进行传输。所述计算机可读存储介质可以是计算机能够存取的任何可用介质或者是包括一个或多个可用介质集成的服务器、数据中心等数据存储设备。所述可用介质可以是磁性介质(例如，软盘、硬盘、磁带)、光介质(例如数字视频光盘(digital video disc，DVD))、或者半导体介质(例如固态硬盘(solid state disk， SSD))等。In the above-mentioned embodiments, it may be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, it can be implemented in the form of a computer program product in whole or in part. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on the computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website, computer, server, or data center. Transmission to another website, computer, server, or data center via wired (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.). The computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or a data center integrated with one or more available media. The usable medium may be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a digital video disc (DVD)), or a semiconductor medium (for example, a solid state disk (SSD)), etc. .
本领域普通技术人员可以意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，能够以电子硬件、或者计算机软件和电子硬件的结合来实现。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本申请的范围。A person of ordinary skill in the art may realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be implemented by electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed by hardware or software depends on the specific application and design constraint conditions of the technical solution. Professionals and technicians can use different methods for each specific application to implement the described functions, but such implementation should not be considered beyond the scope of this application.
所属领域的技术人员可以清楚地了解到，为描述的方便和简洁，上述描述的***、装置和单元的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。Those skilled in the art can clearly understand that, for the convenience and conciseness of description, the specific working process of the system, device and unit described above can refer to the corresponding process in the foregoing method embodiment, which will not be repeated here.
在本申请所提供的几个实施例中，应该理解到，所揭露的***、装置和方法，可以通过其它的方式实现。例如，以上所描述的装置实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个***，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，装置或单元的间接耦合或通信连接，可以是电性，机械或其它的形式。In the several embodiments provided in this application, it should be understood that the disclosed system, device, and method may be implemented in other ways. For example, the device embodiments described above are merely illustrative, for example, the division of the units is only a logical function division, and there may be other divisions in actual implementation, for example, multiple units or components may be combined or It can be integrated into another system, or some features can be ignored or not implemented. In addition, the displayed or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, devices or units, and may be in electrical, mechanical or other forms.
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the objectives of the solutions of the embodiments.
另外，在本申请各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。In addition, the functional units in the various embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units may be integrated into one unit.
所述功能如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本申请各个实施例所述方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(read-only memory，ROM)、随机存取存储器(random access memory，RAM)、磁碟或者光盘等各种可以存储程序代码的介质。If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on this understanding, the technical solution of the present application essentially or the part that contributes to the existing technology or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including Several instructions are used to make a computer device (which may be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (read-only memory, ROM), random access memory (random access memory, RAM), magnetic disks or optical disks and other media that can store program codes. .
以上所述，仅为本申请的具体实施方式，但本申请的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本申请揭露的技术范围内，可轻易想到变化或替换，都应涵盖在本申请的保护范围之内。因此，本申请的保护范围应以所述权利要求的保护范围为准。The above are only specific implementations of this application, but the protection scope of this application is not limited to this. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application. Should be covered within the scope of protection of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims (11)

  1. 一种人口库分库方法，其特征在于，包括：A method for dividing a population database is characterized in that it includes:
    确定在第一区域内的N个第一摄像机，N为正整数；Determine N first cameras in the first area, where N is a positive integer;
    根据所述N个第一摄像机，确定N个人员集合，其中，所述N个人员集合中的第i个人员集合中的人员出现在第i个摄像机的概率大于零，所述N个人员集合中的第i个人员集合与所述第i个摄像机对应，i的取值是[1，N]中的每一个值；According to the N first cameras, determine N sets of people, where the probability that a person in the i-th set of people in the N sets of people appears in the i-th camera is greater than zero, and the N sets of people The i-th person set in is corresponding to the i-th camera, and the value of i is each value in [1, N];
    根据所述N个人员集合，确定所述第一区域的第一人口分库；Determine the first population sub-database of the first area according to the N person sets;
    存储所述第一人口分库，所述第一人口分库用于确定所述N个第一摄像机采集到的第一人脸数据对应的人员身份信息。The first population sub-database is stored, and the first population sub-database is used to determine the personal identity information corresponding to the first face data collected by the N first cameras.
  2. 根据权利要求1所述的方法，其特征在于，在确定第一区域的第一人口分库之前，所述方法还包括：The method according to claim 1, wherein before determining the first population sub-database of the first area, the method further comprises:
    获取M个第二摄像机拍摄的第二人脸数据，所述第二摄像机为所述第i个人员集合中的人员进行区域迁移之前所在的摄像机；Acquiring second face data captured by M second cameras, where the second camera is a camera where a person in the i-th person set was located before area migration;
    根据所述第二人脸数据和人员迁移概率，确定所述第i个人员集合中的人员出现在所述第i个摄像机的概率，所述人员迁移概率为人员从所述第二摄像机迁移至所述第i个第一摄像机的概率。According to the second face data and the probability of personnel migration, the probability that a person in the i-th person set appears in the i-th camera is determined, and the person migration probability is that the person migrates from the second camera to the The probability of the i-th first camera.
  3. 根据权利要求2所述的方法，其特征在于，根据所述第二人脸数据和人员迁移概率，确定所述第i个人员集合中的人员出现在所述第i个摄像机的概率，包括：The method according to claim 2, wherein the determining the probability that a person in the i-th person set appears in the i-th camera according to the second face data and the person migration probability comprises:
    比较所述第二人脸数据与第二人口分库中的人脸数据之间的相似性，得到所述第二人脸数据的置信度，所述置信度用于指示所述第二人脸数据与第二人口分库中的人脸数据属于同一人员的置信概率，所述第二人口分库为所述第二摄像机对应的人口分库；The similarity between the second face data and the face data in the second population sub-database is compared to obtain the confidence of the second face data, and the confidence is used to indicate the second face The confidence probability that the data and the face data in the second population sub-database belong to the same person, and the second population sub-database is the population sub-database corresponding to the second camera;
    根据所述置信度和所述人员迁移概率，得到所述第i个人员集合中的人员出现在所述第i个摄像机的概率。According to the confidence level and the personnel migration probability, the probability that a person in the i-th person set appears in the i-th camera is obtained.
  4. 根据权利要求2或3所述的方法，其特征在于，所述方法还包括：The method according to claim 2 or 3, wherein the method further comprises:
    根据所述第一区域的常住人口库和/或流动人口库中的每个人员在预设时间段内的历史时空轨迹数据，确定人员从所述第二摄像机迁移至所述第i个第一摄像机的概率。According to the historical spatio-temporal trajectory data of each person in the permanent population database and/or the floating population database in the first area within a preset time period, it is determined that the person migrates from the second camera to the i-th first Probability of the camera.
  5. 根据权利要求1至4中任一项所述的方法，其特征在于，所述方法还包括：The method according to any one of claims 1 to 4, wherein the method further comprises:
    根据所述第一区域的常住人口库和/或流动人口库，确定所述第i个人员集合中的人员出现在所述第i个第一摄像机的初始概率。According to the resident population database and/or the floating population database of the first area, determine the initial probability that a person in the i-th person set appears in the i-th first camera.
  6. 一种人口库分库装置，其特征在于，包括：A population database sub-database device, which is characterized in that it comprises:
    处理单元，用于确定在第一区域内的N个第一摄像机，N为正整数；A processing unit, configured to determine N first cameras in the first area, where N is a positive integer;
    所述处理单元，还用于根据所述N个第一摄像机，确定N个人员集合，其中，所述N个人员集合中的第i个人员集合中的人员出现在第i个摄像机的概率大于零，所述N个人员集合中的第i个人员集合与所述第i个摄像机对应，i的取值是[1，N]中的每一个值；The processing unit is further configured to determine N sets of people according to the N first cameras, wherein the probability of a person in the i-th person set in the N-th person sets appearing in the i-th camera is greater than Zero, the i-th person set in the N person sets corresponds to the i-th camera, and the value of i is each value in [1, N];
    所述处理单元，用于根据所述N个人员集合，确定所述第一区域的第一人口分库；The processing unit is configured to determine the first population sub-database of the first area according to the N person sets;
    存储单元，用于存储所述第一人口分库，所述第一人口分库用于确定所述N个第一摄像机采集到的第一人脸数据对应的人员身份信息。The storage unit is configured to store the first population sub-database, and the first population sub-database is used to determine the personal identity information corresponding to the first face data collected by the N first cameras.
  7. 根据权利要求6所述的装置，其特征在于，所述装置还包括：The device according to claim 6, wherein the device further comprises:
    获取单元，用于在确定第一区域的第一人口分库之前，获取M个第二摄像机拍摄的第二人脸数据，所述第二摄像机为所述第i个人员集合中的人员进行区域迁移之前所在的摄像机；The acquiring unit is configured to acquire second face data shot by M second cameras before determining the first population sub-database of the first area, where the second cameras perform area operations for persons in the i-th person set The camera where it was before the migration;
    所述处理单元，还用于根据所述第二人脸数据和人员迁移概率，确定所述第i个人员集合中的人员出现在所述第i个摄像机的概率，所述人员迁移概率为人员从所述第二摄像机迁移至所述第i个第一摄像机的概率。The processing unit is further configured to determine the probability that a person in the i-th person set appears in the i-th camera according to the second face data and the person migration probability, where the person migration probability is a person The probability of migrating from the second camera to the i-th first camera.
  8. 根据权利要求7所述的装置，其特征在于，所述处理单元具体用于：The device according to claim 7, wherein the processing unit is specifically configured to:
    比较所述第二人脸数据与第二人口分库中的人脸数据之间的相似性，得到所述第二人脸数据的置信度，所述置信度用于指示所述第二人脸数据与第二人口分库中的人脸数据属于同一人员的置信概率，所述第二人口分库为所述第二摄像机对应的人口分库；The similarity between the second face data and the face data in the second population sub-database is compared to obtain the confidence of the second face data, and the confidence is used to indicate the second face The confidence probability that the data and the face data in the second population sub-database belong to the same person, and the second population sub-database is the population sub-database corresponding to the second camera;
    根据所述置信度和所述人员迁移概率，得到所述第i个人员集合中的人员出现在所述第i个摄像机的概率。According to the confidence level and the personnel migration probability, the probability that a person in the i-th person set appears in the i-th camera is obtained.
  9. 根据权利要求7或8所述的装置，其特征在于，所述处理单元还用于：The device according to claim 7 or 8, wherein the processing unit is further configured to:
    根据所述第一区域的常住人口库和/或流动人口库中的每个人员在预设时间段内的历史时空轨迹数据，确定人员从所述第二摄像机迁移至所述第i个第一摄像机的概率。According to the historical spatio-temporal trajectory data of each person in the permanent population database and/or the floating population database in the first area within a preset time period, it is determined that the person migrates from the second camera to the i-th first Probability of the camera.
  10. 根据权利要求6至9中任一项所述的装置，其特征在于，所述处理单元还用于：The device according to any one of claims 6 to 9, wherein the processing unit is further configured to:
    根据所述第一区域的常住人口库和/或流动人口库，确定所述第i个人员集合中的人员出现在所述第i个第一摄像机的初始概率。According to the resident population database and/or the floating population database in the first area, determine the initial probability that a person in the i-th person set appears in the i-th first camera.
  11. 一种计算机可读存储介质，其特征在于，所述存储介质中存储有计算机程序或指令，当所述计算机程序或指令被人口库分库装置执行时，实现如权利要求1至5中任一项所述的方法。A computer-readable storage medium, characterized in that a computer program or instruction is stored in the storage medium, and when the computer program or instruction is executed by a population database sub-library device, the implementation is as in any one of claims 1 to 5. The method described in the item.
PCT/CN2020/097668 2019-09-30 2020-06-23 Person database partitioning method, and device WO2021063037A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910942816.9A CN112579593A (en) 2019-09-30 2019-09-30 Population database sorting method and device
CN201910942816.9 2019-09-30

Publications (1)

Publication Number Publication Date
WO2021063037A1 true WO2021063037A1 (en) 2021-04-08

Family

ID=75116850

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/097668 WO2021063037A1 (en) 2019-09-30 2020-06-23 Person database partitioning method, and device

Country Status (2)

Country Link
CN (1) CN112579593A (en)
WO (1) WO2021063037A1 (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113807223A (en) * 2021-09-07 2021-12-17 南京中兴力维软件有限公司 Face cluster subclass merging method, device and equipment
CN114398606A (en) * 2021-10-09 2022-04-26 武汉烽火信息集成技术有限公司 Block chain-based face verification method, equipment and computer-readable storage medium
CN115470250A (en) * 2022-11-15 2022-12-13 中关村科学城城市大脑股份有限公司 Personnel information processing method and device, electronic equipment and computer readable medium
CN116434313A (en) * 2023-04-28 2023-07-14 北京声迅电子股份有限公司 Face recognition method based on multiple face recognition modules
CN117313981A (en) * 2023-07-11 2023-12-29 厦门身份宝网络科技有限公司 Dynamic urban village population analysis method, device, equipment and readable storage medium

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103353940A (en) * 2013-05-15 2013-10-16 吴玉平 Identification method and system for dynamically adjusting comparison sequence based on probability of occurrence
CN104718749A (en) * 2012-07-31 2015-06-17 日本电气株式会社 Image processing system, image processing method, and program
US20150324822A1 (en) * 2014-05-06 2015-11-12 Mastercard International Incorporated Predicting transient population based on payment card usage
CN109543566A (en) * 2018-11-05 2019-03-29 深圳市商汤科技有限公司 Information processing method and device, electronic equipment and storage medium

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104718749A (en) * 2012-07-31 2015-06-17 日本电气株式会社 Image processing system, image processing method, and program
CN103353940A (en) * 2013-05-15 2013-10-16 吴玉平 Identification method and system for dynamically adjusting comparison sequence based on probability of occurrence
US20150324822A1 (en) * 2014-05-06 2015-11-12 Mastercard International Incorporated Predicting transient population based on payment card usage
CN109543566A (en) * 2018-11-05 2019-03-29 深圳市商汤科技有限公司 Information processing method and device, electronic equipment and storage medium

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113807223A (en) * 2021-09-07 2021-12-17 南京中兴力维软件有限公司 Face cluster subclass merging method, device and equipment
CN113807223B (en) * 2021-09-07 2024-04-09 南京中兴力维软件有限公司 Face cluster subclass merging method, device and equipment
CN114398606A (en) * 2021-10-09 2022-04-26 武汉烽火信息集成技术有限公司 Block chain-based face verification method, equipment and computer-readable storage medium
CN114398606B (en) * 2021-10-09 2024-05-03 武汉烽火信息集成技术有限公司 Face verification method, equipment and computer readable storage medium based on block chain
CN115470250A (en) * 2022-11-15 2022-12-13 中关村科学城城市大脑股份有限公司 Personnel information processing method and device, electronic equipment and computer readable medium
CN116434313A (en) * 2023-04-28 2023-07-14 北京声迅电子股份有限公司 Face recognition method based on multiple face recognition modules
CN116434313B (en) * 2023-04-28 2023-11-14 北京声迅电子股份有限公司 Face recognition method based on multiple face recognition modules
CN117313981A (en) * 2023-07-11 2023-12-29 厦门身份宝网络科技有限公司 Dynamic urban village population analysis method, device, equipment and readable storage medium
CN117313981B (en) * 2023-07-11 2024-05-14 厦门身份宝网络科技有限公司 Dynamic urban village population analysis method, device, equipment and readable storage medium

Also Published As

Publication number Publication date
CN112579593A (en) 2021-03-30

Similar Documents

Publication Publication Date Title
WO2021063037A1 (en) Person database partitioning method, and device
WO2021057797A1 (en) Positioning method and apparatus, terminal and storage medium
CN102959551B (en) Image-processing device
US10719735B2 (en) Information processing method, information processing device and video surveillance system
US9665773B2 (en) Searching for events by attendants
JP2022518459A (en) Information processing methods and devices, storage media
Huang et al. Online surveillance video synopsis
CN109902681B (en) User group relation determining method, device, equipment and storage medium
WO2021114985A1 (en) Companionship object identification method and apparatus, server and system
US20230367808A1 (en) Hyperzoom attribute analytics on the edge
CN110727756A (en) Management method and device of space-time trajectory data
US11900688B2 (en) People and vehicle analytics on the edge
CN112364176A (en) Method, equipment and system for constructing personnel action track
CN112766119A (en) Method for accurately identifying strangers and constructing community security based on multi-dimensional face analysis
KR102375145B1 (en) Integrated video data analysis and management system
KR20170095599A (en) System and method for video searching
CN111061916B (en) Video sharing system based on multi-target library image recognition
CN113065016A (en) Offline store information processing method, device, equipment and system
CN114078269A (en) Face image clustering method, device, server and storage medium
CN111859042A (en) Retrieval method and device and electronic equipment
CN110765221A (en) Management method and device of space-time trajectory data
CN116089448A (en) Real-time population management system for establishing population portraits based on multidimensional perception
JP7307887B2 (en) Information processing device, information processing method and program
CN114677627A (en) Target clue finding method, device, equipment and medium
Constantinou et al. Spatial keyframe extraction of mobile videos for efficient object detection at the edge

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20870479

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20870479

Country of ref document: EP

Kind code of ref document: A1