CN111666441A - Method, device and electronic system for determining personnel identity type - Google Patents

Method, device and electronic system for determining personnel identity type Download PDF

Info

Publication number
CN111666441A
CN111666441A CN202010336918.9A CN202010336918A CN111666441A CN 111666441 A CN111666441 A CN 111666441A CN 202010336918 A CN202010336918 A CN 202010336918A CN 111666441 A CN111666441 A CN 111666441A
Authority
CN
China
Prior art keywords
person
target
image
identity
similarity
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202010336918.9A
Other languages
Chinese (zh)
Inventor
程皓
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Kuangshi Technology Co Ltd
Beijing Megvii Technology Co Ltd
Original Assignee
Beijing Kuangshi Technology Co Ltd
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 Beijing Kuangshi Technology Co Ltd filed Critical Beijing Kuangshi Technology Co Ltd
Priority to CN202010336918.9A priority Critical patent/CN111666441A/en
Publication of CN111666441A publication Critical patent/CN111666441A/en
Priority to PCT/CN2020/119614 priority patent/WO2021212760A1/en
Pending legal-status Critical Current

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
    • G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • Library & Information Science (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Computation (AREA)
  • Evolutionary Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Databases & Information Systems (AREA)
  • Image Analysis (AREA)
  • Collating Specific Patterns (AREA)

Abstract

The invention provides a method, a device and an electronic system for determining personnel identity types; the method comprises the steps of obtaining image characteristics of a target image containing target personnel and image characteristics of an image to be compared containing personnel with known identity types; wherein the image features at least comprise human body appearance features of the person; the image characteristics of the images to be compared can reflect the identity types of people with known identity types; comparing the image characteristics of the target image and the image to be compared to obtain characteristic similarity; determining identity similarity of the target person and the person with the known identity type according to the feature similarity; and determining the identity type of the target person according to the identity similarity. The method can determine the identity type of the target person without real-name information, and has high accuracy.

Description

Method, device and electronic system for determining personnel identity type
Technical Field
The invention relates to the technical field of image processing, in particular to a method, a device and an electronic system for determining personnel identity types.
Background
For the captured face images, the images of the same person can be classified into a file in a face recognition mode; if the identity type of the person in the file is required to be known, a real-name photo library is usually provided, the image in the file is compared with the real-name photo library to obtain real-name information of the person in the file, and the identity type of the person is inquired based on the real-name information. However, the real-name photo library can only store real-name photos of a part of persons, and for some mobile persons or persons who do not register real-identity information, the identity type of the person cannot be obtained in this way.
Disclosure of Invention
In view of the above, the present invention provides a method, an apparatus and an electronic system for determining a person identity type, so as to accurately determine an identity type of a target person without real-name information.
In a first aspect, an embodiment of the present invention provides a method for determining a person identity type, where the method includes: acquiring image characteristics of a target image containing target personnel and image characteristics of an image to be compared containing personnel with known identity types; wherein the image features at least comprise human body appearance features of the person; the image characteristics of the images to be compared can reflect the identity types of people with known identity types; comparing the image characteristics of the target image and the image to be compared to obtain characteristic similarity; determining identity similarity of the target person and the person with the known identity type according to the feature similarity; and determining the identity type of the target person according to the identity similarity.
Further, the image to be compared comprises a plurality of images; each image to be compared comprises a person with a known identity type; the identity types of the persons with known identity types in the images to be compared are the same; comparing the image characteristics of the target image and the image to be compared to obtain the characteristic similarity, wherein the step comprises the following steps of: comparing the image characteristics of each target image and each image to be compared to obtain a plurality of characteristic similarities; determining identity similarity of the target person and the person with the known identity type according to the feature similarity, wherein the step comprises the following steps of: and determining the identity similarity of the target person and the person with the known identity type according to the plurality of feature similarities.
Further, the target image includes a plurality of images; the multiple target images contain the same target person; determining identity similarity of the target person and the person with the known identity type according to the plurality of feature similarities, wherein the step comprises the following steps of: aiming at each image to be compared, obtaining the maximum value of the similarity of the image to be compared in a plurality of feature similarities relative to a plurality of target images; and determining the average value or the maximum value of the maximum values of the similarity corresponding to the images to be compared as the identity similarity.
Further, the target image includes a plurality of images; the multiple target images contain the same target person; determining identity similarity of the target person and the person with the known identity type according to the plurality of feature similarities, wherein the step comprises the following steps of: for each target image, obtaining a similarity average value of a plurality of feature similarities of the target image relative to a plurality of images to be compared; and determining the maximum value of the average value of the similarity degrees corresponding to the target images as the identity similarity.
Further, the target image includes a plurality of images; the multiple target images contain the same target person; comparing the image characteristics of the target image and the image to be compared to obtain the characteristic similarity, wherein the step comprises the following steps of: comparing the image characteristics of the target image and the image to be compared to obtain the characteristic similarity corresponding to the target image aiming at each target image; determining identity similarity of the target person and the person with the known identity type according to the feature similarity, wherein the step comprises the following steps of: and determining the maximum value of the similarity in the feature similarities corresponding to the multiple target images as the identity similarity of the target person and the person with the known identity type.
Further, the target image containing the target person is stored in a first person file of the target person in advance; if the target image comprises a plurality of target images, the target images comprising the same person in the plurality of target images are stored in the same first person file.
Further, the image to be compared containing the person with the known identity type is stored in a second person file of the person with the known identity type in advance; if the image to be compared comprises a plurality of images, the images to be compared containing the same person are stored in the same second person file.
Further, the image to be compared comprises a plurality of images; storing a plurality of images to be compared in at least two second personnel files; the identity types of the people with known identity types contained in the images to be compared in the at least two second person files are the same; the target image comprises a plurality of images; storing a plurality of target images in the same first person file; comparing the image characteristics of the target image and the image to be compared to obtain the characteristic similarity, wherein the step comprises the following steps of: for each second person profile, performing the following operations: aiming at each image to be compared in the second personnel file, comparing a plurality of target images with the image to be compared to obtain a plurality of feature similarities of the target images relative to the image to be compared; determining identity similarity of the target person and the person with the known identity type according to the feature similarity, wherein the step comprises the following steps of: for each second person profile, performing the following operations: for each image to be compared in the second personnel file, acquiring the maximum value of the similarity of the image to be compared in a plurality of feature similarities relative to a plurality of target images; determining the average value or the maximum value of a plurality of similarity maximum values corresponding to a plurality of images to be compared in the second personnel file as the identity similarity of the target personnel and the personnel with known identity types corresponding to the second personnel file; and determining the average value of the plurality of identity similarities of the target person and the persons with known identity types corresponding to the plurality of second person profiles as the identity similarity.
Further, the image to be compared comprises a plurality of images; storing a plurality of images to be compared in at least two second personnel files; the identity types of the people with known identity types contained in the images to be compared in the at least two second person files are the same; the target image comprises a plurality of images; storing a plurality of target images in the same first person file; comparing the image characteristics of the target image and the image to be compared to obtain the characteristic similarity, wherein the step comprises the following steps of: for each second person profile, performing the following operations: comparing the plurality of images to be compared in the second personnel file with the target image aiming at each target image to obtain a plurality of feature similarities of the plurality of images to be compared relative to the target image; determining identity similarity of the target person and the person with the known identity type according to the feature similarity, wherein the step comprises the following steps of: for each second person profile, performing the following operations: for each target image, obtaining a similarity average value of a plurality of feature similarities of the target image relative to a plurality of images to be compared in the second personnel file; determining the maximum value of the average values of the similarity degrees corresponding to the target images as the identity similarity degrees of the target person and the person with the known identity type corresponding to the second person file; and determining the average value of the plurality of identity similarities of the target person and the persons with known identity types corresponding to the plurality of second person profiles as the identity similarity.
Further, the step of determining the identity type of the target person according to the identity similarity includes: and if the identity similarity is higher than a preset similarity threshold, determining the identity type of the target person as a known identity type.
Further, the second person profile comprises a plurality of profile groups; the identity types of the second personnel files in the same file group are the same; the identity types of second personnel files of different file groups are different; the target image comprises a plurality of images; storing a plurality of target images in the same first person file; comparing the image characteristics of the target image and the image to be compared to obtain the characteristic similarity, wherein the step comprises the following steps of: aiming at each second person file in each group of file groups and each image to be compared in the second person file, comparing a plurality of target images in a first person file with the image to be compared to obtain a plurality of feature similarities of the first person file relative to the file groups; determining identity similarity of the target person and the person with the known identity type according to the feature similarity, wherein the step comprises the following steps of: for each second person profile in each profile group, performing the following operations: for each image to be compared in the second personnel file, acquiring the maximum value of the similarity of the image to be compared in a plurality of feature similarities relative to a plurality of target images; determining the average value or the maximum value of a plurality of similarity maximum values corresponding to a plurality of images to be compared as the identity similarity of the target person and the person with the known identity type corresponding to the second person file; and determining the average value of the identity similarities of the target person and the persons with known identity types corresponding to the second person profiles in the same profile group as the identity similarity of the target person and the profile group.
Further, the second person profile comprises a plurality of profile groups; the identity types of the second personnel files in the same file group are the same; the identity types of second personnel files of different file groups are different; the target image comprises a plurality of images; storing a plurality of target images in the same first person file; comparing the image characteristics of the target image and the image to be compared to obtain the characteristic similarity, wherein the step comprises the following steps of: for each second person profile in each group of profiles group, performing the following operations: comparing the plurality of images to be compared in the second personnel file with the target image aiming at each target image to obtain a plurality of feature similarities of the plurality of images to be compared relative to the target image; determining identity similarity of the target person and the person with the known identity type according to the feature similarity, wherein the step comprises the following steps of: for each second person profile in each profile group, performing the following operations: for each target image, obtaining a similarity average value of a plurality of feature similarities of the target image relative to a plurality of images to be compared in the second personnel file; determining the maximum value of the average values of the similarity degrees corresponding to the target images as the identity similarity degrees of the target person and the person with the known identity type corresponding to the second person file; and determining the average value of the identity similarities of the target person and the persons with known identity types corresponding to the second person profiles in the same profile group as the identity similarity of the target person and the profile group.
Further, the step of determining the identity type of the target person according to the identity similarity includes: selecting the highest identity similarity from the identity similarities corresponding to the multiple groups of file groups, and determining the identity type of the file group corresponding to the highest identity similarity as the identity type of the target person; or, aiming at each group of file groups, judging whether the identity similarity corresponding to the file group is higher than the similarity threshold corresponding to the file group; and if so, determining the identity type corresponding to the archive group as the identity type of the target person.
Further, the target image in the first person file is provided with acquisition time; the first person file comprises a plurality of target images; after the step of determining the identity type of the target person according to the identity similarity, the method further comprises: and determining the working time of the target personnel according to the corresponding feature similarity of each target image and the acquisition time of the target image.
Further, the step of determining the working time of the target person according to the feature similarity corresponding to each target image and the acquisition time of the target image includes: judging whether the feature similarity of each target image is higher than a preset similarity threshold value or not; if the image is higher than the preset image, determining that the acquisition time corresponding to the target image belongs to the working time of the target personnel; and determining a time period formed by the acquisition time of the working time of the target personnel as the working time of the target personnel.
Further, after the step of determining the working time of the target person according to the feature similarity corresponding to each target image and the acquisition time of the target image, the method further includes: acquiring the working time of target personnel within a preset area range; and determining the number of the target personnel in the working state at the specified time point within the preset area range according to the working time of the target personnel.
Further, the target person includes a plurality of persons; the first person files of a plurality of target persons belong to the same preset area range; after the step of determining the identity type of the target person according to the identity similarity, the method further comprises: and screening the target persons in the preset area range according to the identity type corresponding to the first person file of each target person to obtain the designated person.
In a second aspect, an embodiment of the present invention provides an apparatus for determining a type of a person identity, where the apparatus includes: the characteristic acquisition module is used for acquiring image characteristics of a target image containing target personnel and image characteristics of an image to be compared containing personnel with known identity types; wherein the image features at least comprise human body appearance features of the person; the image characteristics of the images to be compared can reflect the identity types of people with known identity types; the characteristic comparison module is used for comparing the image characteristics of the target image and the image to be compared to obtain the characteristic similarity; the identity similarity determining module is used for determining the identity similarity of the target person and the person with the known identity type according to the feature similarity; and the identity type determining module is used for determining the identity type of the target person according to the identity similarity.
In a third aspect, an embodiment of the present invention provides an electronic system, including: a processing device and a storage device; the storage means has stored thereon a computer program which, when being executed by the processing device, carries out the above-mentioned method of determining the identity type of a person.
In a fourth aspect, an embodiment of the present invention provides a machine-readable storage medium, on which a computer program is stored, where the computer program is executed by a processing device to perform the steps of determining the identity type of the person as described above.
The embodiment of the invention has the following beneficial effects:
the method, the device and the electronic system for determining the identity type of the person firstly acquire the image characteristics of the target image containing the target person and the image characteristics of the image to be compared containing the person with the known identity type; then comparing the image characteristics of the target image and the image to be compared to obtain characteristic similarity; and then according to the feature similarity, determining the identity similarity of the target person and the person with the known identity type, and according to the identity similarity, determining the identity type of the target person. The image features in the method at least comprise human body shape features of people, and the image features of the images to be compared can represent the identity types of people with known identity types, so that the method can predict the identity types of target people with the identity types represented by the shape features, has high accuracy, is favorable for managing the people, improves the efficiency of finding certain specific identity type people, and realizes automatic analysis of the identity of the people.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
In order to make the aforementioned and other objects, features and advantages of the present invention comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a schematic structural diagram of an electronic system according to an embodiment of the present invention;
fig. 2 is a flowchart of a method for determining a person identity type according to an embodiment of the present invention;
fig. 3 is a flowchart of another method for determining the identity type of a person according to an embodiment of the present invention;
fig. 4 is a flowchart of another method for determining the identity type of a person according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of an apparatus for determining a person identity type according to an embodiment of the present invention.
Detailed Description
To make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In order to obtain the identity type of a person without real-name information, the method, the device and the electronic system for determining the identity type of the person provided by the embodiment of the invention can be applied to various fields such as personnel management, city management, public security and the like, and can be realized by adopting corresponding software and hardware, and the embodiment of the invention is described in detail below.
The first embodiment is as follows:
first, an example electronic system 100 for implementing the method, apparatus, and electronic system for determining a person identity type of embodiments of the present invention is described with reference to fig. 1.
As shown in FIG. 1, an electronic system 100 includes one or more processing devices 102, one or more memory devices 104, an input device 106, an output device 108, and one or more image capture devices 110, which are interconnected via a bus system 112 and/or other type of connection mechanism (not shown). It should be noted that the components and structure of the electronic system 100 shown in fig. 1 are exemplary only, and not limiting, and that the electronic system may have other components and structures as desired.
The processing device 102 may be a gateway or an intelligent terminal, or a device including a Central Processing Unit (CPU) or other form of processing unit having data processing capability and/or instruction execution capability, and may process data of other components in the electronic system 100 and may control other components in the electronic system 100 to perform desired functions.
The storage 104 may include one or more computer program products that may include various forms of machine-readable storage media, such as volatile memory and/or non-volatile memory. The volatile memory may include, for example, Random Access Memory (RAM), cache memory (cache), and/or the like. The non-volatile memory may include, for example, Read Only Memory (ROM), hard disk, flash memory, etc. On which one or more computer program instructions may be stored that may be executed by processing device 102 to implement client functionality (implemented by the processing device) and/or other desired functionality in embodiments of the invention described below. Various applications and various data, such as various data used and/or generated by the applications, may also be stored in the machine-readable storage medium.
The input device 106 may be a device used by a user to input instructions and may include one or more of a keyboard, a mouse, a microphone, a touch screen, and the like.
The output device 108 may output various information (e.g., images or sounds) to the outside (e.g., a user), and may include one or more of a display, a speaker, and the like.
The image capture device 110 may capture preview video frames or image data and store the captured preview video frames or image data in the storage 104 for use by other components.
For example, the devices in the electronic system for implementing the method, the apparatus, and the electronic system for determining the identity type of a person according to the embodiments of the present invention may be integrally disposed, or may be disposed in a distributed manner, such as integrally disposing the processing device 102, the storage device 104, the input device 106, and the output device 108, and disposing the image capturing device 110 at a designated position where a target image can be captured. When the above-described devices in the electronic system are integrally provided, the electronic system may be implemented as an intelligent terminal such as a camera, a smart phone, a tablet computer, a vehicle-mounted terminal, and the like.
Example two:
referring to fig. 2, a flow chart of a method for determining the identity type of a person is shown, which is performed by a processing device in the electronic system; the processing device may be any device or chip having data processing capabilities. The processing equipment can independently process the received information, can also be connected with the server, jointly analyzes and processes the information, and uploads a processing result to the cloud. The method comprises the following steps:
step S202, acquiring image characteristics of a target image containing target personnel and image characteristics of an image to be compared containing personnel with known identity types; wherein the image features at least comprise human body appearance features of the person; the image characteristics of the images to be compared can reflect the identity types of people with known identity types;
wherein the target person is a person whose identity type needs to be determined in this embodiment; because the real-name information of the target person cannot be obtained, the identity type of the target person needs to be inferred according to the image characteristics of other persons with known identity types. The image characteristics of the target image containing the target personnel can be obtained through a preset characteristic extraction network, and similarly, the image characteristics of the image to be compared containing the personnel with known identity types can also be obtained through the characteristic extraction network. The image features are obtained by extracting through a feature extraction network, and the image features can usually represent various feature attributes of a target person, such as the appearance features of clothing, stature, hair style, face shape and the like of the person in the image, and can also represent the environmental features of the person, the environment, articles, buildings and the like around the person.
Wherein the image features at least comprise human body appearance features of the person; the human body appearance characteristics can be specifically characteristics related to the identity type of a person or the industry, such as characteristics of clothes, hats, shoes or articles worn by the person; the image features may also include environmental features of the environment where the person is located, which may be features associated with the identity type of the person or the industry, such as vehicles, buildings, environment types around the person, and the like; the environment type can be specifically community, highway, field and other environment types. Generally, the image characteristics of the images to be compared can reflect the identity types of people with known identity types; for example, a uniform capable of representing the known identity type is usually worn by the person in the image to be compared, or the person in the image to be compared is in an environment capable of representing the known identity type, and at this time, the image feature of the image to be compared may represent the identity type of the person with the known identity type.
The target image containing the target person may be an original panoramic image obtained by shooting by a camera device, or an image obtained by performing pedestrian detection on the panoramic image through a pedestrian detection algorithm and performing screenshot on an image area containing pedestrians in the panoramic image. When the target image is a panoramic image, the image characteristics of the target image can usually embody more comprehensive attribute characteristics of the target person, such as appearance characteristics and environmental characteristics of the target person; when the target image is a screenshot image, the target image may only include a part of attribute features of the target person, such as only appearance features of the target person, and a small part of environmental features, even no environmental features.
Step S204, comparing the image characteristics of the target image and the image to be compared to obtain characteristic similarity;
in actual implementation, the image characteristics of the target image and the image to be compared can be compared in a mode of calculating the image characteristic distance; the image characteristic distance, namely the characteristic similarity, can be specifically calculated by an Euclidean distance, a Manhattan distance and a Chebyshev distance equidistance calculation mode, and the image characteristic distance between the target image and the image to be compared is calculated. Generally, the closer the image feature distance between the target image and the image to be compared is, the higher the feature similarity between the target image and the image to be compared is; the farther the image feature distance between the target image and the image to be compared is, the lower the feature similarity between the target image and the image to be compared is.
Step S206, determining identity similarity of the target person and the person with the known identity type according to the feature similarity;
it can be understood that the higher the feature similarity of the image features of the target image and the image to be compared, the higher the identity similarity of the target person and the person with the known identity type.
And S208, determining the identity type of the target person according to the identity similarity.
For example, if the identity similarity is high, the identity type of the target person may be determined to be the same as the identity type of the person included in the image to be compared, that is, the identity type of the target person is the known identity type. If the identity similarity is low, the identity type of the target person can be determined, and is different from the identity types of the persons contained in the images to be compared, namely the identity type of the target person is not the known identity type.
In practical implementation, the above known identity types may be various; the image characteristics of the target image containing the target person can be compared with the image characteristics of the image to be compared containing each type of person with known identity type one by one to obtain the identity similarity between the target person and each type of person with known identity type, and the type of the known identity with higher identity similarity is selected to be determined as the identity type of the target person. The identity type can be identity types of express delivery takeout personnel, environmental sanitation personnel, security personnel, armed police, hospital staff, police and the like. The shapes, environments and the like of the personnel with the identity types generally have more uniform characteristics, and the identity types of the personnel without real-name information can be obtained through reasoning by comparing image characteristics.
It should be noted that, in the method for determining the identity type of the person in this embodiment, usually only the identity type of the target person can be obtained, and the real-name information of the target person cannot be directly obtained; however, obtaining the identity type of the target person is also helpful for narrowing the investigation scope of the target person, and is convenient for managing the target person.
The method for determining the identity type of the personnel obtains the image characteristics of the target image containing the target personnel and the image characteristics of the image to be compared containing the personnel with known identity types; then comparing the image characteristics of the target image and the image to be compared to obtain characteristic similarity; and then according to the feature similarity, determining the identity similarity of the target person and the person with the known identity type, and according to the identity similarity, determining the identity type of the target person. The image features in the method at least comprise human body shape features of people, and the image features of the images to be compared can represent the identity types of people with known identity types, so that the method can predict the identity types of target people with the identity types represented by the shape features, has high accuracy, is favorable for managing the people, improves the efficiency of finding certain specific identity type people, and realizes automatic analysis of the identity of the people.
Example three:
in the method, in order to improve the accuracy of determining the identity type of a target person, a to-be-compared image includes a plurality of images; each image to be compared comprises a person with a known identity type; the identity types of the persons with known identity types contained in the images to be compared are the same.
When comparing the image characteristics of the target image and the image to be compared, comparing the image characteristics of each target image and each image to be compared to obtain a plurality of characteristic similarities; and further determining the identity similarity of the target person and the person with the known identity type according to the plurality of feature similarities. For example, the target image includes one image, the image to be compared includes a plurality of images, and after the target image is compared with the image features of each image to be compared, the feature similarity corresponding to each image to be compared, that is, the feature similarities are obtained; in one mode, the multiple feature similarities can be subjected to subsequent processing, so that the identity similarities of the target person and the persons with known identity types are obtained; for example, the identity similarity may be obtained by averaging, weighted averaging, and the like of the plurality of feature similarities; in the process of averaging and weighted averaging, one or more maximum values of the plurality of feature similarities may be removed, or one or more minimum values of the plurality of feature similarities may be removed, and then the remaining feature similarities may be averaged, weighted averaged, and so on. In addition, the maximum value or the minimum value of the feature similarity can be determined as the identity similarity.
The accuracy and the rationality of subsequently conjecturing the identity type of the target person can be further improved by comparing the characteristics of the plurality of images to be compared with the target image. For example, assuming that the images to be compared include 10 images, the identity types of the persons in the 10 images to be compared are all security, and the identity of the target person in the target image is not actually security; if the target person in the target image is compared with one of the 10 images to be compared, and the person in the image to be compared and the target person are provided with similar accessories, the similarity of the characteristics of the person and the target person is high, the identity type of the target identity is possibly determined to be security, and the identity type estimation is wrong. If the target image is compared with 10 images to be compared, the similarity of the characteristics of the target image and most of the images to be compared is low in high probability, and at the moment, the identity type of the target person cannot be judged as security.
In this embodiment, the target image may include a plurality of target images or one target image, and when the target image includes a plurality of target images, the plurality of target images include the same target person, and the target person is a person whose identity type needs to be determined. Then, aiming at each image to be compared, obtaining the maximum value of the similarity of the image to be compared in a plurality of feature similarities relative to a plurality of target images; and determining the average value or the maximum value of the maximum values of the similarity corresponding to the images to be compared as the identity similarity.
When the target image comprises a plurality of images, each image contains different quantities of image features capable of representing the identity type; for example, some target persons in the target image wear uniforms capable of representing the identity types, and the amount of image features capable of representing the identity types contained in the target image is large; the target person in some target images wears the casual clothes, and the amount of image features which can embody the identity type and are contained in the target images is small; after the image to be compared is compared with each target image in a characteristic way, the maximum value of the similarity of the characteristic similarity is selected, and the target image which can reflect the identity type most can be selected from the multiple target images; taking the feature similarity corresponding to the target image as the maximum value of the similarity corresponding to the current image to be compared; and determining the average value or the maximum value of the maximum values of the similarity of the plurality of images to be compared as the identity similarity.
As an example, assume that the target image includes three, respectively target image a, target image B, and target image C; the images to be compared comprise four images which are respectively an image 1 to be compared, an image 2 to be compared, an image 3 to be compared and an image 4 to be compared. The results of the image characteristics of the three target images and the four images to be compared are shown in table 1 below.
TABLE 1
Object image A Target image B Object image C
Image to be compared 1 0.8 0.9 0.7
Image to be compared 2 0.6 0.7 0.5
Image to be compared 3 0.9 0.9 0.8
Image to be compared 4 0.5 0.5 0.6
The table 1 includes the feature similarity of the image features of each target image and each image to be compared, for example, the feature similarity of the image features of the image to be compared 1 and the target image a is 0.8. Aiming at each image to be compared, obtaining the maximum value of the similarity of the image to be compared in a plurality of feature similarities relative to a plurality of target images; for example, the maximum value of the similarity corresponding to the image to be compared 1 is 0.9, the maximum value of the similarity corresponding to the image to be compared 2 is 0.7, the maximum value of the similarity corresponding to the image to be compared 3 is 0.9, and the maximum value of the similarity corresponding to the image to be compared 4 is 0.6; and then, determining the average value or the maximum value of the maximum values of the similarity corresponding to the images to be compared as the identity similarity. For example, the maximum values of the similarity corresponding to the four images to be compared are 0.9, 0.7, 0.9 and 0.6 respectively; in one mode, the maximum value, namely 0.9, is selected from the four maximum values of similarity as the identity similarity; in another mode, the average of the four maximum similarity values, i.e., (0.9+0.7+0.9+0.6)/4, is determined as the identity similarity.
When identity similarity is determined, there is another implementation manner, for example, when the target image includes a plurality of target images, the plurality of target images contain the same target person; for each target image, obtaining a similarity average value of a plurality of feature similarities of the target image relative to a plurality of images to be compared; and determining the maximum value of the average value of the similarity degrees corresponding to the target images as the identity similarity. Also taking the above table 1 as an example, the average similarity value of the feature similarities of the target image a with respect to the images to be compared is (0.8+0.6+0.9+0.5)/4 is 0.7; the average similarity value of the target image B relative to the feature similarities of the images to be compared is (0.9+0.7+0.9+ 0.5)/4-0.75; the average similarity value of the target image C relative to the feature similarities of the images to be compared is (0.7+0.5+0.8+ 0.6)/4-0.65; wherein the maximum value is 0.75, i.e. identity similarity.
In the method, aiming at each target image, a plurality of images to be compared and the target image are subjected to feature comparison to obtain a plurality of feature similarities of the target image; determining the average value or the maximum value of the feature similarities as the feature similarity corresponding to the target image; and determining the maximum value of the feature similarity corresponding to the target images as the identity similarity.
In addition, another way of comparing image features and determining identity similarity is provided in this embodiment, where the target image includes a plurality of target images, the plurality of target images include the same target person, and the image to be compared includes one target person, and at this time, for each target image, the image features of the target image and the image to be compared are compared to obtain the feature similarity corresponding to the target image; and determining the maximum value of the similarity in the feature similarities corresponding to the multiple target images as the identity similarity of the target person and the person with the known identity type. Also taking the table 1 as an example, assuming that only the image 1 to be compared exists, after the three target images are compared with the image features of the image 1 to be compared, three feature similarities are obtained, which are 0.8, 0.9 and 0.7 respectively, wherein the maximum value of the similarity is 0.9, which is the identity similarity between the target person and the person with the known identity type.
In the above manner, the image features of the target image and the image to be compared are compared in multiple manners to obtain feature similarity, and according to the feature similarity, the identity similarity between the target person and the person with the known identity type is determined, so as to determine the identity type of the target person. The method can deduce the identity type of the target personnel with the identity type through appearance characteristics, and has high accuracy, thereby being beneficial to managing the personnel, improving the efficiency of finding certain personnel with specific identity types and realizing the automatic analysis of the identity of the personnel.
Example four:
in this embodiment, a target image including a target person is pre-stored in a first person profile of the target person; if the target image comprises a plurality of target images, the target images comprising the same person in the plurality of target images are stored in the same first person file. Namely, a one-to-one correspondence relationship exists between the target person and the first person profile; storing the target images of the same target person in a first person file; the target images of different target persons are stored in different first person profiles. It is understood that, in a plurality of target images in the same first person profile, a target image capable of representing an identity type may be included, or a target image capable of representing an identity type may not be included, and both a target image capable of representing an identity type (for example, in the target image, the target person wears a uniform) and a target image incapable of representing an identity type (for example, in the target image, the target person does not wear a uniform) may be included.
In actual implementation, the multiple target images comprise the same person and are located in the same first person file; in the plurality of images to be compared, if the personnel contained in each image to be compared are different, at the moment, the plurality of images to be compared cannot be located in the file. At this time, the specific implementation manner of performing the feature comparison between the multiple target images and the multiple images to be compared can refer to the foregoing embodiment, and is not described herein again.
In another implementation manner, the plurality of images to be compared can be further divided into a plurality of groups, the images to be compared in each group have the same identity type, and the images to be compared in different groups have different identity types; for example, the plurality of images to be compared may be divided into a sanitation worker group, a security group, a takeaway personnel group, and the like. And comparing the target image with each group of images to be compared to obtain the identity similarity of the target image and each group of images to be compared, and determining the identity type of the group with the highest identity similarity as the identity type of the target person in the target image.
In addition, the image to be compared containing the person with the known identity type can be stored in a second person file of the person with the known identity type in advance; if the image to be compared comprises a plurality of images, the images to be compared containing the same person are stored in the same second person file. That is, there is a one-to-one correspondence between the known identity type person and the second person profile. Storing the images to be compared of the same person with the known identity type in a second person file; the images to be compared of different persons with known identity types are stored in different second person files.
In a specific implementation, the embodiment adopts at least two second personnel files to determine the identity type of the target personnel; at this time, the images to be compared comprise a plurality of images; storing a plurality of images to be compared in at least two second personnel files; the identity types of the people with known identity types contained in the images to be compared in the at least two second person files are the same; each second personnel file corresponds to a person with a known identity type, and the persons with the known identity types corresponding to different second personnel files are different; therefore, in this embodiment, the images to be compared of a plurality of persons with the same identity type are compared with the target image. The target image comprises a plurality of images; the plurality of target images are stored in the same first person profile. For example, two second person files are obtained, each second person file corresponds to a person, the identity types of the two persons are the same, if both persons are sanitation workers, and whether the target person is also sanitation workers is determined through images to be compared in the two second person files.
As shown in fig. 3, the method for determining the identity type of the person in this embodiment includes the following steps:
step S302, acquiring image characteristics of a target image containing target personnel and image characteristics of an image to be compared containing personnel with known identity types; wherein the image features at least comprise human body appearance features of the person; the image characteristics of the images to be compared can reflect the identity types of people with known identity types;
step S304, for each second person profile, performing the following operations: aiming at each image to be compared in the second personnel file, comparing a plurality of target images with the image to be compared to obtain a plurality of feature similarities of the target images relative to the image to be compared;
for each second personnel file, one or more images to be compared are included, and a plurality of feature similarities of each image to be compared relative to a plurality of target images can be obtained through the steps; for example, a second person file includes three images to be compared, and a target image includes two images, where each image to be compared can obtain two feature similarities.
Step S306, for each second person profile, performing the following operations: for each image to be compared in the second personnel file, acquiring the maximum value of the similarity of the image to be compared in a plurality of feature similarities relative to a plurality of target images; determining the average value or the maximum value of a plurality of similarity maximum values corresponding to a plurality of images to be compared in the second personnel file as the identity similarity of the target personnel and the personnel with known identity types corresponding to the second personnel file;
as can be seen from the above, each image to be compared corresponds to a plurality of feature similarities, and a maximum value, that is, the maximum value of the similarity, is selected from the plurality of feature similarities, and at this time, each image to be compared in the second person file corresponds to a maximum value of the similarity; if the images to be compared in the second personnel file comprise a plurality of images, selecting a maximum value from the maximum values of the similarity corresponding to each image to be compared, and determining the maximum value as the identity similarity of the target personnel and the personnel with known identity types corresponding to the second personnel file; or, calculating an average value of the maximum similarity values corresponding to each image to be compared, and determining the average value as the identity similarity between the target person and the person with the known identity type corresponding to the second person file. At this time, if the second person profile is multiple, each second person profile corresponds to an identity similarity.
Step S308, determining the average value of the identity similarities of the target person and the persons with known identity types corresponding to the second person profiles as the identity similarity.
Step S310, if the identity similarity is higher than a preset similarity threshold, determining that the identity type of the target person is a known identity type.
For example, if the known identity type corresponding to the second person profile is a sanitation worker, and the identity similarity between the target person and the known identity type corresponding to the second person profile is higher than the preset similarity threshold, it can be presumed that the identity type of the target person is also a sanitation worker; if the identity similarity of the known identity type corresponding to the target person and the second person profile is lower than or equal to a preset similarity threshold, the identity type of the target person can be presumed not to be a sanitation worker; at this time, the second person file with other identity type, such as security, can be obtained, and whether the target person is security is continuously judged through the above steps until the identity type of the target person is presumed.
It should be noted that, if the second personnel files of multiple identity types are included, the target image can be compared with the images to be compared in the second personnel files of each identity type one by one according to a certain sequence; the target image can also be compared with the image to be compared in the second personnel file of each identity type at the same time.
Corresponding to the method of fig. 3, there is another implementation manner that, when comparing the image features of the target image and the image to be compared, the following operations may be performed for each second person file: comparing the plurality of images to be compared in the second personnel file with the target image aiming at each target image to obtain a plurality of feature similarities of the plurality of images to be compared relative to the target image; in the method, comparison is performed from the dimensions of the target images, that is, each target image is compared with a plurality of images to be compared in one second person file, so that a plurality of feature similarities corresponding to each target image are obtained.
Then, for each second person profile, the following operations are performed: for each target image, obtaining a similarity average value of a plurality of feature similarities of the target image relative to a plurality of images to be compared in the second personnel file; determining the maximum value of the average values of the similarity degrees corresponding to the target images as the identity similarity degrees of the target person and the person with the known identity type corresponding to the second person file; and finally, determining the average value of the plurality of identity similarities of the target person and the persons with known identity types corresponding to the plurality of second person profiles as the identity similarity.
For example, the target image includes three images, and a certain second person file includes four images to be compared; at this time, each target image corresponds to four feature similarities, and similarity average values are calculated among the four feature similarities, so that each target image corresponds to one similarity average value, and three target images have three similarity average values in total; then, the maximum value is selected from the three similarity averages, and the maximum value is the identity similarity of the target person and the person with the known identity type corresponding to the second person profile. If a total of two second person profiles are obtained, two identity similarities are obtained, and the average value of the two identity similarities is determined as the final identity similarity of the target person and the person with the known identity type.
In the mode, the image characteristics are compared from the perspective of the file, the identity similarity is determined, the identity type of the target personnel with the identity type can be inferred by the mode, the accuracy is high, the personnel can be managed conveniently, the efficiency of finding certain personnel with specific identity types is improved, and the automatic analysis of the identity of the personnel is realized.
Example five:
the embodiment provides another method for determining the identity type of a person, and in the embodiment, a profile group is introduced, that is, a second person profile comprises a plurality of profile groups; the identity types of the second personnel files in the same file group are the same; the identity types of second personnel files of different file groups are different; the target image comprises a plurality of images; the plurality of target images are stored in the same first person profile. In this embodiment, the file groups and the identity types have a one-to-one correspondence relationship, that is, one file group corresponds to one identity type, and the target image is compared with the images to be compared in the plurality of file groups, so as to determine the identity type of the target person in the target image.
Firstly, aiming at each second person file in each group of file groups and each image to be compared in the second person file, comparing a plurality of target images in a first person file with the image to be compared to obtain a plurality of feature similarities of the first person file relative to the file groups; each image to be compared corresponds to a plurality of feature similarities, and if the second personnel file comprises a plurality of images to be compared, the number of the feature similarities corresponding to the second personnel file is the sum of the feature similarities corresponding to each image to be compared; when the profile group includes a plurality of second person profiles, the number of feature similarities corresponding to the profile group is the sum of the number of feature similarities corresponding to each second person profile. The number of the feature similarities of the first person profile with respect to the profile group is the sum of the number of the feature similarities corresponding to each second person profile.
Then, for each second person profile in each profile group, the following operations are performed: for each image to be compared in the second personnel file, acquiring the maximum value of the similarity of the image to be compared in a plurality of feature similarities relative to a plurality of target images; and determining the average value or the maximum value of the maximum values of the similarity corresponding to the images to be compared as the identity similarity of the target person and the person with the known identity type corresponding to the second person file. At this time, each second person file corresponds to an identity similarity, and when the file group includes a plurality of second person files, the file group corresponds to a plurality of identity similarities.
And finally, determining the average value of the identity similarities of the target person and the persons with known identity types corresponding to the second person profiles in the same profile group as the identity similarity of the target person and the profile group. As can be seen from the above, if the identity types corresponding to the second person files in the same file group are the same, the identity similarity between the target person and the file group, i.e. the identity similarity between the identity type of the target person and the identity type corresponding to the file group.
In addition, another implementation may be possible for determining the identity type of the target person based on the profile group.
Firstly, for each second person profile in each group of profiles, the following operations are performed: comparing the plurality of images to be compared in the second personnel file with the target image aiming at each target image to obtain a plurality of feature similarities of the plurality of images to be compared relative to the target image; the method compares image features from the dimensions of the target images to obtain a plurality of feature similarities corresponding to each target image.
Then, for each second person profile in each profile group, the following operations are performed: for each target image, obtaining a similarity average value of a plurality of feature similarities of the target image relative to a plurality of images to be compared in the second personnel file; determining the maximum value of the average values of the similarity degrees corresponding to the target images as the identity similarity degrees of the target person and the person with the known identity type corresponding to the second person file;
and finally, determining the average value of the identity similarities of the target person and the persons with known identity types corresponding to the second person profiles in the same profile group as the identity similarity of the target person and the profile group.
Based on the above, when the identity type of the target person is determined according to the identity similarity, there may be two implementation manners: selecting the highest identity similarity from the identity similarities corresponding to the multiple groups of file groups, and determining the identity type of the file group corresponding to the highest identity similarity as the identity type of the target person; for example, there are three file groups, and the identity types corresponding to the three file groups are respectively sanitation workers, security guards and takeout personnel; the identity similarity corresponding to the three file groups is respectively 0.8, 0.9 and 0.6, and at this time, the identity similarity corresponding to the file group with the security type is the highest, so that the identity type of the target person can be determined to be security.
In a second mode, aiming at each group of file groups, judging whether the identity similarity corresponding to the file group is higher than the similarity threshold corresponding to the file group; and if so, determining the identity type corresponding to the archive group as the identity type of the target person. The identity type of the target person determined by the method can be one or more; continuing with the above example, if the similarity threshold of each group of profile groups is 0.85, the identity similarity of the profile group corresponding to the security guard is higher than the similarity threshold, and at this time, the identity type of the target person may be determined to be security guard.
In the above manner, image features are compared from the perspective of file groups, and identity similarity is determined, each file group corresponding to an identity type; the method can deduce the identity type of the target personnel with the identity type through appearance characteristics, and has high accuracy, thereby being beneficial to managing the personnel, improving the efficiency of finding certain personnel with specific identity types and realizing the automatic analysis of the identity of the personnel.
Example six:
based on the method for determining the identity type of the person provided by the embodiment, the embodiment provides an application embodiment, and first, a target image in a first person file is provided with acquisition time; the first person file comprises a plurality of target images; after the step of determining the identity type of the target person according to the identity similarity, the working time of the target person can be determined according to the feature similarity corresponding to each target image and the acquisition time of the target image.
Specifically, it may be determined whether the feature similarity of each target image is higher than a preset similarity threshold; if the image is higher than the preset image, determining that the acquisition time corresponding to the target image belongs to the working time of the target personnel; and determining a time period formed by the acquisition time of the working time of the target personnel as the working time of the target personnel.
For example, suppose the target person is in a working state, wears a designated work uniform, or carries a designated work tool, or is in a designated environment; at this time, the image features of the target image containing the target person usually show the relevant features of the identity type of the target person in the working state; when the target person is in a non-working state, the target person usually wears a convenient clothes, or does not carry a working tool, or is not in a designated environment, and at the moment, the image characteristics including the target image of the target person are difficult to embody the relevant characteristics of the identity type of the target person in the working state. The image to be compared with the target image may be an image of the target person in a working state, or an image of other persons in a working state, where the identity types of the other persons in the working state are the same as those of the target person.
The target images of the target personnel can be collected at regular intervals, the collecting time of the target images is recorded, if the feature similarity of a certain target image and the image to be compared is low, the target personnel in the target image is not in a working state, and the collecting time corresponding to the target image does not belong to the working time of the target personnel. Based on the above, the working time of the target person in a day, a week or other periods can be counted, so as to check the attendance of the target person. The attendance checking mode can be used for more accurately counting the working time of the target personnel.
In another application embodiment, after the step of determining the working time of the target person according to the feature similarity corresponding to each target image and the acquisition time of the target image, the working time of the target person within a preset area range can be obtained; and then determining the number of the target personnel in the working state at the specified time point within the preset area range according to the working time of the target personnel. For example, when the identity type of the target person is security, the security factor of a certain geographic area can be evaluated by determining the number of security guards in operation in a specific time period within the geographic area in the manner described above.
In another application embodiment, the target person comprises a plurality of persons; the first person files of a plurality of target persons belong to the same preset area range; after the identity types of the target persons are determined according to the identity similarity, the designated persons can be obtained by screening the target persons in the preset area range according to the identity type corresponding to the first person file of each target person. For example, a plurality of target persons are often present in the same cell, and these target persons may be delivery persons or cell residents, and then after the identity type of each target person is determined in the above manner, the cell residents can be screened from these target persons.
Example seven:
the embodiment provides another method for determining the identity type of the person, which can be used for solving the problem that the identity type of the person is difficult to determine by a person file with unknown identity information, and the identity type of the person with unknown identity type is deduced through the file of the person with known identity type, so that the identity information of the person with unknown identity type can be mastered, and the method can be used for multidimensional research and judgment and data fusion application. As shown in fig. 4, the method includes the steps of:
step S402, acquiring a file of a person with an unknown identity type; generally, a person creates a file, wherein the file comprises an image of the person and image characteristics of the image; the image can be a human body snapshot image, a human face snapshot image and the like of a person; the file can also comprise human body characteristics, human face characteristics and the like of the personnel.
Step S404, acquiring a plurality of personnel images with known identity types; the identity type can be specifically express takeaway personnel, and can also be other identity types. If the person image includes a plurality of persons, the person corresponding to the identity type needs to be identified. In order to improve the accuracy of the person identity type inference, the number of person images of known identity types is usually large, and may be ten, for example.
Step S406, extracting image characteristics of the personnel image with known identity type; the image features may specifically include human features of a person of known identity type; and comparing the image characteristics with the image characteristics of the images in the file. If the file includes M images, the M images respectively correspond to the M image characteristics; the number of the person images with known identity types is N, and the N images correspond to N image characteristics. N rounds of comparison are needed, and when each round of comparison is conducted, the image characteristics of the person image with the known identity type are obtained and compared with the image characteristics of the images in the M files to obtain M characteristic similarities.
Step S408, selecting the highest one of the M feature similarities obtained by one comparison, and obtaining N highest feature similarities which are M1, M2, M3, … and MN; and averaging the N highest feature similarities to obtain the similarity Np between the person in the file and the person with the known identity type. It should be noted that, a plurality of feature similarities that are the highest among the M feature similarities obtained by one comparison may also be selected, for example, two feature similarities, where 2N highest feature similarities are obtained altogether. In addition, weights may be set for the selected several highest feature similarities (for example, in N person images of known identity types, 3 images have the best quality, and the weights of the 3 images corresponding to the similarities may be increased), and each feature similarity is multiplied by the corresponding weight and then summed or averaged, so as to obtain the similarity Np between the person in the archive and the person of the known identity type.
In step S410, if the similarity Np between the person in the profile and the person with the known identity type is greater than the preset threshold R, the identity type of the person in the profile can be considered to be the same as that of the person with the known identity type.
According to the steps S402-S410, whether the person in the file is the identity type of express delivery takeout personnel, sanitation personnel, security personnel, hospital staff and the like can be determined.
In the above manner, the identity type information of unknown personnel is deduced from the identity information of known personnel, so that identity attribute identification of the personnel not mastered is completed; the method can help the public security to obtain massive personnel identity type information, greatly help the public security to improve the efficiency of finding suspicious personnel, and realize the automatic analysis of a portrait big data system. Meanwhile, the method solves the problem that the traditional portrait big data system highly depends on personnel identity information provided by public security, realizes the application of big data with unknown reasoning, greatly improves the usability of personnel files, constructs a better one-person one-file, and leads the public security information to be researched and judged more accurately and more efficiently.
Example eight:
corresponding to the above method embodiment, referring to fig. 5, a schematic structural diagram of an apparatus for determining a person identity type is shown, where the apparatus includes:
the feature obtaining module 50 is configured to obtain image features of a target image including a target person and image features of an image to be compared including a person with a known identity; wherein the image features at least comprise human body appearance features of the person; the image characteristics of the images to be compared can reflect the identity types of people with known identity types;
the feature comparison module 51 is configured to compare image features of the target image and the image to be compared to obtain a feature similarity;
the identity similarity determining module 52 is configured to determine identity similarities between the target person and the persons with known identity types according to the feature similarities;
and the identity type determining module 53 is configured to determine the identity type of the target person according to the identity similarity.
The device for determining the identity type of the person obtains the image characteristics of a target image containing the target person and the image characteristics of an image to be compared containing the person with the known identity type; then comparing the image characteristics of the target image and the image to be compared to obtain characteristic similarity; and then according to the feature similarity, determining the identity similarity of the target person and the person with the known identity type, and according to the identity similarity, determining the identity type of the target person. The image features in the method at least comprise human body shape features of people, and the image features of the images to be compared can represent the identity types of people with known identity types, so that the method can predict the identity types of target people with the identity types represented by the shape features, has high accuracy, is favorable for managing the people, improves the efficiency of finding certain specific identity type people, and realizes automatic analysis of the identity of the people.
Further, the image to be compared comprises a plurality of images; each image to be compared comprises a person with a known identity type; the identity types of the persons with known identity types in the images to be compared are the same; the characteristic comparison module is further used for comparing the image characteristics of each target image with the image characteristics of each image to be compared to obtain a plurality of characteristic similarities; the identity similarity determining module is further configured to determine identity similarities between the target person and the persons with known identity types according to the multiple feature similarities.
Further, the target image includes a plurality of images; the multiple target images contain the same target person; the identity similarity determining module is further configured to, for each image to be compared, obtain a maximum similarity value among a plurality of feature similarities of the image to be compared with the plurality of target images; and determining the average value or the maximum value of the maximum values of the similarity corresponding to the images to be compared as the identity similarity.
Further, the target image includes a plurality of images; the multiple target images contain the same target person; the identity similarity determining module is further configured to, for each target image, obtain a similarity average value of a plurality of feature similarities of the target image with respect to a plurality of images to be compared; and determining the maximum value of the average value of the similarity degrees corresponding to the target images as the identity similarity.
Further, the target image includes a plurality of images; the multiple target images contain the same target person; the feature comparison module is further configured to compare, for each target image, image features of the target image and an image to be compared to obtain a feature similarity corresponding to the target image; the identity similarity determining module is further configured to determine a maximum similarity value among the feature similarities corresponding to the plurality of target images as the identity similarity between the target person and the person with the known identity type.
Further, the target image containing the target person is stored in a first person file of the target person in advance; if the target image comprises a plurality of target images, the target images comprising the same person in the plurality of target images are stored in the same first person file.
Further, the image to be compared containing the person with the known identity type is stored in a second person file of the person with the known identity type in advance; if the image to be compared comprises a plurality of images, the images to be compared containing the same person are stored in the same second person file.
Further, the image to be compared comprises a plurality of images; storing a plurality of images to be compared in at least two second personnel files; the identity types of the people with known identity types contained in the images to be compared in the at least two second person files are the same; the target image comprises a plurality of images; storing a plurality of target images in the same first person file; the feature comparison module is further configured to, for each second person file, perform the following operations: aiming at each image to be compared in the second personnel file, comparing a plurality of target images with the image to be compared to obtain a plurality of feature similarities of the target images relative to the image to be compared; the identity similarity determination module is further configured to, for each second person profile, perform the following operations: for each image to be compared in the second personnel file, acquiring the maximum value of the similarity of the image to be compared in a plurality of feature similarities relative to a plurality of target images; determining the average value or the maximum value of a plurality of similarity maximum values corresponding to a plurality of images to be compared in the second personnel file as the identity similarity of the target personnel and the personnel with known identity types corresponding to the second personnel file; and determining the average value of the plurality of identity similarities of the target person and the persons with known identity types corresponding to the plurality of second person profiles as the identity similarity.
Further, the image to be compared comprises a plurality of images; storing a plurality of images to be compared in at least two second personnel files; the identity types of the people with known identity types contained in the images to be compared in the at least two second person files are the same; the target image comprises a plurality of images; storing a plurality of target images in the same first person file; the feature comparison module is further configured to, for each second person file, perform the following operations: comparing the plurality of images to be compared in the second personnel file with the target image aiming at each target image to obtain a plurality of feature similarities of the plurality of images to be compared relative to the target image; the identity similarity determination module is further configured to, for each second person profile, perform the following operations: for each target image, obtaining a similarity average value of a plurality of feature similarities of the target image relative to a plurality of images to be compared in the second personnel file; determining the maximum value of the average values of the similarity degrees corresponding to the target images as the identity similarity degrees of the target person and the person with the known identity type corresponding to the second person file; and determining the average value of the plurality of identity similarities of the target person and the persons with known identity types corresponding to the plurality of second person profiles as the identity similarity.
Further, the identity type determining module is further configured to determine that the identity type of the target person is a known identity type if the identity similarity is higher than a preset similarity threshold.
Further, the second person profile comprises a plurality of profile groups; the identity types of the second personnel files in the same file group are the same; the identity types of second personnel files of different file groups are different; the target image comprises a plurality of images; storing a plurality of target images in the same first person file; the characteristic comparison module is further configured to compare, for each second person file in each group of file groups, a plurality of target images in a first person file with each image to be compared in the second person file, so as to obtain a plurality of characteristic similarities of the first person file with respect to the file group; the identity similarity determination module is further configured to, for each second person profile in each profile group, perform the following operations: for each image to be compared in the second personnel file, acquiring the maximum value of the similarity of the image to be compared in a plurality of feature similarities relative to a plurality of target images; determining the average value or the maximum value of a plurality of similarity maximum values corresponding to a plurality of images to be compared as the identity similarity of the target person and the person with the known identity type corresponding to the second person file; and determining the average value of the identity similarities of the target person and the persons with known identity types corresponding to the second person profiles in the same profile group as the identity similarity of the target person and the profile group.
Further, the second person profile comprises a plurality of profile groups; the identity types of the second personnel files in the same file group are the same; the identity types of second personnel files of different file groups are different; the target image comprises a plurality of images; storing a plurality of target images in the same first person file; the characteristic comparison module is further configured to execute the following operations for each second person file in each group of file groups: comparing the plurality of images to be compared in the second personnel file with the target image aiming at each target image to obtain a plurality of feature similarities of the plurality of images to be compared relative to the target image; the identity similarity determination module is further configured to, for each second person profile in each profile group, perform the following operations: for each target image, obtaining a similarity average value of a plurality of feature similarities of the target image relative to a plurality of images to be compared in the second personnel file; determining the maximum value of the average values of the similarity degrees corresponding to the target images as the identity similarity degrees of the target person and the person with the known identity type corresponding to the second person file; and determining the average value of the identity similarities of the target person and the persons with known identity types corresponding to the second person profiles in the same profile group as the identity similarity of the target person and the profile group.
Further, the identity type determining module is further configured to select the highest identity similarity from the identity similarities corresponding to the multiple groups of file groups, and determine the identity type of the file group corresponding to the highest identity similarity as the identity type of the target person; or, aiming at each group of file groups, judging whether the identity similarity corresponding to the file group is higher than the similarity threshold corresponding to the file group; and if so, determining the identity type corresponding to the archive group as the identity type of the target person.
Further, the target image in the first person file is provided with acquisition time; the first person file comprises a plurality of target images; the device also comprises a working time determining module which is used for determining the working time of the target personnel according to the corresponding feature similarity of each target image and the acquisition time of the target image.
Further, the working time determining module is further configured to determine, for each target image, whether the feature similarity of the target image is higher than a preset similarity threshold; if the image is higher than the preset image, determining that the acquisition time corresponding to the target image belongs to the working time of the target personnel; and determining a time period formed by the acquisition time of the working time of the target personnel as the working time of the target personnel.
The device also comprises a personnel number determining module, a judging module and a judging module, wherein the personnel number determining module is used for acquiring the working time of the target personnel in the preset area range; and determining the number of the target personnel in the working state at the specified time point within the preset area range according to the working time of the target personnel.
Further, the target person includes a plurality of persons; the first person files of a plurality of target persons belong to the same preset area range; the device further comprises a screening module for screening the designated persons from the target persons within the preset area range according to the identity type corresponding to the first person file of each target person.
The present embodiment further provides an electronic system, including: a processing device and a storage device; the storage means has stored thereon a computer program which, when being executed by the processing device, performs the method of determining the identity type of a person as described above.
The present embodiment also provides a machine-readable storage medium, on which a computer program is stored, which when executed by a processing device performs the steps of determining the identity type of a person as described above.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the system and the apparatus described above may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In addition, in the description of the embodiments of the present invention, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly, e.g., as meaning either a fixed connection, a removable connection, or an integral connection; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meaning of the above terms in the present invention can be understood in specific cases for those skilled in the art.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
Finally, it should be noted that: although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art will understand that the following embodiments are merely illustrative of the present invention, and not restrictive, and the scope of the present invention is not limited thereto: any person skilled in the art can modify or easily conceive the technical solutions described in the foregoing embodiments or equivalent substitutes for some technical features within the technical scope of the present disclosure; such modifications, changes or substitutions do not depart from the spirit and scope of the embodiments of the present invention, and they should be construed as being included therein. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims (20)

1. A method of determining a person identity type, the method comprising:
acquiring image characteristics of a target image containing target personnel and image characteristics of an image to be compared containing personnel with known identity types; wherein the image features comprise at least human body contour features of the person; the image characteristics of the images to be compared can reflect the identity types of the people with known identity types;
comparing the image characteristics of the target image and the image to be compared to obtain characteristic similarity;
determining identity similarity of the target person and the known identity type person according to the feature similarity;
and determining the identity type of the target person according to the identity similarity.
2. The method according to claim 1, wherein the images to be compared comprise a plurality of images; each image to be compared comprises a person with a known identity type; the identity types of the persons with known identity types contained in the images to be compared are the same;
the step of comparing the image characteristics of the target image and the image to be compared to obtain the characteristic similarity comprises the following steps: comparing the image characteristics of each target image and each image to be compared to obtain a plurality of characteristic similarities;
the step of determining the identity similarity between the target person and the person with the known identity type according to the feature similarity comprises the following steps: and determining the identity similarity of the target person and the known identity type person according to the plurality of feature similarities.
3. The method of claim 2, wherein the target image comprises a plurality of sheets; a plurality of target images contain the same target person;
the step of determining identity similarities of the target person and the known identity type person according to the feature similarities includes:
for each image to be compared, acquiring the maximum value of the similarity of the image to be compared in a plurality of feature similarities relative to a plurality of target images;
and determining the average value or the maximum value of the similarity maximum values corresponding to the images to be compared as the identity similarity.
4. The method of claim 2, wherein the target image comprises a plurality of sheets; a plurality of target images contain the same target person;
the step of determining identity similarities of the target person and the known identity type person according to the feature similarities includes:
for each target image, acquiring a similarity average value of a plurality of feature similarities of the target image relative to a plurality of images to be compared;
and determining the maximum value of the similarity average values corresponding to the target images as the identity similarity.
5. The method of claim 1, wherein the target image comprises a plurality of sheets; a plurality of target images contain the same target person;
the step of comparing the image characteristics of the target image and the image to be compared to obtain the characteristic similarity comprises the following steps: comparing the image characteristics of the target image and the image to be compared to obtain the characteristic similarity corresponding to the target image aiming at each target image;
the step of determining the identity similarity between the target person and the person with the known identity type according to the feature similarity comprises the following steps: and determining the maximum similarity value in the feature similarities corresponding to the target images as the identity similarity of the target person and the person with the known identity type.
6. The method according to any one of claims 1 to 5, wherein the target image containing the target person is stored in advance in a first person profile of the target person;
if the target image comprises a plurality of target images, the target images comprising the same person in the plurality of target images are stored in the same first person file.
7. The method according to claim 6, wherein the image to be compared containing the person with known identity type is pre-stored in a second person profile of the person with known identity type;
if the image to be compared comprises a plurality of images, the image to be compared comprising the same person is stored in the same second person file.
8. The method according to claim 7, wherein the images to be compared comprise a plurality of images; storing a plurality of images to be compared in at least two second personnel files; the identity types of the people with known identity types contained in the images to be compared in at least two second person files are the same; the target image comprises a plurality of images; storing a plurality of target images in the same first person file;
the step of comparing the image characteristics of the target image and the image to be compared to obtain the characteristic similarity comprises the following steps: for each of the second person profiles, performing the following operations: aiming at each image to be compared in the second personnel file, comparing a plurality of target images with the image to be compared to obtain a plurality of feature similarities of the target images relative to the image to be compared;
the step of determining the identity similarity between the target person and the person with the known identity type according to the feature similarity comprises the following steps:
for each of the second person profiles, performing the following: for each image to be compared in the second personnel file, acquiring the maximum value of the similarity of the image to be compared in a plurality of feature similarities relative to a plurality of target images; determining the average value or the maximum value of the maximum values of the similarity corresponding to the images to be compared in the second personnel file as the identity similarity of the target personnel and the personnel with the known identity type corresponding to the second personnel file;
and determining the average value of the identity similarities of the target person and the persons with known identity types corresponding to the second person profiles as the identity similarity.
9. The method according to claim 7, wherein the images to be compared comprise a plurality of images; storing a plurality of images to be compared in at least two second personnel files; the identity types of the people with known identity types contained in the images to be compared in at least two second person files are the same; the target image comprises a plurality of images; storing a plurality of target images in the same first person file;
the step of comparing the image characteristics of the target image and the image to be compared to obtain the characteristic similarity comprises the following steps: for each of the second person profiles, performing the following operations: comparing the plurality of images to be compared in the second personnel file with the target image aiming at each target image to obtain a plurality of feature similarities of the plurality of images to be compared relative to the target image;
the step of determining the identity similarity between the target person and the person with the known identity type according to the feature similarity comprises the following steps:
for each of the second person profiles, performing the following: for each target image, obtaining a similarity average value of a plurality of feature similarities of the target image relative to a plurality of images to be compared in the second personnel file; determining the maximum value of the average value of the similarity degrees corresponding to the target images as the identity similarity degree of the target person and the person with the known identity type corresponding to the second person file;
and determining the average value of the identity similarities of the target person and the persons with known identity types corresponding to the second person profiles as the identity similarity.
10. The method of claim 1, wherein the step of determining the identity type of the target person according to the identity similarity comprises:
and if the identity similarity is higher than a preset similarity threshold, determining the identity type of the target person as the known identity type.
11. The method of any of claims 7-9, wherein the second person profile comprises a plurality of profile groups; the identity types of the second personnel files in the same file group are the same; the identity types of second personnel files of different file groups are different; the target image comprises a plurality of images; storing a plurality of target images in the same first person file;
the step of comparing the image characteristics of the target image and the image to be compared to obtain the characteristic similarity comprises the following steps: aiming at each second person file in each group of file groups and each image to be compared in the second person file, comparing a plurality of target images in the first person file with the image to be compared to obtain a plurality of feature similarities of the first person file relative to the file groups;
the step of determining the identity similarity between the target person and the person with the known identity type according to the feature similarity comprises the following steps:
for each of the second person profiles in each of the profile groups, performing the following operations:
for each image to be compared in the second personnel file, acquiring the maximum value of the similarity of the image to be compared in a plurality of feature similarities relative to a plurality of target images; determining the average value or the maximum value of the maximum values of the similarity corresponding to the images to be compared as the identity similarity of the target person and the person with the known identity type corresponding to the second person file;
and determining the average value of the identity similarities of the target person and the persons with known identity types corresponding to the second person profiles in the same profile group as the identity similarity of the target person and the profile group.
12. The method of any of claims 7-9, wherein the second person profile comprises a plurality of profile groups; the identity types of the second personnel files in the same file group are the same; the identity types of second personnel files of different file groups are different; the target image comprises a plurality of images; storing a plurality of target images in the same first person file;
the step of comparing the image characteristics of the target image and the image to be compared to obtain the characteristic similarity comprises the following steps: for each second person profile in each group of profiles group, performing the following operations: comparing the plurality of images to be compared in the second personnel file with the target image aiming at each target image to obtain a plurality of feature similarities of the plurality of images to be compared relative to the target image;
the step of determining the identity similarity between the target person and the person with the known identity type according to the feature similarity comprises the following steps:
for each of the second person profiles in each of the profile groups, performing the following operations:
for each target image, obtaining a similarity average value of a plurality of feature similarities of the target image relative to a plurality of images to be compared in the second personnel file; determining the maximum value of the average value of the similarity degrees corresponding to the target images as the identity similarity degree of the target person and the person with the known identity type corresponding to the second person file;
and determining the average value of the identity similarities of the target person and the persons with known identity types corresponding to the second person profiles in the same profile group as the identity similarity of the target person and the profile group.
13. The method according to claim 11 or 12, wherein the step of determining the identity type of the target person according to the identity similarity comprises:
selecting the highest identity similarity from the identity similarities corresponding to the multiple groups of archive groups, and determining the identity type of the archive group corresponding to the highest identity similarity as the identity type of the target person;
or, aiming at each group of file groups, judging whether the identity similarity corresponding to the file group is higher than the similarity threshold corresponding to the file group; and if so, determining the identity type corresponding to the archive group as the identity type of the target person.
14. The method of claim 6, wherein a target image in the first person profile is provided with an acquisition time; the first person file comprises a plurality of target images;
after the step of determining the identity type of the target person according to the identity similarity, the method further includes:
and determining the working time of the target personnel according to the corresponding feature similarity of each target image and the acquisition time of the target image.
15. The method according to claim 14, wherein the step of determining the working time of the target person according to the feature similarity corresponding to each target image and the acquisition time of the target image comprises:
judging whether the feature similarity of each target image is higher than a preset similarity threshold value or not; if the target image is higher than the preset target image, determining that the acquisition time corresponding to the target image belongs to the working time of the target personnel;
and determining a time period formed by the acquisition time of the working time of the target person as the working time of the target person.
16. The method according to claim 14, wherein after the step of determining the working time of the target person according to the corresponding feature similarity of each target image and the acquisition time of the target image, the method further comprises:
acquiring the working time of target personnel within a preset area range;
and determining the number of the target personnel in the working state at the specified time point within the preset area range according to the working time of the target personnel.
17. The method of claim 6, wherein the target person comprises a plurality; the first person files of the target persons belong to the same preset area range;
after the step of determining the identity type of the target person according to the identity similarity, the method further includes:
and screening the target persons in the preset area range to obtain the designated persons according to the identity type corresponding to the first person file of each target person.
18. An apparatus for determining the identity type of a person, the apparatus comprising:
the characteristic acquisition module is used for acquiring image characteristics of a target image containing target personnel and image characteristics of an image to be compared containing personnel with known identity types; wherein the image features comprise at least human body contour features of the person; the image characteristics of the images to be compared can reflect the identity types of the people with known identity types;
the characteristic comparison module is used for comparing the image characteristics of the target image and the image to be compared to obtain characteristic similarity;
the identity similarity determining module is used for determining the identity similarity of the target person and the known identity type person according to the feature similarity;
and the identity type determining module is used for determining the identity type of the target person according to the identity similarity.
19. An electronic system, characterized in that the electronic system comprises: a processing device and a storage device;
the storage means has stored thereon a computer program which, when being executed by the processing device, carries out the method of determining a person identity type according to any one of claims 1-17.
20. A machine readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processing device, performs the steps of determining a person identity type according to any one of claims 1-17.
CN202010336918.9A 2020-04-24 2020-04-24 Method, device and electronic system for determining personnel identity type Pending CN111666441A (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN202010336918.9A CN111666441A (en) 2020-04-24 2020-04-24 Method, device and electronic system for determining personnel identity type
PCT/CN2020/119614 WO2021212760A1 (en) 2020-04-24 2020-09-30 Method and apparatus for determining identity type of person, and electronic system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010336918.9A CN111666441A (en) 2020-04-24 2020-04-24 Method, device and electronic system for determining personnel identity type

Publications (1)

Publication Number Publication Date
CN111666441A true CN111666441A (en) 2020-09-15

Family

ID=72382791

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010336918.9A Pending CN111666441A (en) 2020-04-24 2020-04-24 Method, device and electronic system for determining personnel identity type

Country Status (2)

Country Link
CN (1) CN111666441A (en)
WO (1) WO2021212760A1 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112258363A (en) * 2020-10-16 2021-01-22 浙江大华技术股份有限公司 Identity information confirmation method and device, storage medium and electronic device
WO2021212760A1 (en) * 2020-04-24 2021-10-28 北京旷视科技有限公司 Method and apparatus for determining identity type of person, and electronic system

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103207870A (en) * 2012-01-17 2013-07-17 华为技术有限公司 Method, server, device and system for photo sort management
CN109726765A (en) * 2019-01-02 2019-05-07 京东方科技集团股份有限公司 A kind of sample extraction method and device of visual classification problem
WO2019153193A1 (en) * 2018-02-08 2019-08-15 深圳前海达闼云端智能科技有限公司 Taxi operation monitoring method, device, storage medium, and system
CN110378092A (en) * 2019-07-26 2019-10-25 北京积加科技有限公司 Identification system and client, server and method
CN110730330A (en) * 2019-09-27 2020-01-24 深圳市大拿科技有限公司 Sound processing method and related product

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP6448223B2 (en) * 2014-06-12 2019-01-09 キヤノン株式会社 Image recognition system, image recognition apparatus, image recognition method, and computer program
CN106577350B (en) * 2016-11-22 2020-10-09 深圳市沃特沃德股份有限公司 Pet type identification method and device
CN111666441A (en) * 2020-04-24 2020-09-15 北京旷视科技有限公司 Method, device and electronic system for determining personnel identity type

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103207870A (en) * 2012-01-17 2013-07-17 华为技术有限公司 Method, server, device and system for photo sort management
WO2019153193A1 (en) * 2018-02-08 2019-08-15 深圳前海达闼云端智能科技有限公司 Taxi operation monitoring method, device, storage medium, and system
CN109726765A (en) * 2019-01-02 2019-05-07 京东方科技集团股份有限公司 A kind of sample extraction method and device of visual classification problem
CN110378092A (en) * 2019-07-26 2019-10-25 北京积加科技有限公司 Identification system and client, server and method
CN110730330A (en) * 2019-09-27 2020-01-24 深圳市大拿科技有限公司 Sound processing method and related product

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021212760A1 (en) * 2020-04-24 2021-10-28 北京旷视科技有限公司 Method and apparatus for determining identity type of person, and electronic system
CN112258363A (en) * 2020-10-16 2021-01-22 浙江大华技术股份有限公司 Identity information confirmation method and device, storage medium and electronic device

Also Published As

Publication number Publication date
WO2021212760A1 (en) 2021-10-28

Similar Documents

Publication Publication Date Title
CN110807385B (en) Target detection method, target detection device, electronic equipment and storage medium
CN104246786B (en) Field selection in mode discovery
US20170300744A1 (en) Method and apparatus for determining identity identifier of face in face image, and terminal
CN112001932B (en) Face recognition method, device, computer equipment and storage medium
CN109325429B (en) Method, device, storage medium and terminal for associating feature data
CN106709047B (en) Object searching method and device
CN114581491B (en) Pedestrian trajectory tracking method, system and related device
CN112016623B (en) Face clustering method, device, equipment and storage medium
CN109426785A (en) A kind of human body target personal identification method and device
CN111221991B (en) Method and device for determining personnel identity attribute and electronic equipment
CN111079648A (en) Data set cleaning method and device and electronic system
WO2020172870A1 (en) Method and apparatus for determining motion trajectory of target object
CN111666441A (en) Method, device and electronic system for determining personnel identity type
CN106504162A (en) Same pedestrian's association analysis method and device based on station MAC scan datas
CN111753826B (en) Vehicle and license plate association method, device and electronic system
CN112966652A (en) Trajectory convergence method and device, computer equipment and storage medium
CN109241316B (en) Image retrieval method, image retrieval device, electronic equipment and storage medium
CN114124484B (en) Network attack identification method, system, device, terminal equipment and storage medium
CN110263830B (en) Image processing method, device and system and storage medium
CN109886239A (en) Portrait clustering method, apparatus and system
CN115062186A (en) Video content retrieval method, device, equipment and storage medium
CN114048344A (en) Similar face searching method, device, equipment and readable storage medium
CN110825893A (en) Target searching method, device, system and storage medium
CN113674152A (en) Image processing method, image processing device, electronic equipment and computer readable storage medium
CN110968719B (en) Face clustering method and device

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination