WO2019128883A1 - 一种身份标定系统和方法 - Google Patents

一种身份标定系统和方法 Download PDF

Info

Publication number
WO2019128883A1
WO2019128883A1 PCT/CN2018/122883 CN2018122883W WO2019128883A1 WO 2019128883 A1 WO2019128883 A1 WO 2019128883A1 CN 2018122883 W CN2018122883 W CN 2018122883W WO 2019128883 A1 WO2019128883 A1 WO 2019128883A1
Authority
WO
WIPO (PCT)
Prior art keywords
identity information
node
node identity
face feature
feature
Prior art date
Application number
PCT/CN2018/122883
Other languages
English (en)
French (fr)
Inventor
郑天航
胡飏
颜王辉
王巍
林彬
Original Assignee
苏州欧普照明有限公司
欧普照明股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from CN201711446279.6A external-priority patent/CN108009530B/zh
Priority claimed from CN201721872711.3U external-priority patent/CN207764814U/zh
Application filed by 苏州欧普照明有限公司, 欧普照明股份有限公司 filed Critical 苏州欧普照明有限公司
Publication of WO2019128883A1 publication Critical patent/WO2019128883A1/zh
Priority to US16/910,040 priority Critical patent/US11113509B2/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/161—Detection; Localisation; Normalisation
    • G06V40/166—Detection; Localisation; Normalisation using acquisition arrangements
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/40—Scenes; Scene-specific elements in video content
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/168—Feature extraction; Face representation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/172—Classification, e.g. identification

Definitions

  • the invention relates to the field of information processing technology, in particular to an identity calibration system and method.
  • the current statistics of shopping mall traffic and the movement tracking of passenger flow are based on face recognition technology. Because face recognition requires high computing power, face recognition is performed by sending the video stream captured by the camera to the background server. The recognition speed can be improved. After receiving the video stream, the background server analyzes the information in the video, can identify and extract relevant feature values and store the comparison, and then display the relevant analysis information. Its specific architecture is shown in Figure 1.
  • the problem with the system shown in FIG. 1 is that the video stream is an application that occupies bandwidth.
  • the network load is large, which has a great influence on the stability and accuracy of the system, and the background server analyzes the multi-channel video.
  • the background server analyzes the multi-channel video.
  • the video stream is uploaded to the backend server, it also involves user privacy issues. Therefore, these problems need to be solved urgently.
  • the present invention has been made in order to provide an identity calibration system and method that overcomes the above problems or at least partially solves the above problems.
  • the identity calibration system improves system efficiency and security through distributed computing of compute nodes and servers. .
  • An embodiment of the present invention provides an identity calibration system, including: a computing node including an image collection device, and a server that establishes a communication connection with the computing node;
  • the computing node is configured to use the image capturing device thereon to collect an image according to a preset image capturing frequency; perform face recognition on the collected image, identify a facial feature in the image, and according to the recognized facial feature Determining node identity information corresponding to the face feature; and then uploading the node identity information corresponding to the face feature and the face feature to the server;
  • the server is configured to calibrate the identity of the facial feature based on the facial feature uploaded by the computing node and the node identity information corresponding to the facial feature.
  • the computing node is further configured to:
  • the node identity information corresponding to the matching facial feature in the corresponding relationship is determined as the node identity information corresponding to the recognized facial feature;
  • the computing node is further configured to:
  • the node identity information After determining the node identity information corresponding to the face feature, the node identity information, the appearance time of the node identity information, and the face center coordinates corresponding to the node identity information are saved in the local node log table.
  • the computing node is further configured to:
  • the node identity information in the node log table is different, the first node identity information and the second node identity information are recorded;
  • the first node identity information and the second node identity information are used.
  • the node identity information is determined to be substantially the same and recorded in the node identity information merge table.
  • the server is further configured to:
  • the server is further configured to:
  • the face feature information corresponding to the face feature uploaded by any of the plurality of computing nodes and the node identity information corresponding to the face feature is received again, it is determined whether the node identity information that is the same as the uploaded node identity information exists in the mapping relationship, If yes, the calibration identity information corresponding to the same node identity information is used as the calibration identity of the uploaded node identity information according to the mapping relationship; if not, assigning a new identity different from the calibration identity information included in the mapping relationship
  • the identity information is calibrated, the new identifiable identity information is used as the calibration identity of the uploaded node identity information, and the new calibrated identity information, the uploaded node identity information, and the corresponding facial features are recorded into the mapping relationship.
  • the computing node is further configured to: upload the node log table to the server;
  • the server is further configured to: when receiving the face feature and the node identity information corresponding to the face feature uploaded by any of the plurality of computing nodes, according to the node log table, if the arbitrary computing node is If the number of identity information of the same node determined in the preset duration is greater than the preset threshold, the same node identity information is calibrated as valid identity information, and the node identity information calibrated as valid identity information is further determined whether the mapping relationship exists.
  • the node identity information that is the same as the valid identity information if yes, the calibration identity information corresponding to the same node identity information is used as the calibration identity of the valid identity information according to the mapping relationship; if not, the assignment is in the mapping relationship
  • the new calibration identity information includes different calibration identity information, the new calibration identity information is used as the calibration identity of the valid identity information, and the new calibration identity information, the valid identity information, and the corresponding facial features are recorded into the mapping relationship.
  • the server is further configured to:
  • querying node identity information of the node identity information merge table corresponding to the newly calibrated identity information is substantially the same node identity information, and adding the substantially identical node identity information to In the mapping relationship.
  • the computing node is further configured to:
  • the computing node is further configured to:
  • Uploading according to the preset uploading frequency, the node identity information corresponding to the local face feature and the face feature to the server, and generating an upload log, so as to save the subsequent upload locally and not upload according to the upload log.
  • the node identity information corresponding to the face feature and the face feature of the server is uploaded to the server.
  • An embodiment of the present invention further provides an identity calibration method, which is applied to a computing node including an image collection device, and the method includes:
  • the node identity information corresponding to the face feature and the face feature is uploaded to the server, so that the server calibrates the identity of the face feature based on the face feature uploaded by the computing node and the node identity information corresponding to the face feature.
  • the determining, according to the identified facial features, the node identity information corresponding to the facial features including:
  • the node identity information corresponding to the matching facial feature in the corresponding relationship is determined as the node identity information corresponding to the recognized facial feature;
  • the method further includes:
  • the node identity information, the appearance time of the node identity information, and the face center coordinates corresponding to the node identity information are saved in the local node log table.
  • the method further includes:
  • the node identity information in the node log table is different, the first node identity information and the second node identity information are recorded;
  • the first node identity information and the second node identity information are used.
  • the node identity information is determined to be substantially the same and recorded in the node identity information merge table.
  • the method further includes:
  • performing face recognition on the collected image, identifying a facial feature in the image, and determining node identity information corresponding to the facial feature according to the recognized facial feature including:
  • the uploading the node identity information corresponding to the face feature and the face feature to the server includes:
  • Uploading according to the preset uploading frequency, the node identity information corresponding to the local face feature and the face feature to the server, and generating an upload log, so as to save the subsequent upload locally and not upload according to the upload log.
  • the node identity information corresponding to the face feature and the face feature of the server is uploaded to the server.
  • the embodiment of the invention further provides an identity calibration method, which is applied to a server, and includes:
  • the computing node including the image collecting device, the face feature and the node identity information corresponding to the face feature; wherein the computing node uses the image capturing device thereon to collect the image according to the preset image capturing frequency; and the collected image Performing face recognition, identifying facial features in the image, and determining node identity information corresponding to the facial features according to the recognized facial features; and then uploading the node identity information corresponding to the facial features and the facial features to The server;
  • the identity of the face feature is calibrated based on the face feature uploaded by the computing node and the node identity information corresponding to the face feature.
  • the identity of the facial feature is calibrated based on the facial feature uploaded by the computing node and the node identity information corresponding to the facial feature, including:
  • the identity of the facial feature is calibrated based on the facial feature uploaded by the computing node and the node identity information corresponding to the facial feature, including:
  • the face feature information corresponding to the face feature uploaded by any of the plurality of computing nodes and the node identity information corresponding to the face feature is received again, it is determined whether the node identity information that is the same as the uploaded node identity information exists in the mapping relationship, If yes, the calibration identity information corresponding to the same node identity information is used as the calibration identity of the uploaded node identity information according to the mapping relationship; if not, assigning a new identity different from the calibration identity information included in the mapping relationship
  • the identity information is calibrated, the new identifiable identity information is used as the calibration identity of the uploaded node identity information, and the new calibrated identity information, the uploaded node identity information, and the corresponding facial features are recorded into the mapping relationship.
  • the identity of the facial feature is calibrated based on the facial feature uploaded by the computing node and the node identity information corresponding to the facial feature, including:
  • the node feature information corresponding to the face feature and the face feature uploaded by any of the plurality of computing nodes is received again, according to the node log table uploaded by the computing node, if the arbitrary computing node is at a preset duration If the number of identity information of the same node determined in the same is greater than the preset threshold, the same node identity information is calibrated as valid identity information, and the node identity information calibrated as valid identity information is further determined whether the mapping relationship exists and the valid identity
  • the node identity information with the same information, if present, the calibration identity information corresponding to the same node identity information is used as the calibration identity of the valid identity information according to the mapping relationship; if not, the calibration and the calibration included in the mapping relationship are included
  • the newly-identified identity information with different identity information uses the newly-identified identity information as the calibration identity of the valid identity information, and records the new calibration identity information, the valid identity information, and the corresponding facial features into the mapping relationship.
  • the method further includes:
  • querying node identity information of the node identity information merge table corresponding to the newly calibrated identity information is substantially the same node identity information, and adding the substantially identical node identity information to In the mapping relationship.
  • the identity calibration system includes a computing node including an image collection device, and a server that establishes a communication connection with the computing node. After the computing node collects an image and performs face recognition of the image, the facial feature and the analysis data are extracted. Uploading to the server; the server performs feature matching and association to realize identity calibration of the facial features; thus, the distributed computing of the computing node and the server improves the efficiency and security of the system.
  • the embodiment of the present invention further analyzes the node identity information determined by the recognized face feature on the basis of the image analysis algorithm of the face recognition, and solves the problem that the person is within the specified range of the same location point, because the face is The angle and the light will determine the problem of different node identity information, improve the accuracy of the calculation of the node, and thus improve the accuracy of the identity calibration of the entire system.
  • the server by determining the validity of the node identity information of the computing node, the solution is solved within a specified range of the same location point, and the node identity information is determined due to the angle of the face and the light.
  • the problem has improved the accuracy of the identity calibration of the entire system.
  • FIG. 1 is a schematic diagram of a system architecture of hotspot analysis and mobile trajectory tracking in the prior art
  • FIG. 2 is a schematic structural diagram of an identity calibration system according to an embodiment of the invention.
  • FIG. 3 is a schematic flow chart of an identity calibration method applied to a computing node according to an embodiment of the invention
  • FIG. 4 is a schematic diagram showing a T-type rule of node identity information according to an embodiment of the invention.
  • FIG. 5 is a flow chart showing an identity calibration method applied to a server according to an embodiment of the present invention.
  • FIG. 6 is a flow chart showing an identity calibration method applied to a server according to another embodiment of the present invention.
  • an embodiment of the present invention provides an identity calibration system that improves system efficiency and security through distributed computing of compute nodes and servers.
  • the identity calibration system provided by the embodiment of the present invention may include a computing node 210, and a server 220 that establishes a communication connection with the computing node 210, specifically:
  • the computing node 210 includes an image capturing device 211 for acquiring an image according to a preset image capturing frequency by using the image capturing device 211 thereon; performing face recognition on the collected image, identifying a facial feature in the image, and identifying the The facial feature is determined, and the node identity information corresponding to the face feature is determined; afterwards, the node identity information corresponding to the face feature and the face feature is uploaded to the server 220;
  • the server 220 is configured to calibrate the identity of the facial feature based on the facial features uploaded by the computing node 210 and the node identity information corresponding to the facial features.
  • the number of the computing nodes 210 in the identity calibration system provided by the embodiment of the present invention may be one or more, and the four computing nodes 210 in FIG. 2 are only schematic and are not implemented in the embodiment of the present invention. limit.
  • the preset image acquisition frequency may be set or adjusted according to actual needs, for example, the preset image acquisition frequency may be 3 frames for 1 second, or 4 frames for 1 second, etc., The embodiment does not limit this.
  • the computing node 210 performs face recognition on the captured image to identify a facial feature in the image.
  • the face recognition algorithm may be used to perform face recognition on the collected image, and the detected image is detected. After the face is positioned and the key feature points of the face are located, the face region is extracted for preprocessing, and facial features such as face center coordinates, facial features, face shape, angle, and the like are extracted.
  • the specific information may be:
  • the recognized face feature is matched with the face feature in the corresponding relationship to obtain a matching result
  • the node identity information corresponding to the matching facial feature in the corresponding relationship is determined as the node identity information corresponding to the recognized facial feature;
  • the matching result is that there is no matching facial feature in the correspondence relationship, assigning a new node identity information different from the node identity information included in the corresponding relationship, and determining the new node identity information as the node corresponding to the recognized facial feature.
  • Identity information and record the new node identity information and the corresponding face features into the corresponding relationship.
  • the node identity information herein includes the identifier of the computing node, which is the identity information indicating the face feature on the computing node, that is, the corresponding relationship on each node is separately generated and stored, for example, the computing node 210 in the identity calibration system includes Calculating nodes A, B, C, and D, assuming that the face feature a is recognized from the images acquired by the computing nodes A and B, the corresponding node identity information of the face feature a on the computing node A is the node Aa, the face The corresponding node identity information of feature a on compute node B is node Ba.
  • Table 1 The correspondence between the face features of the compute node A and the node identity information is illustrated in Table 1 below.
  • the compute nodes B, C, and D also have a correspondence table similar to that shown in Table 1. It should be noted that the examples herein are merely illustrative and are not intended to limit the embodiments of the present invention.
  • the preset threshold may be set according to actual requirements. For example, the matching degree is measured by a value of 0 to 100, and the preset threshold may be 75 or 80, etc., it should be noted that the example here is only illustrative, and The embodiments of the present invention are not limited.
  • the computing node 210 when performing face recognition on the captured image, if the image includes multiple faces, the computing node 210 starts the same number of multiple working threads as the plurality of faces; a worker thread identifying a face feature of each face in the image, and determining node identity information corresponding to the face feature according to the recognized face feature, wherein a worker thread identifies a face feature of a face, and The node identity information corresponding to the face feature is determined according to the recognized face feature.
  • the multi-threaded parallel operation can improve the processing efficiency of the computing node, thereby improving the processing efficiency of the entire identity calibration system.
  • the computing node 210 may upload in real time, specifically:
  • the computing node 210 uploads the node identity information corresponding to the face feature and the face feature to the server 220 in real time after determining the node identity information corresponding to the face feature.
  • the computing node 210 may also upload according to the preset upload frequency, specifically:
  • the computing node 210 After determining the node identity information corresponding to the face feature, the computing node 210 saves the node identity information corresponding to the face feature and the face feature locally;
  • the node identity information corresponding to the local face feature and the face feature saved locally is uploaded to the server 220 according to the preset upload frequency.
  • the preset upload frequency can be set according to actual needs.
  • the preset upload frequency can be uploaded every 1 minute, or every 2 minutes. It should be noted that the example here is only illustrative, and The embodiments of the present invention are not limited.
  • the computing node 210 is further configured to upload the node identity information corresponding to the local face feature and the face feature saved to the server 220 according to the preset upload frequency, and upload the file to the server.
  • the node identity information corresponding to the face feature and the face feature of 220 is deleted locally. This saves local storage space.
  • the computing node 210 is further configured to upload the node identity information corresponding to the local face feature and the face feature to the server 220 according to the preset upload frequency, and generate an upload log.
  • the node identity information corresponding to the face feature and the face feature saved locally and not uploaded to the server 220 is uploaded to the server 220 according to the upload log. In this way, the problem of repeated uploading can be avoided, the transmission resources can be saved, and the processing efficiency on the server side can be improved.
  • the computing node 210 may set the node identity information, the appearance time of the node identity information, and the face center coordinate corresponding to the node identity information. Saved in the local node log table.
  • Table 2 illustrates the node log table of the computing node A.
  • the computing nodes B, C, and D also have node logs similar to those shown in Table 2. table. It should be noted that the examples herein are merely illustrative and are not intended to limit the embodiments of the present invention.
  • the local node log table is saved for later use. For example, it can solve the problem that people have different node identity information due to the angle of the face and the light within the specified range of the same location point. For example, the valid identity information in the node identity information corresponding to the single node can be determined, so as to be used to count the user's stay in the scenario when applied to the actual scenario, and the like. The details will be described separately below.
  • the computing node 210 is further configured to:
  • the preset number here can be set according to actual needs, such as 10; the preset number of frames can also be set according to actual needs, such as 4 and so on. It should be noted that the examples herein are merely illustrative and are not intended to limit the embodiments of the present invention.
  • Table 3 illustrates the node identity information combining table of the computing node A.
  • the node Aa, the node Ac, and the node Ae are substantially the same nodes.
  • Identity information; the node Ab and the node Ad are substantially the same node identity information.
  • compute nodes B, C, and D also have a node identity information merge table similar to that shown in Table 3. It should be noted that the examples herein are merely illustrative and are not intended to limit the embodiments of the present invention.
  • the computing node 210 may upload its own node log table to the server 220, and the server 220 may determine the same according to the node log table uploaded by the computing node 210. If the number of the node identity information is greater than the preset number threshold, if the number of the same node identity information determined by the computing node 210 within the preset duration T is greater than the preset number threshold M, the same node identity information is calibrated as valid identity information. For example, if the number of the same node identity information determined within 10 seconds of the preset duration is greater than the preset number threshold 7, the same node identity information is calibrated as valid identity information. It should be noted that T and M may be set according to actual needs, and the examples herein are merely illustrative and are not intended to limit the embodiments of the present invention.
  • the server 220 can further count the number of people according to the valid identity information.
  • the image collection device of the computing node 210 is an image of the collected mall, and the server 220 can count the traffic of the mall according to the valid identity information.
  • the image collection device of the computing node 210 is an image of the store in the collected mall, and the server 220 can count the traffic of the store in the mall according to the valid identity information.
  • the shopping malls listed herein are only schematic, and the user can deploy the identity calibration system provided by the embodiments of the present invention in corresponding scenarios, such as popular scenic spots, public places, etc., by using an identity calibration system. The calibration of the identity of the person is realized, and then the statistics of the traffic of the scene and the detection of the hotspot are implemented according to the identity of the calibration.
  • the identity calibration system provided by the embodiment of the present invention can be implemented, specifically:
  • the image acquisition device of the computing node 210 is used to collect the image of the exhibition or the product display, perform face recognition on the collected image, identify the facial features in the image, and determine the facial features corresponding to the recognized facial features.
  • Node identity information after which the node identity information, the appearance time of the node identity information, and the face center coordinates corresponding to the node identity information are saved in the local node log table;
  • the computing node 210 uploads its own node log table to the server 220;
  • the server 220 may determine, according to the node log table uploaded by the computing node 210, whether the number of identity information of the same node is greater than a preset threshold. If the number of identity information of the same node determined by the computing node 210 within the preset duration T is greater than the preset number. Threshold M, the same node identity information is calibrated as valid identity information. For example, if the number of the same node identity information determined within 10 seconds of the preset duration is greater than the preset number threshold 7, the same node identity information is calibrated as valid identity information. Thus, the server 220 can count the visitors who are staying based on the valid identity information.
  • the server 220 may perform matching on the node identity information between the nodes by calculating the node identity information corresponding to the face feature and the face feature uploaded by the computing node. .
  • the corresponding node identity information of the face feature a on the computing node A is the node Aa
  • the node identity information corresponding to the face feature a on the computing node B is the node Ba
  • the server 220 can be based on the face feature.
  • a calibration node identity information node Aa and node Ba are the same person, and assign unified calibration identity information to gid-a.
  • the server 220 can be specifically implemented by the following steps:
  • the node identity information of the same face feature is assigned the same calibration identity information according to the face feature uploaded by the plurality of computing nodes 210 and the node identity information corresponding to the face feature, and the calibration identity information, the face feature, and the node identity information are established. Mapping relationship;
  • the node identity information of the node identity information merge table uploaded by the computing node is substantially the same as the identity information of each node, and the substantially identical node identity information is added to the mapping relationship.
  • the following table 4 illustrates the initial mapping relationship established on the server 220. It is to be noted that the examples are merely illustrative and are not intended to limit the embodiments of the present invention.
  • the server 220 may query, according to the node identity information in the mapping relationship shown in Table 4, the node identity information in the node identity information combination table of the computing node A shown in Table 3, which is substantially the same as the identity information of each node.
  • the substantially identical node identity information is added to the mapping relationship, and the mapping relationship shown in Table 5 is obtained.
  • the server 220 may further query, according to the node identity information in the mapping relationship shown in Table 4, the node identity information in the node identity information combination table of the computing nodes B, C, and D, which is substantially the same as the identity information of each node.
  • the substantially identical node identity information is added to the mapping relationship.
  • the server 220 may determine whether the mapping relationship exists or not. If the node identity information of the uploaded node identity information is the same, the calibration identity information corresponding to the same node identity information is used as the calibration identity of the uploaded node identity information according to the mapping relationship; if not, the assignment and mapping relationship are included. The new calibration identity information of the calibration identity information is used as the calibration identity of the uploaded node identity information, and the new calibration identity information, the uploaded node identity information and the corresponding facial features are recorded into the mapping relationship.
  • the plurality of computing nodes 210 upload the respective node log tables to the server 220, and the server 220 receives the facial features and faces uploaded by any of the plurality of computing nodes 210 again.
  • the node identity information corresponding to the feature may be obtained according to the node log table of the arbitrary computing node, and if the number of the same node identity information determined by the arbitrary computing node within the preset duration is greater than a preset number threshold, the same node identity information is used.
  • the data is calibrated as valid identity information, and the node identity information calibrated as valid identity information is further determined whether there is node identity information in the mapping relationship that is the same as the valid identity information.
  • the same identity identity information corresponding to the node identity information is mapped according to the mapping relationship.
  • the information is used as the calibration identity of the valid identity information; if not, the new calibration identity information different from the calibration identity information included in the mapping relationship is assigned, the new calibration identity information is used as the calibration identity of the valid identity information, and the new identity information is newly calibrated.
  • valid identity information and corresponding facial features Record mapping relationship.
  • the server 220 may further determine the validity of the node identity information in the initial mapping relationship established according to the node log table of each of the plurality of computing nodes 210, if the node identity information is valid. Sex, then reserved; if the node identity information is not valid, delete it.
  • the server 220 may further query, according to the node identity information corresponding to the newly calibrated identity information in the mapping relationship, the node identity information of the node identity information merge table corresponding to the newly calibrated identity information. Information, the substantially identical node identity information is added to the mapping relationship.
  • the server 220 can further count the number of people according to the calibration identity information.
  • the image collection device of the plurality of computing nodes 210 is an image of the collected shopping mall, and the server 220 can count the traffic of the shopping mall according to the calibration identity information.
  • the image collection devices of the plurality of computing nodes 210 are images of the shops in the collected shopping malls, and the server 220 may count the traffic of the shops in the shopping malls according to the calibration identity information.
  • the shopping malls listed herein are only schematic, and the user can deploy the identity calibration system provided by the embodiments of the present invention in corresponding scenarios, such as popular scenic spots, public places, etc., by using an identity calibration system. The calibration of the identity of the person is realized, and then the statistics of the traffic of the scene and the detection of the hotspot are implemented according to the identity of the calibration.
  • a network-based distributed management system may be deployed in the computing node 210 and the server 220, respectively, and the computing node 210 may upload a face recognition and a captured image through the distributed management system.
  • the analysis results of the analysis such as the face feature, the node identity information corresponding to the face feature, the coordinate position of the face in the image, the acquisition time of the image, and the like, are not limited thereto.
  • FIG. 3 is a schematic flowchart of an identity calibration method applied to a computing node according to an embodiment of the present invention.
  • the identity calibration method applied to a computing node including an image collection device may include the following steps S302 to S306. .
  • Step S302 Acquire an image according to a preset image acquisition frequency by using an image acquisition device.
  • Step S304 performing face recognition on the collected image, identifying a facial feature in the image, and determining node identity information corresponding to the facial feature according to the recognized facial feature.
  • the captured image is subjected to face recognition. If the face is not recognized, the image acquired by the image acquisition device is continuously acquired, and the image is subjected to face recognition.
  • Step S306 Upload the node identity information corresponding to the face feature and the face feature to the server, so that the server calibrates the identity of the face feature based on the face feature uploaded by the computing node and the node identity information corresponding to the face feature.
  • the preset image acquisition frequency mentioned in the above step S302 can be set or adjusted according to actual needs.
  • the preset image acquisition frequency may be 3 frames of images for 1 second, or 4 frames of images for 1 second, etc., in the embodiment of the present invention. There is no limit to this.
  • the captured image is subjected to face recognition in step S304, and the facial features in the image are recognized.
  • the face recognition algorithm may be used to perform face recognition on the collected image, and the detected image is detected. After the face is positioned and the key feature points of the face are located, the face region is extracted for preprocessing, and facial features such as face center coordinates, facial features, face shape, angle, and the like are extracted.
  • the node identity information corresponding to the face feature is determined according to the recognized face feature in step S304, and specifically, the correspondence between the face feature and the node identity information is recognized.
  • the face features are matched with the face features in the corresponding relationship to obtain a matching result; if the matching result is that there is a matching face feature in the corresponding relationship, the node identity information corresponding to the matching face features in the corresponding relationship is determined.
  • the node identity information corresponding to the recognized face feature if the matching result is that there is no matching face feature in the correspondence relationship, assigning a new node identity information different from the node identity information included in the corresponding relationship, the new node is The identity information is determined as the node identity information corresponding to the recognized face feature, and the new node identity information and the corresponding face feature are recorded into the corresponding relationship.
  • the node identity information herein includes an identifier of the computing node, which is identity information indicating that the face feature is on the computing node.
  • the computing node in the identity calibration system includes computing nodes A, B, C, and D, assuming that the computing node A and The face feature a is recognized in the image acquired by B, and the node identity information corresponding to the face feature a on the computing node A is the node Aa, and the node identity information corresponding to the face feature a on the computing node B is the node Ba.
  • the examples herein are merely illustrative and are not intended to limit the embodiments of the present invention.
  • the preset threshold may be set according to actual requirements. For example, the matching degree is measured by a value of 0 to 100, and the preset threshold may be 75 or 80, etc., it should be noted that the example here is only illustrative, and The embodiments of the present invention are not limited.
  • step S304 when performing face recognition on the collected image in step S304, if it is detected that the image includes multiple faces, the same number of working threads as the plurality of faces are activated; a worker thread identifying a face feature of each face in the image, and determining node identity information corresponding to the face feature according to the recognized face feature, wherein a worker thread identifies a face feature of a face, and The node identity information corresponding to the face feature is determined according to the recognized face feature.
  • the multi-threaded parallel operation can improve the processing efficiency of the computing node, thereby improving the processing efficiency of the entire identity calibration system.
  • the node identity information corresponding to the face feature and the face feature is uploaded to the server in step S306, and may be uploaded in real time. Specifically, the node identity information corresponding to the face feature may be determined each time. After that, the node identity information corresponding to the face feature and the face feature is uploaded to the server in real time.
  • the node identity information corresponding to the face feature and the face feature is uploaded to the server in step S306, and may also be uploaded according to a preset upload frequency, and specifically may be a node corresponding to the determined face feature.
  • the node identity information corresponding to the face feature and the face feature is saved locally; according to the preset upload frequency, the node identity information corresponding to the local face feature and the face feature is uploaded to the server.
  • the preset upload frequency can be set according to actual needs.
  • the preset upload frequency can be uploaded every 1 minute, or every 2 minutes. It should be noted that the example here is only illustrative, and The embodiments of the present invention are not limited.
  • the preset upload frequency may be saved locally.
  • the node identity information corresponding to the face feature and the face feature is uploaded to the server, and the node identity information corresponding to the face feature and the face feature uploaded to the server is deleted locally. This saves local storage space.
  • the preset upload frequency may be saved locally.
  • the node identity information corresponding to the face feature and the face feature is uploaded to the server, and an upload log is generated, so as to save the node identity corresponding to the face feature and the face feature that is not locally uploaded to the server according to the upload log.
  • the information is uploaded to the server. In this way, the problem of repeated uploading can be avoided, the transmission resources can be saved, and the processing efficiency on the server side can be improved.
  • the node identity information, the appearance time of the node identity information, and the face center corresponding to the node identity information may be performed.
  • the coordinates are saved in the local node log table for later use. For example, it can solve the problem that people have different node identity information due to the angle of the face and the light within the specified range of the same location point.
  • the valid identity information in the node identity information corresponding to the single node can be determined, so as to be used to count the user's stay in the scenario when applied to the actual scenario, and the like. The details will be described separately below.
  • the embodiment of the present invention may also Includes the following steps:
  • the present embodiment can solve the problem that a person can determine different node identity information due to the angle of the face and the light in the specified range of the same location point, that is, the above step is called a T-type algorithm, and the T-type algorithm Three conditions of composition:
  • the node identity information in the node log table is different, and is recorded as the first node identity information and the second node identity information;
  • the offset of the face center coordinate corresponding to the second node identity information with respect to the face center coordinate offset corresponding to the first node identity information is smaller than the preset number of pixel points
  • condition 3 When the conditions 1) and 2) are satisfied, if the change rule of the node identity information is as shown in condition 3), and the number of frames of the second node identity information between the same first node identity information continues to be smaller than the preset frame
  • the first node identity information and the second node identity information are determined as the same node identity information, which is used to increase the accuracy of the system.
  • the preset number here can be set according to actual needs, such as 10; the preset number of frames can also be set according to actual needs, such as 4 and so on. It should be noted that the examples herein are merely illustrative and are not intended to limit the embodiments of the present invention.
  • FIG. 5 is a schematic flowchart of an identity calibration method applied to a server according to an embodiment of the present invention.
  • the identity calibration method applied to the server may include the following steps S502 to S504.
  • Step S502 receiving the face feature information corresponding to the face feature and the face feature uploaded by the computing node of the image collecting device, wherein the computing node uses the image capturing device thereon to collect the image according to the preset image capturing frequency;
  • the image is used for face recognition, the face features in the image are recognized, and the node identity information corresponding to the face feature is determined according to the recognized face features; after that, the node identity information corresponding to the face feature and the face feature is uploaded.
  • the server receives face feature information corresponding to the face feature and the face feature uploaded.
  • Step S504 the identity of the facial feature is calibrated based on the facial feature uploaded by the computing node and the node identity information corresponding to the facial feature.
  • the identity of the facial features is calibrated based on the facial features uploaded by the computing node and the node identity information corresponding to the facial features in step S504.
  • the number of the identity information of the same node is greater than the preset number threshold according to the node log table uploaded by the computing node. If the number of identity information of the same node determined by the computing node in the preset time length T is greater than the preset number threshold M, Then the same node identity information is calibrated as valid identity information. For example, if the number of the same node identity information determined within 10 seconds of the preset duration is greater than the preset number threshold 7, the same node identity information is calibrated as valid identity information.
  • T and M may be set according to actual needs, and the examples herein are merely illustrative and are not intended to limit the embodiments of the present invention.
  • the server can further count the number of people according to the valid identity information.
  • the image collection device of the computing node is an image of the collected mall, and the server can count the traffic of the mall according to the valid identity information.
  • the image collection device of the computing node is an image of the store in the collected mall, and the server may calculate the traffic of the store in the mall according to the valid identity information.
  • the shopping malls listed herein are only schematic, and the user can deploy the identity calibration system provided by the embodiments of the present invention in corresponding scenarios, such as popular scenic spots, public places, etc., by using an identity calibration system. The calibration of the identity of the person is realized, and then the statistics of the traffic of the scene and the detection of the hotspot are implemented according to the identity of the calibration.
  • the image acquisition device of the computing node is used to collect the image of the exhibition or the product display, perform face recognition on the collected image, identify the facial features in the image, and determine the node corresponding to the facial feature according to the recognized facial features.
  • Identity information after which the node identity information, the appearance time of the node identity information, and the face center coordinates corresponding to the node identity information are saved in the local node log table; the computing node uploads its own node log table to the server;
  • the server may determine, according to the node log table uploaded by the computing node, whether the number of identity information of the same node is greater than a preset number threshold. If the number of identity information of the same node determined by the computing node in the preset duration T is greater than a preset threshold value M, Then the same node identity information is calibrated as valid identity information. For example, if the number of the same node identity information determined within 10 seconds of the preset duration is greater than the preset number threshold 7, the same node identity information is calibrated as valid identity information. Thus, the server can count the visitors who are staying based on the valid identity information.
  • step S504 may specifically calculate the node identity information between the nodes by using the node identity information corresponding to the face feature and the face feature uploaded by the computing node. match.
  • the corresponding node identity information of the face feature a on the computing node A is the node Aa
  • the corresponding node identity information of the face feature a on the computing node B is the node Ba
  • the server can according to the face feature a
  • the calibration node identity information node Aa and node Ba are the same person, and the unified calibration identity information is assigned to gid-a.
  • Step S504 can be specifically implemented by the following steps:
  • the node identity information of the same face feature is assigned the same calibration identity information, and the calibration identity information, the face feature, and the node identity information are established.
  • the node identity information of the node identity information merge table uploaded by the computing node is substantially the same as the identity information of each node, and the substantially identical node identity information is added to the mapping relationship.
  • the server when the server receives the face feature uploaded by any computing node of the plurality of computing nodes and the node identity information corresponding to the face feature, the server may also determine whether the mapping relationship exists and is uploaded.
  • the node identity information with the same node identity information if present, uses the calibration identity information corresponding to the same node identity information as the calibration identity of the uploaded node identity information according to the mapping relationship; if not, the calibration included in the mapping relationship
  • the new identity information is different from the identity information, and the newly-identified identity information is used as the calibration identity of the uploaded node identity information, and the new calibration identity information, the uploaded node identity information, and the corresponding facial features are recorded into the mapping relationship.
  • the plurality of computing nodes upload the respective node log tables to the server, and the server again receives the face features and the faces corresponding to the face features uploaded by any of the plurality of computing nodes.
  • the identity information may be based on the node log table of the arbitrary computing node. If the number of the same node identity information determined by the arbitrary computing node within a preset duration is greater than a preset number threshold, the identity information of the same node is calibrated to a valid identity. The information further determines whether the node identity information with the same valid identity information exists in the mapping relationship, and if yes, the calibration identity information corresponding to the same node identity information is used as the valid identity according to the mapping relationship.
  • the calibration identity of the information if it does not exist, assign new calibration identity information different from the calibration identity information contained in the mapping relationship, and use the newly-calibrated identity information as the calibration identity of the valid identity information, and newly calibrate the identity information and the valid identity information. And the corresponding face feature record In the relationship.
  • the server may further determine the validity of the node identity information in the initial mapping relationship established according to the node log table of each of the plurality of computing nodes, and if the node identity information is valid, Then, if the node identity information is not valid, delete it.
  • the server may further query the node identity information that is substantially the same as the node identity information corresponding to the new calibration identity information in the node identity information merge table according to the node identity information corresponding to the newly calibrated identity information in the mapping relationship.
  • the substantially identical node identity information is added to the mapping relationship.
  • the server can further count the number of people according to the calibration identity information.
  • the image collection device of the plurality of computing nodes is an image of the collected shopping mall, and the server can calculate the traffic of the shopping mall according to the calibration identity information.
  • the image collection device of the plurality of computing nodes is an image of the store in the collected mall, and the server may calculate the traffic of the store in the mall according to the calibration identity information.
  • the shopping malls listed herein are only schematic, and the user can deploy the identity calibration system provided by the embodiments of the present invention in corresponding scenarios, such as popular scenic spots, public places, etc., by using an identity calibration system. The calibration of the identity of the person is realized, and then the statistics of the traffic of the scene and the detection of the hotspot are implemented according to the identity of the calibration.
  • FIG. 6 is a schematic flowchart diagram of an identity calibration method applied to a server according to another embodiment of the present invention.
  • the identity calibration method applied to the server may include the following steps S602 to S606.
  • Step S602 receiving the face feature information corresponding to the face feature and the face feature uploaded by the computing node of the image collecting device, wherein the computing node uses the image capturing device thereon to collect the image according to the preset image capturing frequency;
  • the image is used for face recognition, the face features in the image are recognized, and the node identity information corresponding to the face feature is determined according to the recognized face features; after that, the node identity information corresponding to the face feature and the face feature is uploaded.
  • the server receives face feature information corresponding to the face feature and the face feature uploaded.
  • Step S604 analyzing whether the node identity information is valid identity information in the node identity information corresponding to the face feature uploaded by the computing node, and if yes, proceeding to step S606; if not, returning to step S602.
  • the same node identity information is calibrated as valid identity information. For example, if the number of the same node identity information determined within 10 seconds of the preset duration is greater than the preset number threshold 7, the same node identity information is calibrated as valid identity information.
  • T and M may be set according to actual needs, and the examples herein are merely illustrative and are not intended to limit the embodiments of the present invention.
  • Step S606 matching the gid list, and then, according to the matching result, calibrating the calibration identity information of the valid identity information obtained by the analysis.
  • the gid list includes the mapping relationship between the calibration identity information, the face feature, and the node identity information, and can determine whether there is node identity information in the mapping relationship that is the same as the valid identity information, and if so, according to the mapping relationship
  • the calibration identity information corresponding to the same node identity information is used as the calibration identity of the valid identity information; if not, the new calibration identity information different from the calibration identity information included in the mapping relationship is assigned, and the new calibration identity information is used as the valid identity information.
  • the identity is calibrated, and the new calibration identity information, valid identity information, and corresponding facial features are recorded into the mapping relationship.
  • the node identity information corresponding to the newly identifiable identity information in the node identity information merge table may be queried according to the node identity information corresponding to the newly calibrated identity information in the mapping relationship, The substantially identical node identity information is added to the mapping relationship.
  • the embodiments of the present invention can achieve the following beneficial effects:
  • the identity calibration system includes a computing node including an image collection device, and a server that establishes a communication connection with the computing node. After the computing node collects an image and performs face recognition of the image, the facial feature and the analysis data are extracted. Uploading to the server; the server performs feature matching and association to realize identity calibration of the facial features; thus, the distributed computing of the computing node and the server improves the efficiency and security of the system.
  • the embodiment of the present invention further analyzes the node identity information determined by the recognized face feature on the basis of the image analysis algorithm of the face recognition, and solves the problem that the person is within the specified range of the same location point, because the face is The angle and the light will determine the problem of different node identity information, improve the accuracy of the calculation of the node, and thus improve the accuracy of the identity calibration of the entire system.
  • the server by determining the validity of the node identity information of the computing node, the solution is solved within a specified range of the same location point, and the node identity information is determined due to the angle of the face and the light.
  • the problem has improved the accuracy of the identity calibration of the entire system.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Computer Interaction (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Image Analysis (AREA)
  • Collating Specific Patterns (AREA)

Abstract

一种身份标定系统和方法，该系统包括：包含图像采集装置（211）的计算节点（210），与所述计算节点（210）建立通信连接的服务器（220）；所述计算节点（210），用于利用其上的图像采集装置（211），按照预设图像采集频率采集图像；对采集的图像进行人脸识别，识别出图像中的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息；之后，将人脸特征及人脸特征对应的节点身份信息上传至所述服务器（220）；所述服务器（220），用于基于所述计算节点（210）上传的人脸特征及人脸特征对应的节点身份信息，对人脸特征的身份进行标定。由此，通过计算节点（210）和服务器（220）的分布式计算提高了系统的效率以及安全性。

Description

一种身份标定系统和方法 技术领域
本发明涉及信息处理技术领域，特别是一种身份标定系统和方法。
背景技术
传统的ibeacon低耗能的蓝牙技术定位，只能提供简单的位置信息，无法提供更详细的分析数据。随着图像处理技术越来越成熟，此类商业的应用也越来越多。尤其是其中的人脸识别，应用于商场对客流量的统计和商场中商铺的热点检测等是非常有价值的。
目前的商场客流量的统计和客流的移动轨迹追踪的实现是以人脸识别技术为基础，由于人脸识别需要较高的运算能力，通过将摄像头拍摄的视频流发送至后台服务器进行人脸识别可以提高识别速度，后台服务器接收到视频流后，分析视频中的信息，可以识别并提取相关的特征值并用于存储比较，然后显示相关的分析信息。其具体架构图如图1所示。
但图1所示的系统的问题在于，视频流是占用带宽的应用，当节点较多时，网络负载较大，对系统的稳定性及准确性有较大的影响，且后台服务器分析多路视频流时也存在计算资源不够的问题。另外，当视频流上传至后台服务器时，还涉及用户隐私问题。因此，亟待解决这些问题。
发明内容
鉴于上述问题，提出了本发明以便提供一种克服上述问题或者至少部分地解决上述问题的身份标定系统和方法，该身份标定系统通过计算节点和服务器的分布式计算提高了系统的效率以及安全性。
本发明实施例提供了一种身份标定系统，包括：包含图像采集装置的计算节点，与所述计算节点建立通信连接的服务器；
所述计算节点，用于利用其上的图像采集装置，按照预设图像采集频率采集图像；对采集的图像进行人脸识别，识别出图像中的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息；之后，将人脸特征及人脸特征对应的节点身份信息上传至所述服务器；
所述服务器，用于基于所述计算节点上传的人脸特征及人脸特征对应的节点身份信息，对人脸特征的身份进行标定。
可选地，所述计算节点还用于：
在人脸特征与节点身份信息的对应关系中，将识别出的人脸特征与所述对应关系中的人脸特征进行匹配；
若所述对应关系中存在匹配的人脸特征，则将所述对应关系中匹配的人脸特征对应的节点身份信息，确定为识别出的人脸特征对应的节点身份信息；
若所述对应关系中不存在匹配的人脸特征，则分配一与所述对应关系中包含的节点身份信息不同的新节点身份信息，将所述新节点身份信息确定为识别出的人脸特征对应的节点身份信息，并将所述新节点身份信息及对应的人脸特征记录到所述对应关系中。
可选地，所述计算节点还用于：
在确定人脸特征对应的节点身份信息之后，将该节点身份信息、该节点身份信息的出现时间以及该节点身份信息对应的人脸中心坐标保存在本地的节点日志表中。
可选地，所述计算节点还用于：
若所述节点日志表中的节点身份信息存在不相同的情况，则记为第一节点身份信息和第二节点身份信息；
从所述节点日志表中查询所述第一节点身份信息和所述第二节点身份信息各自对应的人脸中心坐标以及出现时间；
判断所述第二节点身份信息对应的人脸中心坐标相对所述第一节点身份信息对应的人脸中心坐标偏移的偏移量是否小于预设数量的像素点；
若是，并且相同的所述第一节点身份信息之间的所述第二节点身份信息所持续的帧数小于预设帧数，则将所述第一节点身份信息和所述第二节点身份信息确定为实质相同的节点身份信息，并记录在节点身份信息合并表中。
可选地，所述服务器还用于：
根据多个所述计算节点上传的人脸特征及人脸特征对应的节点身份信息，将相同人脸特征的节点身份信息分配相同的标定身份信息，建立标定身份信息、人脸特征以及节点身份信息间的映射关系；
根据所述映射关系中的各节点身份信息，查询所述计算节点上传的节点身份信息合并表中与各节点身份信息实质相同的节点身份信息，将该实质相同的节点身份信息增加到所述映射关系中。
可选地，所述服务器还用于：
当再次接收到多个所述计算节点中任意计算节点上传的人脸特征及人脸特征对应的节点身份信息时，判断所述映射关系中是否存在与上传的节点身份信息相同的节点身份信息，若存在，则根据所述映射关系将相同的节点身份信息对应的标定身份信息作为上传的节点身份信息的标定身份；若不存在，则分配与所述映射关系中包含的标定身份信 息不同的新标定身份信息，将新标定身份信息作为上传的节点身份信息的标定身份，并将新标定身份信息、上传的节点身份信息以及对应的人脸特征记录到所述映射关系中。
可选地，所述计算节点还用于：将所述节点日志表上传至所述服务器；
所述服务器还用于：当再次接收到多个所述计算节点中任意计算节点上传的人脸特征及人脸特征对应的节点身份信息时，根据所述节点日志表，若该任意计算节点在预设时长内确定的同一节点身份信息的数量大于预设数量阈值，则将该同一节点身份信息标定为有效身份信息，对标定为有效身份信息的节点身份信息进一步判断所述映射关系中是否存在与有效身份信息相同的节点身份信息，若存在，则根据所述映射关系将相同的节点身份信息对应的标定身份信息作为有效身份信息的标定身份；若不存在，则分配与所述映射关系中包含的标定身份信息不同的新标定身份信息，将新标定身份信息作为有效身份信息的标定身份，并将新标定身份信息、有效身份信息以及对应的人脸特征记录到所述映射关系中。
可选地，所述服务器还用于：
根据所述映射关系中新标定身份信息对应的节点身份信息，查询节点身份信息合并表中与新标定身份信息对应的节点身份信息实质相同的节点身份信息，将该实质相同的节点身份信息增加到所述映射关系中。
可选地，所述计算节点还用于：
在对采集的图像进行人脸识别时，若检测到图像中包含多个人脸，则启动与多个人脸相同数量的多个工作线程；
利用所述多个工作线程，识别出图像中多个人脸各自的人脸特征，根据识别出的人脸特征，确定人脸特征对应的节点身份信息，其中，一个工作线程识别一个人脸的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息。
可选地，所述计算节点还用于：
根据预设上传频率，将保存在本地的人脸特征及人脸特征对应的节点身份信息上传至所述服务器，并生成上传日志，从而根据所述上传日志，将后续保存在本地的且未上传至所述服务器的人脸特征及人脸特征对应的节点身份信息上传至所述服务器。
本发明实施例还提供了一种身份标定方法，应用于包含图像采集装置的计算节点，所述方法包括：
利用图像采集装置，按照预设图像采集频率采集图像；
对采集的图像进行人脸识别，识别出图像中的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息；
将人脸特征及人脸特征对应的节点身份信息上传至服务器，从而所述服务器基于所 述计算节点上传的人脸特征及人脸特征对应的节点身份信息，对人脸特征的身份进行标定。
可选地，所述根据识别出的人脸特征，确定人脸特征对应的节点身份信息，包括：
在人脸特征与节点身份信息的对应关系中，将识别出的人脸特征与所述对应关系中的人脸特征进行匹配；
若所述对应关系中存在匹配的人脸特征，则将所述对应关系中匹配的人脸特征对应的节点身份信息，确定为识别出的人脸特征对应的节点身份信息；
若所述对应关系中不存在匹配的人脸特征，则分配一与所述对应关系中包含的节点身份信息不同的新节点身份信息，将所述新节点身份信息确定为识别出的人脸特征对应的节点身份信息，并将所述新节点身份信息及对应的人脸特征记录到所述对应关系中。
可选地，在确定人脸特征对应的节点身份信息之后，所述方法还包括：
将该节点身份信息、该节点身份信息的出现时间以及该节点身份信息对应的人脸中心坐标保存在本地的节点日志表中。
可选地，所述方法还包括：
若所述节点日志表中的节点身份信息存在不相同的情况，则记为第一节点身份信息和第二节点身份信息；
从所述节点日志表中查询所述第一节点身份信息和所述第二节点身份信息各自对应的人脸中心坐标以及出现时间；
判断所述第二节点身份信息对应的人脸中心坐标相对所述第一节点身份信息对应的人脸中心坐标偏移的偏移量是否小于预设数量的像素点；
若是，并且相同的所述第一节点身份信息之间的所述第二节点身份信息所持续的帧数小于预设帧数，则将所述第一节点身份信息和所述第二节点身份信息确定为实质相同的节点身份信息，并记录在节点身份信息合并表中。
可选地，所述方法还包括：
将所述节点日志表以及所述节点身份信息合并表上传至所述服务器。
可选地，所述对采集的图像进行人脸识别，识别出图像中的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息，包括：
在对采集的图像进行人脸识别时，若检测到图像中包含多个人脸，则启动与多个人脸相同数量的多个工作线程；
利用所述多个工作线程，识别出图像中多个人脸各自的人脸特征，根据识别出的人脸特征，确定人脸特征对应的节点身份信息，其中，一个工作线程识别一个人脸的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息。
可选地，所述将人脸特征及人脸特征对应的节点身份信息上传至服务器，包括：
根据预设上传频率，将保存在本地的人脸特征及人脸特征对应的节点身份信息上传至所述服务器，并生成上传日志，从而根据所述上传日志，将后续保存在本地的且未上传至所述服务器的人脸特征及人脸特征对应的节点身份信息上传至所述服务器。
本发明实施例还提供了一种身份标定方法，应用于服务器，包括：
接收包含图像采集装置的计算节点上传的人脸特征及人脸特征对应的节点身份信息；其中，所述计算节点利用其上的图像采集装置，按照预设图像采集频率采集图像；对采集的图像进行人脸识别，识别出图像中的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息；之后，将人脸特征及人脸特征对应的节点身份信息上传至所述服务器；
基于所述计算节点上传的人脸特征及人脸特征对应的节点身份信息，对人脸特征的身份进行标定。
可选地，基于所述计算节点上传的人脸特征及人脸特征对应的节点身份信息，对人脸特征的身份进行标定，包括：
根据多个所述计算节点上传的人脸特征及人脸特征对应的节点身份信息，将相同人脸特征的节点身份信息分配相同的标定身份信息，建立标定身份信息、人脸特征以及节点身份信息间的映射关系；
根据所述映射关系中的各节点身份信息，查询所述计算节点上传的节点身份信息合并表中与各节点身份信息实质相同的节点身份信息，将该实质相同的节点身份信息增加到所述映射关系中。
可选地，基于所述计算节点上传的人脸特征及人脸特征对应的节点身份信息，对人脸特征的身份进行标定，包括：
当再次接收到多个所述计算节点中任意计算节点上传的人脸特征及人脸特征对应的节点身份信息时，判断所述映射关系中是否存在与上传的节点身份信息相同的节点身份信息，若存在，则根据所述映射关系将相同的节点身份信息对应的标定身份信息作为上传的节点身份信息的标定身份；若不存在，则分配与所述映射关系中包含的标定身份信息不同的新标定身份信息，将新标定身份信息作为上传的节点身份信息的标定身份，并将新标定身份信息、上传的节点身份信息以及对应的人脸特征记录到所述映射关系中。
可选地，基于所述计算节点上传的人脸特征及人脸特征对应的节点身份信息，对人脸特征的身份进行标定，包括：
当再次接收到多个所述计算节点中任意计算节点上传的人脸特征及人脸特征对应的节点身份信息时，根据所述计算节点上传的节点日志表，若该任意计算节点在预设时长 内确定的同一节点身份信息的数量大于预设数量阈值，则将该同一节点身份信息标定为有效身份信息，对标定为有效身份信息的节点身份信息进一步判断所述映射关系中是否存在与有效身份信息相同的节点身份信息，若存在，则根据所述映射关系将相同的节点身份信息对应的标定身份信息作为有效身份信息的标定身份；若不存在，则分配与所述映射关系中包含的标定身份信息不同的新标定身份信息，将新标定身份信息作为有效身份信息的标定身份，并将新标定身份信息、有效身份信息以及对应的人脸特征记录到所述映射关系中。
可选地，所述方法还包括：
根据所述映射关系中新标定身份信息对应的节点身份信息，查询节点身份信息合并表中与新标定身份信息对应的节点身份信息实质相同的节点身份信息，将该实质相同的节点身份信息增加到所述映射关系中。
本发明实施例提供的身份标定系统包括了包含图像采集装置的计算节点，以及与计算节点建立通信连接的服务器，在计算节点采集图像并进行图像的人脸识别后，提取人脸特征及分析数据上传至服务器；服务器进行特征匹配及关联，实现人脸特征的身份标定；由此，通过计算节点和服务器的分布式计算提高了系统的效率以及安全性。
并且，本发明实施例在计算节点处，还在人脸识别的图像分析算法的基础上对识别人脸特征确定的节点身份信息进一步分析，解决人在同一位置点的指定范围内，由于人脸的角度以及光线，会确定出不同的节点身份信息的问题，提高计算节点识别的准确性，进而提高了整个系统的身份标定的准确性。
进一步，本发明实施例在服务器处，通过对计算节点的节点身份信息的有效性确定，解决人在同一位置点的指定范围内，由于人脸的角度以及光线，会确定出不同的节点身份信息的问题，提高了整个系统的身份标定的准确性。
上述说明仅是本发明技术方案的概述，为了能够更清楚了解本发明的技术手段，而可依照说明书的内容予以实施，并且为了让本发明的上述和其它目的、特征和优点能够更明显易懂，以下特举本发明的具体实施方式。
根据下文结合附图对本发明具体实施例的详细描述，本领域技术人员将会更加明了本发明的上述以及其他目的、优点和特征。
附图说明
通过阅读下文优选实施方式的详细描述，各种其他的优点和益处对于本领域普通技术人员将变得清楚明了。附图仅用于示出优选实施方式的目的，而并不认为是对本发明的限制。而且在整个附图中，用相同的参考符号表示相同的部件。在附图中：
图1所示为现有技术中热点分析与移动轨迹追踪的系统架构示意图；
图2所示为根据本发明一实施例的身份标定系统的结构示意图；
图3所示为根据本发明一实施例的应用于计算节点的身份标定方法的流程示意图；
图4所示为根据本发明一实施例的节点身份信息的T型规律示意图；
图5所示为根据本发明一实施例的应用于服务器的身份标定方法的流程示意图；以及
图6所示为根据本发明另一实施例的应用于服务器的身份标定方法的流程示意图。
具体实施方式
下面将参照附图更详细地描述本公开的示例性实施例。虽然附图中显示了本公开的示例性实施例，然而应当理解，可以以各种形式实现本公开而不应被这里阐述的实施例所限制。相反，提供这些实施例是为了能够更透彻地理解本公开，并且能够将本公开的范围完整的传达给本领域的技术人员。
为了解决上述技术问题，本发明实施例提供了一种身份标定系统，该身份标定系统通过计算节点和服务器的分布式计算提高了系统的效率以及安全性。如图2所示，本发明实施例提供的身份标定系统可以包括计算节点210，与计算节点210建立通信连接的服务器220，具体地：
计算节点210包含图像采集装置211，用于利用其上的图像采集装置211，按照预设图像采集频率采集图像；对采集的图像进行人脸识别，识别出图像中的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息；之后，将人脸特征及人脸特征对应的节点身份信息上传至服务器220；
服务器220，用于基于计算节点210上传的人脸特征及人脸特征对应的节点身份信息，对人脸特征的身份进行标定。
需要说明的是，本发明实施例提供的身份标定系统中的计算节点210的数量可以是一个或多个，图2中的四个计算节点210仅是示意性的，并不对本发明实施例进行限制。
在本发明的可选实施例中，预设图像采集频率可以根据实际需求来设置或调整，如预设图像采集频率可以是1秒采集3帧图像，或者1秒采集4帧图像等，本发明实施例对此不作限制。
在本发明的可选实施例中，计算节点210对采集的图像进行人脸识别，识别出图像中的人脸特征，具体可以利用人脸识别算法对采集的图像进行人脸识别，在检测到人脸并定位面部关键特征点后，提取人脸区域进行预处理，提取人脸特征，如人脸中心坐标、脸部的五官位置、脸型、角度等。
在本发明的可选实施例中，计算节点210在根据识别出的人脸特征，确定人脸特征对应的节点身份信息时，具体可以是：
在人脸特征与节点身份信息的对应关系中，将识别出的人脸特征与对应关系中的人脸特征进行匹配，得到匹配结果；
若匹配结果为对应关系中存在匹配的人脸特征，则将对应关系中匹配的人脸特征对应的节点身份信息，确定为识别出的人脸特征对应的节点身份信息；
若匹配结果为对应关系中不存在匹配的人脸特征，则分配一与对应关系中包含的节点身份信息不同的新节点身份信息，将新节点身份信息确定为识别出的人脸特征对应的节点身份信息，并将新节点身份信息及对应的人脸特征记录到对应关系中。
这里的节点身份信息中包含计算节点的标识，是表示人脸特征在计算节点上的身份信息，即各节点上的对应关系是分别独立生成并存储的，例如身份标定系统中的计算节点210包括计算节点A、B、C和D，假设从计算节点A和B采集的图像中均识别出人脸特征a，则人脸特征a在计算节点A上对应的节点身份信息是节点A-a，人脸特征a在计算节点B上对应的节点身份信息是节点B-a。下面的表1中示意了计算节点A的人脸特征与节点身份信息的对应关系，同样地，计算节点B、C和D也存在类似表1所示的对应关系表。需要说明的是，此处举例仅是示意性的，并不对本发明实施例进行限制。
表1
Figure PCTCN2018122883-appb-000001
在本发明的可选实施例中，若对应关系中存在与识别出的人脸特征的匹配度大于预设阈值的人脸特征，则确定对应关系中存在匹配的人脸特征，否则确定对应关系中不存在匹配的人脸特征。这里的预设阈值可以根据实际需求来设置，例如匹配度以0至100的数值来衡量，则预设阈值可以是75或80等，需要说明的是，此处举例仅是示意性的，并不对本发明实施例进行限制。
在本发明的可选实施例中，计算节点210在对采集的图像进行人脸识别时，若检测到图像中包含多个人脸，则启动与多个人脸相同数量的多个工作线程；利用多个工作线程，识别出图像中多个人脸各自的人脸特征，根据识别出的人脸特征，确定人脸特征对 应的节点身份信息，其中，一个工作线程识别一个人脸的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息。这样，通过多线程并行工作的方式能够提高计算节点的处理效率，从而提高整个身份标定系统的处理效率。
在本发明的可选实施例中，计算节点210在将人脸特征及人脸特征对应的节点身份信息上传至服务器220时，可以实时上传，具体地：
计算节点210每当确定人脸特征对应的节点身份信息后，实时将人脸特征及人脸特征对应的节点身份信息上传至服务器220。
在本发明的可选实施例中，计算节点210在将人脸特征及人脸特征对应的节点身份信息上传至服务器220时，也可以按照预设上传频率上传，具体地：
计算节点210在确定人脸特征对应的节点身份信息之后，将人脸特征及人脸特征对应的节点身份信息保存在本地；
根据预设上传频率，将保存在本地的人脸特征及人脸特征对应的节点身份信息上传至服务器220。
这里的预设上传频率可以按照实际需求进行设置，如预设上传频率可以是每隔1分钟上传一次，或者每隔2分钟上传一次，需要说明的是，此处举例仅是示意性的，并不对本发明实施例进行限制。
在本发明的可选实施例中，计算节点210还用于根据预设上传频率，将保存在本地的人脸特征及人脸特征对应的节点身份信息上传至服务器220，并将已上传至服务器220的人脸特征及人脸特征对应的节点身份信息从本地删除。这样，能够节省本地的存储空间。
在本发明的可选实施例中，计算节点210还用于根据预设上传频率，将保存在本地的人脸特征及人脸特征对应的节点身份信息上传至服务器220，并生成上传日志，从而根据上传日志，将后续保存在本地的且未上传至服务器220的人脸特征及人脸特征对应的节点身份信息上传至服务器220。这样，能够避免重复上传的问题，能够节省传输资源，并提高服务器侧的处理效率。
在本发明的可选实施例中，计算节点210在确定人脸特征对应的节点身份信息之后，可以将该节点身份信息、该节点身份信息的出现时间以及该节点身份信息对应的人脸中心坐标保存在本地的节点日志表中。仍以上述的计算节点A、B、C和D为例，下面的表2示意了计算节点A的节点日志表，同样地，计算节点B、C和D也存在类似表2所示的节点日志表。需要说明的是，此处举例仅是示意性的，并不对本发明实施例进行限制。
表2
Figure PCTCN2018122883-appb-000002
进一步地，保存在本地的节点日志表，可以供后续使用。例如，能够解决人在同一位置点的指定范围内，由于人脸的角度以及光线，会确定出不同的节点身份信息的问题。又例如，能够确定单节点对应的节点身份信息中的有效身份信息，从而在应用到实际场景时，用来统计用户在该场景下的停留情况，等等。下面将分别进行详细介绍。
在本发明的可选实施例中，在解决人在同一位置点的指定范围内，由于人脸的角度以及光线，会确定出不同的节点身份信息的问题时，具体地：
若节点日志表中的节点身份信息存在不相同的情况，记为第一节点身份信息和第二节点身份信息，则计算节点210还用于：
从节点日志表中查询第一节点身份信息和第二节点身份信息各自对应的人脸中心坐标以及出现时间；
判断第二节点身份信息对应的人脸中心坐标相对第一节点身份信息对应的人脸中心坐标偏移的偏移量是否小于预设数量的像素点；
若是，并且相同的第一节点身份信息之间的第二节点身份信息所持续的帧数小于预设帧数，则将第一节点身份信息和第二节点身份信息确定为实质相同的节点身份信息，并记录在节点身份信息合并表中。
这里的预设数量可以根据实际需求来设置，如10等；预设帧数也可以根据实际需求来设置，如4等。需要说明的是，此处举例仅是示意性的，并不对本发明实施例进行限制。
仍以上述的计算节点A、B、C和D为例，下面的表3示意了计算节点A的节点身份信息合并表，在表3中，节点A-a、节点A-c以及节点A-e为实质相同的节点身份信息；节点A-b和节点A-d为实质相同的节点身份信息。同样地，计算节点B、C和D也存在类似表3所示的节点身份信息合并表。需要说明的是，此处举例仅是示意性的，并不对本发明实施例进行限制。
表3
Figure PCTCN2018122883-appb-000003
在本发明的可选实施例中，若计算节点210的数量为一个，计算节点210可以将自身的节点日志表上传至服务器220，服务器220可以根据计算节点210上传的节点日志表，判断某同一节点身份信息的数量是否大于预设数量阈值，若计算节点210在预设时长T内确定的同一节点身份信息的数量大于预设数量阈值M，则将该同一节点身份信息标定为有效身份信息。例如，在预设时长10秒内确定的同一节点身份信息的数量大于预设数量阈值7，则将该同一节点身份信息标定为有效身份信息。需要说明的是，T和M可以根据实际需求来设置，此处的举例仅是示意性的，并不对本发明实施例进行限制。
这样，服务器220进一步可以根据有效身份信息来统计人数，例如，计算节点210的图像采集装置是采集的商场的图像，则服务器220可以根据有效身份信息来统计商场的客流量。又例如，计算节点210的图像采集装置是采集的商场中商铺的图像，则服务器220可以根据有效身份信息来统计商场中商铺的客流量。需要说明的是，此处列举的商场仅是示意性的，用户可以根据实际需求将本发明实施例提供的身份标定系统部署在相应的场景，如热门景点、公共场所等，利用身份标定系统来实现人物身份的标定，进而根据标定的身份来实现场景的客流量的统计和热点的检测。
在一个具体的应用场景中，对于展会或产品陈列时，通常需要了解参观人员对产品或展品的关注度，因此，需要统计有停留的参观人员，不统计那些没有停留的。采用本发明实施例提供的身份标定系统可以来实现，具体地：
采用计算节点210的图像采集装置来采集展会或产品陈列的图像，对采集的图像进行人脸识别，识别出图像中的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息，之后，将该节点身份信息、该节点身份信息的出现时间以及该节点身份信息对应的人脸中心坐标保存在本地的节点日志表中；
计算节点210将自身的节点日志表上传至服务器220；
服务器220可以根据计算节点210上传的节点日志表，判断某同一节点身份信息的数量是否大于预设数量阈值，若计算节点210在预设时长T内确定的同一节点身份信息的数量大于预设数量阈值M，则将该同一节点身份信息标定为有效身份信息。例如，在预设时长10秒内确定的同一节点身份信息的数量大于预设数量阈值7，则将该同一节点 身份信息标定为有效身份信息。从而，服务器220可以根据有效身份信息来统计有停留的参观人员。
在本发明的可选实施例中，若计算节点210的数量为多个，服务器220可以对计算节点上传的人脸特征和人脸特征对应的节点身份信息进行计算节点间的节点身份信息的匹配。仍然以上文列举为例，人脸特征a在计算节点A上对应的节点身份信息是节点A-a，人脸特征a在计算节点B上对应的节点身份信息是节点B-a，服务器220可以根据人脸特征a标定节点身份信息节点A-a和节点B-a是同一个人，并分配统一的标定身份信息为gid-a。服务器220具体可以通过以下步骤来实现：
根据多个计算节点210上传的人脸特征及人脸特征对应的节点身份信息，将相同人脸特征的节点身份信息分配相同的标定身份信息，建立标定身份信息、人脸特征以及节点身份信息间的映射关系；
根据映射关系中的各节点身份信息，查询计算节点上传的节点身份信息合并表中与各节点身份信息实质相同的节点身份信息，将该实质相同的节点身份信息增加到映射关系中。
下面的表4示意了服务器220上建立的初始的映射关系，需要说明的是，此处举例仅是示意性的，并不对本发明实施例进行限制。
表4
Figure PCTCN2018122883-appb-000004
进一步地，服务器220根据表4所示的映射关系中的各节点身份信息，可以查询表3所示的计算节点A的节点身份信息合并表中，与各节点身份信息实质相同的节点身份信息，将该实质相同的节点身份信息增加到映射关系中，得到表5所示的映射关系。
表5
Figure PCTCN2018122883-appb-000005
同样地，服务器220根据表4所示的映射关系中的各节点身份信息，还可以查询计算节点B、C和D的节点身份信息合并表中，与各节点身份信息实质相同的节点身份信息，将该实质相同的节点身份信息增加到映射关系中。
在本发明的可选实施例中，服务器220当再次接收到多个计算节点210中任意计算节点上传的人脸特征及人脸特征对应的节点身份信息时，还可以判断映射关系中是否存在与上传的节点身份信息相同的节点身份信息，若存在，则根据映射关系将相同的节点身份信息对应的标定身份信息作为上传的节点身份信息的标定身份；若不存在，则分配与映射关系中包含的标定身份信息不同的新标定身份信息，将新标定身份信息作为上传的节点身份信息的标定身份，并将新标定身份信息、上传的节点身份信息以及对应的人脸特征记录到映射关系中。
在本发明的可选实施例中，多个计算节点210将各自的节点日志表上传至服务器220，则服务器220当再次接收到多个计算节点210中任意计算节点上传的人脸特征及人脸特征对应的节点身份信息时，可以根据该任意计算节点的节点日志表，若该任意计算节点在预设时长内确定的同一节点身份信息的数量大于预设数量阈值，则将该同一节点身份信息标定为有效身份信息，对标定为有效身份信息的节点身份信息进一步判断映射关系中是否存在与有效身份信息相同的节点身份信息，若存在，则根据映射关系将相同的节点身份信息对应的标定身份信息作为有效身份信息的标定身份；若不存在，则分配与映射关系中包含的标定身份信息不同的新标定身份信息，将新标定身份信息作为有效身份信息的标定身份，并将新标定身份信息、有效身份信息以及对应的人脸特征记录到映射关系中。
在本发明的可选实施例中，服务器220还可以根据多个计算节点210各自的节点日志表，对建立的初始的映射关系中的节点身份信息的有效性进行判断，若节点身份信息具备有效性，则保留；若节点身份信息不具备有效性，则删除。
在本发明的可选实施例中，服务器220还可以根据映射关系中新标定身份信息对应的节点身份信息，查询节点身份信息合并表中与新标定身份信息对应的节点身份信息实质相同的节点身份信息，将该实质相同的节点身份信息增加到映射关系中。
这样，服务器220进一步可以根据标定身份信息来统计人数，例如，多个计算节点210的图像采集装置是采集的商场的图像，则服务器220可以根据标定身份信息来统计商场的客流量。又例如，多个计算节点210的图像采集装置是采集的商场中商铺的图像，则服务器220可以根据标定身份信息来统计商场中商铺的客流量。需要说明的是，此处列举的商场仅是示意性的，用户可以根据实际需求将本发明实施例提供的身份标定系统部署在相应的场景，如热门景点、公共场所等，利用身份标定系统来实现人物身份的标 定，进而根据标定的身份来实现场景的客流量的统计和热点的检测。
在本发明的可选实施例中，可以分别在计算节点210及服务器220中部署实现基于网络的分布式管理系统，计算节点210可以通过此分布式管理系统上传对采集的图像进行人脸识别和分析的分析结果，如人脸特征、人脸特征对应的节点身份信息、人脸在图像中的坐标位置、图像的采集时间等等，本发明实施例不限于此。
基于同一发明构思，本发明实施例还提供了一种身份标定方法。如图3所示为根据本发明一实施例的应用于计算节点的身份标定方法的流程示意图，参见图3，该应用于包含图像采集装置的计算节点的身份标定方法可以包括如下步骤S302至S306。
步骤S302，利用图像采集装置，按照预设图像采集频率采集图像。
步骤S304，对采集的图像进行人脸识别，识别出图像中的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息。
在该步骤中，对采集的图像进行人脸识别，若未识别出人脸，则继续获取图像采集装置采集的图像，并对图像进行人脸识别。
步骤S306，将人脸特征及人脸特征对应的节点身份信息上传至服务器，从而服务器基于计算节点上传的人脸特征及人脸特征对应的节点身份信息，对人脸特征的身份进行标定。
上文步骤S302中提及的预设图像采集频率可以根据实际需求来设置或调整，如预设图像采集频率可以是1秒采集3帧图像，或者1秒采集4帧图像等，本发明实施例对此不作限制。
在本发明的可选实施例中，步骤S304中对采集的图像进行人脸识别，识别出图像中的人脸特征，具体可以利用人脸识别算法对采集的图像进行人脸识别，在检测到人脸并定位面部关键特征点后，提取人脸区域进行预处理，提取人脸特征，如人脸中心坐标、脸部的五官位置、脸型、角度等。
在本发明的可选实施例中，步骤S304中根据识别出的人脸特征，确定人脸特征对应的节点身份信息，具体可以是在人脸特征与节点身份信息的对应关系中，将识别出的人脸特征与对应关系中的人脸特征进行匹配，得到匹配结果；若匹配结果为对应关系中存在匹配的人脸特征，则将对应关系中匹配的人脸特征对应的节点身份信息，确定为识别出的人脸特征对应的节点身份信息；若匹配结果为对应关系中不存在匹配的人脸特征，则分配一与对应关系中包含的节点身份信息不同的新节点身份信息，将新节点身份信息确定为识别出的人脸特征对应的节点身份信息，并将新节点身份信息及对应的人脸特征记录到对应关系中。
这里的节点身份信息中包含计算节点的标识，是表示人脸特征在计算节点上的身份 信息，例如身份标定系统中的计算节点包括计算节点A、B、C和D，假设从计算节点A和B采集的图像中均识别出人脸特征a，则人脸特征a在计算节点A上对应的节点身份信息是节点A-a，人脸特征a在计算节点B上对应的节点身份信息是节点B-a，需要说明的是，此处举例仅是示意性的，并不对本发明实施例进行限制。
在本发明的可选实施例中，若对应关系中存在与识别出的人脸特征的匹配度大于预设阈值的人脸特征，则确定对应关系中存在匹配的人脸特征，否则确定对应关系中不存在匹配的人脸特征。这里的预设阈值可以根据实际需求来设置，例如匹配度以0至100的数值来衡量，则预设阈值可以是75或80等，需要说明的是，此处举例仅是示意性的，并不对本发明实施例进行限制。
在本发明的可选实施例中，步骤S304中在对采集的图像进行人脸识别时，若检测到图像中包含多个人脸，则启动与多个人脸相同数量的多个工作线程；利用多个工作线程，识别出图像中多个人脸各自的人脸特征，根据识别出的人脸特征，确定人脸特征对应的节点身份信息，其中，一个工作线程识别一个人脸的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息。这样，通过多线程并行工作的方式能够提高计算节点的处理效率，从而提高整个身份标定系统的处理效率。
在本发明的可选实施例中，上文步骤S306中将人脸特征及人脸特征对应的节点身份信息上传至服务器，可以实时上传，具体可以是每当确定人脸特征对应的节点身份信息后，实时将人脸特征及人脸特征对应的节点身份信息上传至服务器。
在本发明的可选实施例中，步骤S306中将人脸特征及人脸特征对应的节点身份信息上传至服务器，也可以按照预设上传频率上传，具体可以是在确定人脸特征对应的节点身份信息之后，将人脸特征及人脸特征对应的节点身份信息保存在本地；根据预设上传频率，将保存在本地的人脸特征及人脸特征对应的节点身份信息上传至服务器。
这里的预设上传频率可以按照实际需求进行设置，如预设上传频率可以是每隔1分钟上传一次，或者每隔2分钟上传一次，需要说明的是，此处举例仅是示意性的，并不对本发明实施例进行限制。
在本发明的可选实施例中，在根据预设上传频率，将保存在本地的人脸特征及人脸特征对应的节点身份信息上传至服务器时，可以根据预设上传频率，将保存在本地的人脸特征及人脸特征对应的节点身份信息上传至服务器，并将已上传至服务器的人脸特征及人脸特征对应的节点身份信息从本地删除。这样，能够节省本地的存储空间。
在本发明的可选实施例中，在根据预设上传频率，将保存在本地的人脸特征及人脸特征对应的节点身份信息上传至服务器时，可以根据预设上传频率，将保存在本地的人脸特征及人脸特征对应的节点身份信息上传至服务器，并生成上传日志，从而根据上传 日志，将后续保存在本地的且未上传至服务器的人脸特征及人脸特征对应的节点身份信息上传至服务器。这样，能够避免重复上传的问题，能够节省传输资源，并提高服务器侧的处理效率。
在本发明的可选实施例中，上文步骤S304在确定人脸特征对应的节点身份信息之后，可以将该节点身份信息、该节点身份信息的出现时间以及该节点身份信息对应的人脸中心坐标保存在本地的节点日志表中，以供后续使用。例如，能够解决人在同一位置点的指定范围内，由于人脸的角度以及光线，会确定出不同的节点身份信息的问题。又例如，能够确定单节点对应的节点身份信息中的有效身份信息，从而在应用到实际场景时，用来统计用户在该场景下的停留情况，等等。下面将分别进行详细介绍。
在本发明的可选实施例中，若保存在本地的节点日志表中的节点身份信息存在不相同的情况，记为第一节点身份信息和第二节点身份信息，则本发明实施例还可以包括以下步骤：
从节点日志表中查询第一节点身份信息和第二节点身份信息各自对应的人脸中心坐标以及出现时间；
判断第二节点身份信息对应的人脸中心坐标相对第一节点身份信息对应的人脸中心坐标偏移的偏移量是否小于预设数量的像素点；
若是，并且相同的第一节点身份信息之间的第二节点身份信息所持续的帧数小于预设帧数，则将第一节点身份信息和第二节点身份信息确定为实质相同的节点身份信息，并记录在节点身份信息合并表中。
可以看到，本实施例能够解决人在同一位置点的指定范围内，由于人脸的角度以及光线，会确定出不同的节点身份信息的问题，即将上述步骤称为T型算法，T型算法构成的三个条件：
1)节点日志表中的节点身份信息存在不相同的情况，记为第一节点身份信息和第二节点身份信息；
2)第二节点身份信息对应的人脸中心坐标相对第一节点身份信息对应的人脸中心坐标偏移的偏移量小于预设数量的像素点；
3)节点身份信息跳动规律，符合梯形规律，如图4所示。
在满足条件1)和2)时，如果节点身份信息的变化规律如条件3)所示，并且，相同的第一节点身份信息之间的第二节点身份信息所持续的帧数小于预设帧数，则将第一节点身份信息和第二节点身份信息确定为相同的节点身份信息，用来增加系统的准确性。
这里的预设数量可以根据实际需求来设置，如10等；预设帧数也可以根据实际需求来设置，如4等。需要说明的是，此处举例仅是示意性的，并不对本发明实施例进行限 制。
基于同一发明构思，本发明实施例还提供了一种身份标定方法。如图5所示为根据本发明一实施例的应用于服务器的身份标定方法的流程示意图，参见图5，该应用于服务器的身份标定方法可以包括如下步骤S502至S504。
步骤S502，接收包含图像采集装置的计算节点上传的人脸特征及人脸特征对应的节点身份信息；其中，计算节点利用其上的图像采集装置，按照预设图像采集频率采集图像；对采集的图像进行人脸识别，识别出图像中的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息；之后，将人脸特征及人脸特征对应的节点身份信息上传至服务器。
步骤S504，基于计算节点上传的人脸特征及人脸特征对应的节点身份信息，对人脸特征的身份进行标定。
在本发明的可选实施例中，若计算节点的数量为一个，则步骤S504中基于计算节点上传的人脸特征及人脸特征对应的节点身份信息，对人脸特征的身份进行标定，具体可以是根据计算节点上传的节点日志表，判断某同一节点身份信息的数量是否大于预设数量阈值，若计算节点在预设时长T内确定的同一节点身份信息的数量大于预设数量阈值M，则将该同一节点身份信息标定为有效身份信息。例如，在预设时长10秒内确定的同一节点身份信息的数量大于预设数量阈值7，则将该同一节点身份信息标定为有效身份信息。需要说明的是，T和M可以根据实际需求来设置，此处的举例仅是示意性的，并不对本发明实施例进行限制。
这样，服务器进一步可以根据有效身份信息来统计人数，例如，计算节点的图像采集装置是采集的商场的图像，则服务器可以根据有效身份信息来统计商场的客流量。又例如，计算节点的图像采集装置是采集的商场中商铺的图像，则服务器可以根据有效身份信息来统计商场中商铺的客流量。需要说明的是，此处列举的商场仅是示意性的，用户可以根据实际需求将本发明实施例提供的身份标定系统部署在相应的场景，如热门景点、公共场所等，利用身份标定系统来实现人物身份的标定，进而根据标定的身份来实现场景的客流量的统计和热点的检测。
在一个具体的应用场景中，对于展会或产品陈列时，通常需要了解参观人员对产品或展品的关注度，因此，需要统计有停留的参观人员，不统计那些没有停留的。具体可以通过以下步骤来实现：
采用计算节点的图像采集装置来采集展会或产品陈列的图像，对采集的图像进行人脸识别，识别出图像中的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息，之后，将该节点身份信息、该节点身份信息的出现时间以及该节点身份信 息对应的人脸中心坐标保存在本地的节点日志表中；计算节点将自身的节点日志表上传至服务器；
服务器可以根据计算节点上传的节点日志表，判断某同一节点身份信息的数量是否大于预设数量阈值，若计算节点在预设时长T内确定的同一节点身份信息的数量大于预设数量阈值M，则将该同一节点身份信息标定为有效身份信息。例如，在预设时长10秒内确定的同一节点身份信息的数量大于预设数量阈值7，则将该同一节点身份信息标定为有效身份信息。从而，服务器可以根据有效身份信息来统计有停留的参观人员。
在本发明的可选实施例中，若计算节点的数量为多个，则步骤S504具体可以对计算节点上传的人脸特征和人脸特征对应的节点身份信息进行计算节点间的节点身份信息的匹配。仍然以上文列举为例，人脸特征a在计算节点A上对应的节点身份信息是节点A-a，人脸特征a在计算节点B上对应的节点身份信息是节点B-a，服务器可以根据人脸特征a标定节点身份信息节点A-a和节点B-a是同一个人，并分配统一的标定身份信息为gid-a。步骤S504具体可以通过以下步骤来实现：
根据多个计算节点上传的人脸特征及人脸特征对应的节点身份信息，将相同人脸特征的节点身份信息分配相同的标定身份信息，建立标定身份信息、人脸特征以及节点身份信息间的映射关系；
根据映射关系中的各节点身份信息，查询计算节点上传的节点身份信息合并表中与各节点身份信息实质相同的节点身份信息，将该实质相同的节点身份信息增加到映射关系中。
在本发明的可选实施例中，服务器当再次接收到多个计算节点中任意计算节点上传的人脸特征及人脸特征对应的节点身份信息时，还可以判断映射关系中是否存在与上传的节点身份信息相同的节点身份信息，若存在，则根据映射关系将相同的节点身份信息对应的标定身份信息作为上传的节点身份信息的标定身份；若不存在，则分配与映射关系中包含的标定身份信息不同的新标定身份信息，将新标定身份信息作为上传的节点身份信息的标定身份，并将新标定身份信息、上传的节点身份信息以及对应的人脸特征记录到映射关系中。
在本发明的可选实施例中，多个计算节点将各自的节点日志表上传至服务器，则服务器当再次接收到多个计算节点中任意计算节点上传的人脸特征及人脸特征对应的节点身份信息时，可以根据该任意计算节点的节点日志表，若该任意计算节点在预设时长内确定的同一节点身份信息的数量大于预设数量阈值，则将该同一节点身份信息标定为有效身份信息，对标定为有效身份信息的节点身份信息进一步判断映射关系中是否存在与有效身份信息相同的节点身份信息，若存在，则根据映射关系将相同的节点身份信息对 应的标定身份信息作为有效身份信息的标定身份；若不存在，则分配与映射关系中包含的标定身份信息不同的新标定身份信息，将新标定身份信息作为有效身份信息的标定身份，并将新标定身份信息、有效身份信息以及对应的人脸特征记录到映射关系中。
在本发明的可选实施例中，服务器还可以根据多个计算节点各自的节点日志表，对建立的初始的映射关系中的节点身份信息的有效性进行判断，若节点身份信息具备有效性，则保留；若节点身份信息不具备有效性，则删除。
在本发明的可选实施例中，服务器还可以根据映射关系中新标定身份信息对应的节点身份信息，查询节点身份信息合并表中与新标定身份信息对应的节点身份信息实质相同的节点身份信息，将该实质相同的节点身份信息增加到映射关系中。
这样，服务器进一步可以根据标定身份信息来统计人数，例如，多个计算节点的图像采集装置是采集的商场的图像，则服务器可以根据标定身份信息来统计商场的客流量。又例如，多个计算节点的图像采集装置是采集的商场中商铺的图像，则服务器可以根据标定身份信息来统计商场中商铺的客流量。需要说明的是，此处列举的商场仅是示意性的，用户可以根据实际需求将本发明实施例提供的身份标定系统部署在相应的场景，如热门景点、公共场所等，利用身份标定系统来实现人物身份的标定，进而根据标定的身份来实现场景的客流量的统计和热点的检测。
图6所示为根据本发明另一实施例的应用于服务器的身份标定方法的流程示意图，参见图6，该应用于服务器的身份标定方法可以包括如下步骤S602至S606。
步骤S602，接收包含图像采集装置的计算节点上传的人脸特征及人脸特征对应的节点身份信息；其中，计算节点利用其上的图像采集装置，按照预设图像采集频率采集图像；对采集的图像进行人脸识别，识别出图像中的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息；之后，将人脸特征及人脸特征对应的节点身份信息上传至服务器。
步骤S604，在计算节点上传的人脸特征对应的节点身份信息中，分析节点身份信息是否为有效身份信息，若是，则继续执行步骤S606；若否，则返回步骤S602。
在该步骤中，根据计算节点的节点日志表，若计算节点在预设时长T内确定的同一节点身份信息的数量大于预设数量阈值M，则将该同一节点身份信息标定为有效身份信息。例如，在预设时长10秒内确定的同一节点身份信息的数量大于预设数量阈值7，则将该同一节点身份信息标定为有效身份信息。需要说明的是，T和M可以根据实际需求来设置，此处的举例仅是示意性的，并不对本发明实施例进行限制。
步骤S606，匹配gid列表，进而根据匹配结果，标定分析得到的有效身份信息的标定身份信息。
在该步骤中，gid列表中包括标定身份信息、人脸特征以及节点身份信息间的映射关系，可以判断映射关系中是否存在与有效身份信息相同的节点身份信息，若存在，则根据映射关系将相同的节点身份信息对应的标定身份信息作为有效身份信息的标定身份；若不存在，则分配与映射关系中包含的标定身份信息不同的新标定身份信息，将新标定身份信息作为有效身份信息的标定身份，并将新标定身份信息、有效身份信息以及对应的人脸特征记录到映射关系中。
在本发明的可选实施例中，还可以根据映射关系中新标定身份信息对应的节点身份信息，查询节点身份信息合并表中与新标定身份信息对应的节点身份信息实质相同的节点身份信息，将该实质相同的节点身份信息增加到映射关系中。
根据上述任意一个可选实施例或多个可选实施例的组合，本发明实施例能够达到如下有益效果：
本发明实施例提供的身份标定系统包括了包含图像采集装置的计算节点，以及与计算节点建立通信连接的服务器，在计算节点采集图像并进行图像的人脸识别后，提取人脸特征及分析数据上传至服务器；服务器进行特征匹配及关联，实现人脸特征的身份标定；由此，通过计算节点和服务器的分布式计算提高了系统的效率以及安全性。
并且，本发明实施例在计算节点处，还在人脸识别的图像分析算法的基础上对识别人脸特征确定的节点身份信息进一步分析，解决人在同一位置点的指定范围内，由于人脸的角度以及光线，会确定出不同的节点身份信息的问题，提高计算节点识别的准确性，进而提高了整个系统的身份标定的准确性。
进一步，本发明实施例在服务器处，通过对计算节点的节点身份信息的有效性确定，解决人在同一位置点的指定范围内，由于人脸的角度以及光线，会确定出不同的节点身份信息的问题，提高了整个系统的身份标定的准确性。
以上所述的实施例，对本发明的目的、技术方案和有益效果进行了进一步详细说明，所应理解的是，以上所述仅为本发明的示例性实施例而已，并不用于限制本发明，凡在本发明的精神和原则之内，所做的任何修改、等同替换、改进等，均应包含在本发明的保护范围之内。

Claims (22)

  1. 一种身份标定系统，包括：包含图像采集装置的计算节点，与所述计算节点建立通信连接的服务器；
    所述计算节点，用于利用其上的图像采集装置，按照预设图像采集频率采集图像；对采集的图像进行人脸识别，识别出图像中的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息；之后，将人脸特征及人脸特征对应的节点身份信息上传至所述服务器；
    所述服务器，用于基于所述计算节点上传的人脸特征及人脸特征对应的节点身份信息，对人脸特征的身份进行标定。
  2. 根据权利要求1所述的系统，其中，所述计算节点还用于：
    在人脸特征与节点身份信息的对应关系中，将识别出的人脸特征与所述对应关系中的人脸特征进行匹配；
    若所述对应关系中存在匹配的人脸特征，则将所述对应关系中匹配的人脸特征对应的节点身份信息，确定为识别出的人脸特征对应的节点身份信息；
    若所述对应关系中不存在匹配的人脸特征，则分配一与所述对应关系中包含的节点身份信息不同的新节点身份信息，将所述新节点身份信息确定为识别出的人脸特征对应的节点身份信息，并将所述新节点身份信息及对应的人脸特征记录到所述对应关系中。
  3. 根据权利要求1所述的系统，其中，所述计算节点还用于：
    在确定人脸特征对应的节点身份信息之后，将该节点身份信息、该节点身份信息的出现时间以及该节点身份信息对应的人脸中心坐标保存在本地的节点日志表中。
  4. 根据权利要求3所述的系统，其中，所述计算节点还用于：
    若所述节点日志表中的节点身份信息存在不相同的情况，则记为第一节点身份信息和第二节点身份信息；
    从所述节点日志表中查询所述第一节点身份信息和所述第二节点身份信息各自对应的人脸中心坐标以及出现时间；
    判断所述第二节点身份信息对应的人脸中心坐标相对所述第一节点身份信息对应的人脸中心坐标偏移的偏移量是否小于预设数量的像素点；
    若是，并且相同的所述第一节点身份信息之间的所述第二节点身份信息所持续的帧数小于预设帧数，则将所述第一节点身份信息和所述第二节点身份信息确定为实质相同的节点身份信息，并记录在节点身份信息合并表中。
  5. 根据权利要求4所述的系统，其中，所述服务器还用于：
    根据多个所述计算节点上传的人脸特征及人脸特征对应的节点身份信息，将相同人脸特征的节点身份信息分配相同的标定身份信息，建立标定身份信息、人脸特征以及节点身份信息间的映射关系；
    根据所述映射关系中的各节点身份信息，查询所述计算节点上传的节点身份信息合并表中与各节点身份信息实质相同的节点身份信息，将该实质相同的节点身份信息增加到所述映射关系中。
  6. 根据权利要求5所述的系统，其中，所述服务器还用于：
    当再次接收到多个所述计算节点中任意计算节点上传的人脸特征及人脸特征对应的节点身份信息时，判断所述映射关系中是否存在与上传的节点身份信息相同的节点身份信息，若存在，则根据所述映射关系将相同的节点身份信息对应的标定身份信息作为上传的节点身份信息的标定身份；若不存在，则分配与所述映射关系中包含的标定身份信息不同的新标定身份信息，将新标定身份信息作为上传的节点身份信息的标定身份，并将新标定身份信息、上传的节点身份信息以及对应的人脸特征记录到所述映射关系中。
  7. 根据权利要求5所述的系统，其中，
    所述计算节点还用于：将所述节点日志表上传至所述服务器；
    所述服务器还用于：当再次接收到多个所述计算节点中任意计算节点上传的人脸特征及人脸特征对应的节点身份信息时，根据所述节点日志表，若该任意计算节点在预设时长内确定的同一节点身份信息的数量大于预设数量阈值，则将该同一节点身份信息标定为有效身份信息，对标定为有效身份信息的节点身份信息进一步判断所述映射关系中是否存在与有效身份信息相同的节点身份信息，若存在，则根据所述映射关系将相同的节点身份信息对应的标定身份信息作为有效身份信息的标定身份；若不存在，则分配与所述映射关系中包含的标定身份信息不同的新标定身份信息，将新标定身份信息作为有效身份信息的标定身份，并将新标定身份信息、有效身份信息以及对应的人脸特征记录到所述映射关系中。
  8. 根据权利要求6或7所述的系统，其中，所述服务器还用于：
    根据所述映射关系中新标定身份信息对应的节点身份信息，查询节点身份信息合并表中与新标定身份信息对应的节点身份信息实质相同的节点身份信息，将该实质相同的节点身份信息增加到所述映射关系中。
  9. 根据权利要求1所述的系统，其中，所述计算节点还用于：
    在对采集的图像进行人脸识别时，若检测到图像中包含多个人脸，则启动与多个人脸相同数量的多个工作线程；
    利用所述多个工作线程，识别出图像中多个人脸各自的人脸特征，根据识别出的人 脸特征，确定人脸特征对应的节点身份信息，其中，一个工作线程识别一个人脸的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息。
  10. 根据权利要求1所述的系统，其中，所述计算节点还用于：
    根据预设上传频率，将保存在本地的人脸特征及人脸特征对应的节点身份信息上传至所述服务器，并生成上传日志，从而根据所述上传日志，将后续保存在本地的且未上传至所述服务器的人脸特征及人脸特征对应的节点身份信息上传至所述服务器。
  11. 一种身份标定方法，应用于包含图像采集装置的计算节点，所述方法包括：
    利用图像采集装置，按照预设图像采集频率采集图像；
    对采集的图像进行人脸识别，识别出图像中的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息；
    将人脸特征及人脸特征对应的节点身份信息上传至服务器，从而所述服务器基于所述计算节点上传的人脸特征及人脸特征对应的节点身份信息，对人脸特征的身份进行标定。
  12. 根据权利要求11所述的方法，其中，所述根据识别出的人脸特征，确定人脸特征对应的节点身份信息，包括：
    在人脸特征与节点身份信息的对应关系中，将识别出的人脸特征与所述对应关系中的人脸特征进行匹配；
    若所述对应关系中存在匹配的人脸特征，则将所述对应关系中匹配的人脸特征对应的节点身份信息，确定为识别出的人脸特征对应的节点身份信息；
    若所述对应关系中不存在匹配的人脸特征，则分配一与所述对应关系中包含的节点身份信息不同的新节点身份信息，将所述新节点身份信息确定为识别出的人脸特征对应的节点身份信息，并将所述新节点身份信息及对应的人脸特征记录到所述对应关系中。
  13. 根据权利要求11所述的方法，其中，在确定人脸特征对应的节点身份信息之后，所述方法还包括：
    将该节点身份信息、该节点身份信息的出现时间以及该节点身份信息对应的人脸中心坐标保存在本地的节点日志表中。
  14. 根据权利要求13所述的方法，其中，还包括：
    若所述节点日志表中的节点身份信息存在不相同的情况，则记为第一节点身份信息和第二节点身份信息；
    从所述节点日志表中查询所述第一节点身份信息和所述第二节点身份信息各自对应的人脸中心坐标以及出现时间；
    判断所述第二节点身份信息对应的人脸中心坐标相对所述第一节点身份信息对应的 人脸中心坐标偏移的偏移量是否小于预设数量的像素点；
    若是，并且相同的所述第一节点身份信息之间的所述第二节点身份信息所持续的帧数小于预设帧数，则将所述第一节点身份信息和所述第二节点身份信息确定为实质相同的节点身份信息，并记录在节点身份信息合并表中。
  15. 根据权利要求14所述的方法，其中，还包括：
    将所述节点日志表以及所述节点身份信息合并表上传至所述服务器。
  16. 根据权利要求11所述的方法，其中，所述对采集的图像进行人脸识别，识别出图像中的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息，包括：
    在对采集的图像进行人脸识别时，若检测到图像中包含多个人脸，则启动与多个人脸相同数量的多个工作线程；
    利用所述多个工作线程，识别出图像中多个人脸各自的人脸特征，根据识别出的人脸特征，确定人脸特征对应的节点身份信息，其中，一个工作线程识别一个人脸的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息。
  17. 根据权利要求11所述的方法，其中，所述将人脸特征及人脸特征对应的节点身份信息上传至服务器，包括：
    根据预设上传频率，将保存在本地的人脸特征及人脸特征对应的节点身份信息上传至所述服务器，并生成上传日志，从而根据所述上传日志，将后续保存在本地的且未上传至所述服务器的人脸特征及人脸特征对应的节点身份信息上传至所述服务器。
  18. 一种身份标定方法，应用于服务器，包括：
    接收包含图像采集装置的计算节点上传的人脸特征及人脸特征对应的节点身份信息；其中，所述计算节点利用其上的图像采集装置，按照预设图像采集频率采集图像；对采集的图像进行人脸识别，识别出图像中的人脸特征，并根据识别出的人脸特征，确定人脸特征对应的节点身份信息；之后，将人脸特征及人脸特征对应的节点身份信息上传至所述服务器；
    基于所述计算节点上传的人脸特征及人脸特征对应的节点身份信息，对人脸特征的身份进行标定。
  19. 根据权利要求18所述的方法，其中，基于所述计算节点上传的人脸特征及人脸特征对应的节点身份信息，对人脸特征的身份进行标定，包括：
    根据多个所述计算节点上传的人脸特征及人脸特征对应的节点身份信息，将相同人脸特征的节点身份信息分配相同的标定身份信息，建立标定身份信息、人脸特征以及节点身份信息间的映射关系；
    根据所述映射关系中的各节点身份信息，查询所述计算节点上传的节点身份信息合并表中与各节点身份信息实质相同的节点身份信息，将该实质相同的节点身份信息增加到所述映射关系中。
  20. 根据权利要求19所述的方法，其中，基于所述计算节点上传的人脸特征及人脸特征对应的节点身份信息，对人脸特征的身份进行标定，包括：
    当再次接收到多个所述计算节点中任意计算节点上传的人脸特征及人脸特征对应的节点身份信息时，判断所述映射关系中是否存在与上传的节点身份信息相同的节点身份信息，若存在，则根据所述映射关系将相同的节点身份信息对应的标定身份信息作为上传的节点身份信息的标定身份；若不存在，则分配与所述映射关系中包含的标定身份信息不同的新标定身份信息，将新标定身份信息作为上传的节点身份信息的标定身份，并将新标定身份信息、上传的节点身份信息以及对应的人脸特征记录到所述映射关系中。
  21. 根据权利要求19所述的方法，其中，基于所述计算节点上传的人脸特征及人脸特征对应的节点身份信息，对人脸特征的身份进行标定，包括：
    当再次接收到多个所述计算节点中任意计算节点上传的人脸特征及人脸特征对应的节点身份信息时，根据所述计算节点上传的节点日志表，若该任意计算节点在预设时长内确定的同一节点身份信息的数量大于预设数量阈值，则将该同一节点身份信息标定为有效身份信息，对标定为有效身份信息的节点身份信息进一步判断所述映射关系中是否存在与有效身份信息相同的节点身份信息，若存在，则根据所述映射关系将相同的节点身份信息对应的标定身份信息作为有效身份信息的标定身份；若不存在，则分配与所述映射关系中包含的标定身份信息不同的新标定身份信息，将新标定身份信息作为有效身份信息的标定身份，并将新标定身份信息、有效身份信息以及对应的人脸特征记录到所述映射关系中。
  22. 根据权利要求20或21所述的方法，其中，还包括：
    根据所述映射关系中新标定身份信息对应的节点身份信息，查询节点身份信息合并表中与新标定身份信息对应的节点身份信息实质相同的节点身份信息，将该实质相同的节点身份信息增加到所述映射关系中。
PCT/CN2018/122883 2017-12-27 2018-12-21 一种身份标定系统和方法 WO2019128883A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US16/910,040 US11113509B2 (en) 2017-12-27 2020-06-23 Identity determination system and method

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
CN201711446279.6A CN108009530B (zh) 2017-12-27 2017-12-27 一种身份标定系统和方法
CN201711446279.6 2017-12-27
CN201721872711.3 2017-12-27
CN201721872711.3U CN207764814U (zh) 2017-12-27 2017-12-27 一种身份标定系统

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US16/910,040 Continuation US11113509B2 (en) 2017-12-27 2020-06-23 Identity determination system and method

Publications (1)

Publication Number Publication Date
WO2019128883A1 true WO2019128883A1 (zh) 2019-07-04

Family

ID=67066658

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/122883 WO2019128883A1 (zh) 2017-12-27 2018-12-21 一种身份标定系统和方法

Country Status (2)

Country Link
US (1) US11113509B2 (zh)
WO (1) WO2019128883A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019128883A1 (zh) * 2017-12-27 2019-07-04 苏州欧普照明有限公司 一种身份标定系统和方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9589181B2 (en) * 2013-02-28 2017-03-07 Hitachi Kokusai Electric Inc. Person search method and device for searching person staying on platform
CN106886746A (zh) * 2016-12-27 2017-06-23 浙江宇视科技有限公司 一种识别方法及后端服务器
CN107358160A (zh) * 2017-06-08 2017-11-17 小草数语（北京）科技有限公司 终端监控视频处理方法、监控终端以及服务器
CN108009530A (zh) * 2017-12-27 2018-05-08 欧普照明股份有限公司 一种身份标定系统和方法
CN207764814U (zh) * 2017-12-27 2018-08-24 欧普照明股份有限公司 一种身份标定系统

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2410359A (en) * 2004-01-23 2005-07-27 Sony Uk Ltd Display
TW200929005A (en) * 2007-12-26 2009-07-01 Altek Corp Human face detection and tracking method
CN105578229A (zh) * 2015-12-15 2016-05-11 小米科技有限责任公司 电子设备控制方法和装置
US20170300823A1 (en) * 2016-04-13 2017-10-19 International Business Machines Corporation Determining user influence by contextual relationship of isolated and non-isolated content
US20180069937A1 (en) * 2016-09-02 2018-03-08 VeriHelp, Inc. Event correlation and association using a graph database
US10089521B2 (en) * 2016-09-02 2018-10-02 VeriHelp, Inc. Identity verification via validated facial recognition and graph database
WO2019128883A1 (zh) * 2017-12-27 2019-07-04 苏州欧普照明有限公司 一种身份标定系统和方法
US10452897B1 (en) * 2018-08-06 2019-10-22 Capital One Services, Llc System for verifying the identity of a user

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9589181B2 (en) * 2013-02-28 2017-03-07 Hitachi Kokusai Electric Inc. Person search method and device for searching person staying on platform
CN106886746A (zh) * 2016-12-27 2017-06-23 浙江宇视科技有限公司 一种识别方法及后端服务器
CN107358160A (zh) * 2017-06-08 2017-11-17 小草数语（北京）科技有限公司 终端监控视频处理方法、监控终端以及服务器
CN108009530A (zh) * 2017-12-27 2018-05-08 欧普照明股份有限公司 一种身份标定系统和方法
CN207764814U (zh) * 2017-12-27 2018-08-24 欧普照明股份有限公司 一种身份标定系统

Also Published As

Publication number Publication date
US20200320277A1 (en) 2020-10-08
US11113509B2 (en) 2021-09-07

Similar Documents

Publication Publication Date Title
US10354145B2 (en) Associating a user identity with a mobile device identity
WO2020094091A1 (zh) 一种图像抓拍方法、监控相机及监控系统
JP2020509633A5 (zh)
CN110659560B (zh) 一种关联对象的识别方法及系统
CN102196251B (zh) 智慧城市智能监控方法和系统
TW201508520A (zh) 設置背景圖像的方法及相關的伺服器和系統
US10230922B2 (en) Combining installed audio-visual sensors with ad-hoc mobile audio-visual sensors for smart meeting rooms
US11593920B2 (en) Systems and methods for media privacy
WO2020094088A1 (zh) 一种图像抓拍方法、监控相机及监控系统
WO2018035667A1 (zh) 显示方法、装置、电子设备、计算机程序产品和非暂态计算机可读存储介质
US10360481B2 (en) Unconstrained event monitoring via a network of drones
WO2019128229A1 (zh) 一种视频数据传输、处理方法、装置及终端、服务器
WO2020093829A1 (zh) 开放场景的实时人流统计方法和装置
US20130177219A1 (en) Face Data Acquirer, End User Video Conference Device, Server, Method, Computer Program And Computer Program Product For Extracting Face Data
Yuan Crowd monitoring using mobile phones
US11017241B2 (en) People-flow analysis system and people-flow analysis method
WO2014106384A1 (zh) 一种监控录像信息提供方法、装置及视频监控系统
CN110765134A (zh) 档案建立方法、设备及存储介质
CN108009530B (zh) 一种身份标定系统和方法
CN113516036A (zh) 一种监控区域内目标对象的数量检测方法及装置
WO2019128883A1 (zh) 一种身份标定系统和方法
CN207764814U (zh) 一种身份标定系统
WO2021212760A1 (zh) 确定人员身份类型的方法、装置和电子系统
CN109344281A (zh) 一种基于wifi探针与摄像头技术的数据分析方法
TWI510098B (zh) 攝像裝置及利用其動態追蹤特定物體的方法

Legal Events

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

Ref document number: 18894778

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18894778

Country of ref document: EP

Kind code of ref document: A1