CN111429476B - Method and device for determining action track of target person - Google Patents
Method and device for determining action track of target person Download PDFInfo
- Publication number
- CN111429476B CN111429476B CN201910020875.0A CN201910020875A CN111429476B CN 111429476 B CN111429476 B CN 111429476B CN 201910020875 A CN201910020875 A CN 201910020875A CN 111429476 B CN111429476 B CN 111429476B
- Authority
- CN
- China
- Prior art keywords
- monitoring
- target person
- human body
- face image
- image data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
- H04N7/181—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a plurality of remote sources
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30196—Human being; Person
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30196—Human being; Person
- G06T2207/30201—Face
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method and a device for determining a target person action track, and belongs to the field of intelligent monitoring. The method comprises the following steps: determining a first monitoring point set which is shot to a target person based on reference face image data of the target person and monitoring images shot by all monitoring points; acquiring the human body structural data of the target person, wherein the human body structural data of the target person is acquired according to a monitoring image shot by a monitoring point in the first monitoring point set; determining a second monitoring point set of the target person based on the human body structural data of the target person and monitoring images shot by other monitoring points associated with the monitoring points in the first monitoring point set; and determining the action track of the target person based on the information of the first monitoring point set and the information of the second monitoring point set. By adopting the method and the device, the accuracy of determining the action track of the target person can be improved.
Description
Technical Field
The invention relates to the field of intelligent monitoring, in particular to a method and a device for determining a target person action track.
Background
With the development of image recognition technology, the server can recognize the target person from the monitoring video according to the reference image of the target person.
The server can acquire a plurality of character images from the monitoring video shot by the monitoring point, and further can respectively compare the plurality of character images based on the reference image of the target character to search the target character. When the target person is found, the server can determine the position information of the corresponding monitoring point and the shot time information as track point data, and then determine the action track of the target person according to all track point data so as to assist the case processing of the staff.
In carrying out the invention, the inventors have found that the prior art has at least the following problems:
when determining the action track of a target person, the server generally recognizes a face image of the person. In the actual situation, a plurality of monitoring videos may not shoot face images, and the available track point data is relatively small, so that the accuracy of the action track of the obtained target person is relatively low.
Disclosure of Invention
In order to solve the problems in the prior art, the embodiment of the invention provides a method and a device for determining the action track of a target person. The technical scheme is as follows:
In a first aspect, a method of determining a target person action trajectory is provided, the method comprising:
determining a first monitoring point set which is shot to a target person based on reference face image data of the target person and monitoring images shot by all monitoring points;
acquiring the human body structural data of the target person, wherein the human body structural data of the target person is acquired according to a monitoring image shot by a monitoring point in the first monitoring point set;
determining a second monitoring point set of the target person based on the human body structural data of the target person and monitoring images shot by other monitoring points associated with the monitoring points in the first monitoring point set;
and determining the action track of the target person based on the information of the first monitoring point set and the information of the second monitoring point set.
Optionally, the method further comprises:
if the number of the monitoring points in the first monitoring point set is greater than a first preset threshold, acquiring human body structural data of at least one peer person of the target person according to the monitoring images of the monitoring points in the first monitoring point set, wherein the first preset threshold is greater than 1, and the peer person is a person meeting the action track similarity condition with the target person;
The determining, based on the human body structured data of the target person and the monitoring images captured by the other monitoring points associated with the monitoring points in the first monitoring point set, a second monitoring point set capturing the target person includes:
and determining a second monitoring point set shot to the target person based on the human body structural data of the target person, the human body structural data of the at least one peer person and monitoring images shot by other monitoring points associated with the monitoring points in the first monitoring point set.
Optionally, the obtaining the human body structural data of at least one peer person of the target person according to the monitoring image of the monitoring point in the first monitoring point set includes:
acquiring face image data shot by a plurality of monitoring points in the first monitoring point set and snapshot time corresponding to each piece of face image data;
based on the acquired face image data and the snapshot time corresponding to each piece of face image data, determining face image data meeting the condition that the snapshot time is close to the face image data of the target person;
dividing the face image data with the similarity larger than a second preset threshold value into an image group in the determined face image data;
At least one image group containing the face image data with the number larger than a third preset threshold value is determined, and the human body structural data corresponding to the face image data in the at least one image group is determined to be the human body structural data of at least one peer person of the target person.
Optionally, the determining, based on the human body structural data of the target person, the human body structural data of the at least one peer person, and the monitored images captured by the other monitored points associated with the monitored points in the first monitored point set, the second monitored point set capturing the target person includes:
acquiring human body structural data of other monitoring points related to the monitoring points in the first monitoring point set;
determining all monitoring points which at least meet the following two conditions in the other associated monitoring points as a second monitoring point set:
the human body structural data of the monitoring point has human body structural data with similarity with the human body structural data of the target person being larger than a preset value;
the human body structural data of the monitoring point has human body structural data with similarity larger than a preset value with the human body structural data of the same-row person.
Optionally, the other monitoring points associated with the monitoring points in the first monitoring point set are monitoring points with a distance between any monitoring point in the first monitoring point set being smaller than a fourth preset threshold.
In a second aspect, there is provided an apparatus for determining a target person action trajectory, the apparatus comprising:
the determining module is used for determining a first monitoring point set which is shot to the target person based on the reference face image data of the target person and the monitoring images shot by the monitoring points;
the acquisition module is used for acquiring the human body structural data of the target person, and the human body structural data of the target person is acquired according to the monitoring images shot by the monitoring points in the first monitoring point set;
the determining module is further configured to determine, based on the human body structured data of the target person and monitoring images captured by other monitoring points associated with the monitoring points in the first monitoring point set, that a second monitoring point set of the target person is captured; and determining the action track of the target person based on the information of the first monitoring point set and the information of the second monitoring point set.
Optionally, the acquiring module is further configured to:
If the number of the monitoring points in the first monitoring point set is greater than a first preset threshold, acquiring human body structural data of at least one peer person of the target person according to the monitoring images of the monitoring points in the first monitoring point set, wherein the first preset threshold is greater than 1, and the peer person is a person meeting the action track similarity condition with the target person;
the determining module is used for:
and determining a second monitoring point set shot to the target person based on the human body structural data of the target person, the human body structural data of the at least one peer person and monitoring images shot by other monitoring points associated with the monitoring points in the first monitoring point set.
Optionally, the acquiring module is configured to:
acquiring face image data shot by a plurality of monitoring points in the first monitoring point set and snapshot time corresponding to each piece of face image data;
based on the acquired face image data and the snapshot time corresponding to each piece of face image data, determining face image data meeting the condition that the snapshot time is close to the face image data of the target person;
dividing the face image data with the similarity larger than a second preset threshold value into an image group in the determined face image data;
At least one image group containing the face image data with the number larger than a third preset threshold value is determined, and the human body structural data corresponding to the face image data in the at least one image group is determined to be the human body structural data of at least one peer person of the target person.
Optionally, the determining module is configured to:
acquiring human body structural data of other monitoring points related to the monitoring points in the first monitoring point set;
determining all monitoring points which at least meet the following two conditions in the other associated monitoring points as a second monitoring point set:
the human body structural data of the monitoring point has human body structural data with similarity with the human body structural data of the target person being larger than a preset value;
the human body structural data of the monitoring point has human body structural data with similarity larger than a preset value with the human body structural data of the same-row person.
Optionally, the other monitoring points associated with the monitoring points in the first monitoring point set are monitoring points with a distance between any monitoring point in the first monitoring point set being smaller than a fourth preset threshold.
In a third aspect, a system for determining a target person action trajectory is provided, the system comprising a server and a plurality of monitoring points, wherein:
The monitoring point is used for shooting a monitoring image in a monitoring range and sending the shot monitoring image to the server;
the server is used for determining a first monitoring point set which is shot to the target person based on the reference face imaging data of the target person and the monitoring images shot by each monitoring point; acquiring the human body structural data of the target person, wherein the human body structural data of the target person is acquired according to a monitoring image shot by a monitoring point in the first monitoring point set; determining a second monitoring point set of the target person based on the human body structural data of the target person and monitoring images shot by other monitoring points associated with the monitoring points in the first monitoring point set; and determining the action track of the target person based on the information of the first monitoring point set and the information of the second monitoring point set.
In a fourth aspect, a server is provided, the server comprising a processor and a memory, the memory having stored therein at least one instruction that is loaded and executed by the processor to implement the method of determining a target persona action trajectory as described in the first aspect.
In a fifth aspect, there is provided a computer readable storage medium having stored therein at least one instruction that is loaded and executed by a processor to implement the method of determining a target person action trajectory as described in the first aspect.
The technical scheme provided by the embodiment of the invention has the beneficial effects that:
in the embodiment of the invention, a server determines a first monitoring point set of a shot target person based on reference face image data of the target person and monitoring images shot by all monitoring points, then acquires human body structural data of the target person in the corresponding monitoring images, determines a second monitoring point set of the shot target person based on the human body structural data of the target person and monitoring images shot by other monitoring points, and determines a movement track of the target person based on the first monitoring point set and the second monitoring point set. Therefore, the target person is searched based on the face image data, the human body structural data of the target person is obtained, the target person is searched based on the human body structural data, more monitoring points for shooting the target person can be searched, the action track of the target person is refined, and the accuracy of determining the action track of the target person is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flowchart of a method for determining a target person action trajectory according to an embodiment of the present invention;
FIG. 2 is a schematic illustration of an implementation environment provided by an embodiment of the present invention;
FIG. 3 is a schematic diagram of a target character's action track according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of an apparatus for determining a target person action trajectory according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of a server according to an embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the embodiments of the present invention will be described in further detail with reference to the accompanying drawings.
The embodiment of the invention provides a method for determining the action track of a target person, which can be realized by a server.
The server may include a processor, memory, transceiver, etc. The processor may be a CPU (Central Processing Unit ) or the like, and may be used for processing such as determining the track point data of the target person, acquiring the face image, acquiring the body image, determining the action track of the target person, and the like. The memory may be RAM (Random Access Memory ), flash (Flash memory) or the like, and may be used to store received data, data required in a processing procedure, data generated in a processing procedure, or the like, such as a face image, a body image, track point data, a track of action, or the like. A transceiver, which may be used for data transmission with other devices, may include an antenna, matching circuitry, a modem, etc.
As shown in fig. 1, the process flow of the method may include the following steps:
in step 101, the server determines a first set of monitoring points that are snapped to the target person based on the reference face image data of the target person and the monitoring images captured by the respective monitoring points.
The data in the first monitoring point set may include position information of the monitoring points and capturing time of capturing the target person.
In practice, the monitoring point may be a monitoring camera for capturing monitoring images within a monitoring range. When a worker processes a case, the worker may search the monitoring image of each monitoring point to determine the action track of the target person.
The monitoring points may mainly include three types: face cameras, smart cameras, and normal cameras. The main functions of the three monitoring points can be different, for example, a face camera can acquire face images of the past person through an embedded intelligent algorithm, the intelligent camera can acquire human body images of the past person through the embedded intelligent algorithm, and a common camera can only shoot monitoring images in a monitoring range. However, as long as the angle of the installation of the monitoring camera is proper, the definition of the shot monitoring image meets the requirement, the monitoring image can be processed by a server at the rear end, and the face image or the human body image in the monitoring image can be obtained. The intelligent algorithm for acquiring the face image or the body image may be a deep learning algorithm, which is not limited herein.
When the target person is searched in the monitoring image, the searching is performed based on the face image data, and compared with the searching based on the human body structural data, the searching accuracy is higher. Therefore, in order to find the target person in the monitor image, the user may first upload the reference face image data of the target person to the server. Further, when determining the action track of the target person, the server may acquire the monitoring images captured at the respective monitoring points from the database, and search the monitoring images for the face image data of the target person based on the reference face image data of the target person. The server may add the position information of the monitoring points of the target person and the snapshot time to the first monitoring point set every time the face image data of the target person is found in the monitoring image, indicating that the target person has appeared in the monitoring range. After searching the monitoring video of each monitoring point, the first monitoring point set may only include one monitoring point or may include a plurality of monitoring points.
Optionally, before the face image or the human body image is acquired, a person image in the monitoring image may be acquired, and the corresponding processing may be as follows: and acquiring each person image in the monitoring images shot by each monitoring point, and determining face image data in each person image.
In practice, the monitoring point or server may acquire not only the face image but also the person image of each person when acquiring the face image in the monitoring image. The person image may be an image including the entire human body. When the face image is acquired, the face part of the person image can be acquired; when the human body image is acquired, a human body portion of the human body image may be acquired.
In order to facilitate the calculation of the similarity between images, the images may be mathematically modeled to obtain a string of binary codes, i.e. to obtain an image model, such as a face image model or a body image model. Meanwhile, the attribute information of the image can be obtained through the image recognition technology, for example, the attribute information of the face image can be gender and the like, the attribute information of the human body image can be clothes color, clothes style and the like, and the attribute information can be structural information. The face image data or the human body structured data can comprise the image model and the attribute information, the face camera can acquire the face image model through an intelligent algorithm, and the intelligent camera can acquire the human body image model through the intelligent algorithm in the same way. Face image data or body structured data obtained by a face camera, smart camera or server may be stored in a database. An implementation environment schematic diagram is shown in fig. 2.
The server may include a storage server and a processing server, where the storage server may be configured to store the database, the processing server may be configured to perform a method of determining a target person action trajectory, and the processing server may perform data interaction with the storage server. Of course, both storage and processing may be performed by a single server, and embodiments of the present invention are implemented by a single server as an example.
Optionally, the specific process of determining the first monitoring point set by the server may be as follows: acquiring face image data of each face in a monitoring image shot by each monitoring point; determining the similarity between the reference face image data of the target person and each acquired face image data, and determining target face image data with the similarity larger than a preset threshold value; and adding the position information and the snapshot time of the monitoring points corresponding to the target face image data into the first monitoring point set.
In an implementation, the server may acquire face image data from the database, and then sequentially calculate the similarity between each face image data and the reference face image data of the target person according to a similarity algorithm. Then, the server can select the target face image data with similarity larger than a preset threshold value as the face image data of the searched target person, determine the position information and the snapshot time of the corresponding monitoring points as at least one track point data, and add the track point data into the first monitoring point set so as to be used for generating the action track of the target person.
In step 102, the server obtains the human structured data of the target person.
The human body structural data of the target person are acquired according to monitoring images shot by the monitoring points in the first monitoring point set.
In the implementation, the server searches the face image of the target person in the monitoring image and then obtains the human body structural data of the target person in the corresponding monitoring image. Because the clothing of the target person has a large influence on the human body structural data, the target person can be searched based on the face image data under the condition that the clothing of the target person at the time cannot be known in advance, and then the human body structural data of the target person at the time can be acquired.
Optionally, in the above process, before the face image or the human body image is acquired, the human body image in the monitoring image may be acquired, and then the server may acquire the target human body structural data in the target human body image to which the target face image data belongs as the human body structural data of the target human body.
Of course, if the monitoring point can extract the human body structure data and upload the human body structure data to the server, the server may also acquire the human body structure data sent by the monitoring point. The specific manner of extracting the human body structure data by the monitoring points is the same as the above manner, and will not be described here again.
In step 103, the server determines a second set of monitoring points at which the target person is captured based on the human body structured data of the target person and the monitoring images captured by the other monitoring points associated with the monitoring points in the first set of monitoring points.
The associated monitoring points may be monitoring points satisfying a preset relationship, for example, monitoring points preset as the same block or the same street. Or, other monitoring points associated with the monitoring points in the first monitoring point set may also be monitoring points whose distance from any monitoring point in the first monitoring point set is smaller than a fourth preset threshold. The specific association manner of the monitoring points in this embodiment is not limited.
In an implementation, the number of the human body structural data acquired from the monitoring image may be greater than that of the human face image data, and after the server determines the first monitoring point set according to the human face image data of the target person in the above process, the server may search other human body structural data of the target person in the monitoring images shot by other monitoring points according to the human body structural data of the target person. The server may use the position information of the monitored point of the target person and the photographing time information as one supplementary track point data whenever other human body structural data of the target person is found in the monitored image.
Optionally, the specific process of determining the second monitoring point set by the server may be as follows: acquiring each human body structural data in monitoring images shot by other monitoring points related to the monitoring points in the first monitoring point set; determining the similarity between the human body structural data of the target person and each human body structural data, and determining the human body structural data with the similarity larger than a preset threshold value as other human body structural data of the target person; and adding the position information and the snapshot time of the monitoring points corresponding to other human body structural data into the second monitoring point set.
In an implementation, the server may obtain, from the database, the human body structured data in the monitoring image captured by the other monitoring points associated with the monitoring point in the first monitoring point set, and the specific manner is described above and will not be described herein. Similar to the calculation of the similarity of the face image data, the server may calculate, according to a similarity algorithm, the similarity between the obtained human body structured data and the human body structured data of the target person obtained in step 102. Then, the server can select the human body structural data with similarity larger than a preset threshold value as other human body structural data of the searched target person, and determine the position information and the snapshot time of the corresponding monitoring points as at least one supplementary track point data, and the supplementary track point data is added into the second monitoring point set so as to obtain a more accurate target person action track.
Other human structured data determined in the above process may belong to other figures similar to the target figure's clothing, so that a plurality of possible action trajectories of the target figure may be generated. In some practical cases, the target person may have substantially the same action track as some of the peer persons. Therefore, the human body structural data of the same person can be introduced as an aid, and the supplementary track point data of the action track can be determined so as to determine the action track of the more accurate target person, and the corresponding processing can be as follows: if the number of the monitoring points in the first monitoring point set is larger than a first preset threshold value, acquiring human body structural data of at least one peer person of the target person according to the monitoring images of the monitoring points in the first monitoring point set; a second set of monitoring points at which the target person is captured is determined based on the human body structured data of the target person, the human body structured data of the at least one peer person, and the monitoring images captured by the other monitoring points associated with the monitoring points in the first set of monitoring points.
The first preset threshold is greater than 1, the peer person refers to a person who meets the action track similarity condition with the target person, and the meeting of the action track similarity may mean that the number of identical track points is greater than the third preset threshold.
In an implementation, if the number of the monitoring points in the first monitoring point set is greater than a first preset threshold, where the first preset threshold should be at least greater than 1, people that appear at the monitoring points at the same time except for the target person may be searched in the monitored images of the monitoring points, where the possibility that the people are peer people of the target person is high. Further, the server may extract the human body structured data of at least one peer person from the monitoring image. And then, in the process of searching other human body structural data of the target person, searching whether other human body structural data of the same person exist in the monitoring image at the same time, if so, determining the position information and the snapshot time of the corresponding monitoring points as at least one piece of supplementary track point data, and adding the supplementary track point data into the second monitoring point set.
Alternatively, the specific process of determining the human structured data of the peer person may be as follows: acquiring face image data shot by a plurality of monitoring points in a first monitoring point set and snapshot time corresponding to each piece of face image data; based on the acquired face image data and the snapshot time corresponding to each piece of face image data, determining face image data meeting the condition that the snapshot time approaches to the face image data of the target person; dividing the face image data with the similarity larger than a second preset threshold value into an image group in the determined face image data; at least one image group containing the face image data with the number larger than a third preset threshold value is determined, and the human body structural data corresponding to the face image data in the at least one image group is determined to be the human body structural data of at least one peer person of the target person.
In implementation, for the case that before a face image or a human body image is acquired, a person image in a monitored image may be acquired, and the server may acquire, in monitored points of the first monitored point set, face image data of each person image that satisfies a condition that the snapshot time approaches (for example, within 1 minute before and after) based on each snapshot time in the first monitored point set, to obtain a face image data set (hereinafter referred to as a set) corresponding to each monitored point. Then, the face image data in each set is respectively calculated to be similar to the face image data in other sets. For one face image data, if the face image data with the similarity greater than the second similarity threshold value is found in other sets, the face image data can be considered to belong to the same person, and can be divided into the same image group to represent one person. If the number of face image data in the image group is greater than a third preset threshold, the person may be considered as a peer person of the target person. Furthermore, the server can obtain corresponding human body structural data from the person images of the same-party persons, and the human body structural data of the same-party persons are obtained. Of course, the same person under the determination of the server may be one or a plurality of persons.
Optionally, the specific process of determining the second set of monitoring points in combination with the human body structured data of the peer person may be as follows: acquiring human body structural data of other monitoring points related to the monitoring points in the first monitoring point set; all monitoring points which at least meet the following two conditions in the other associated monitoring points are determined as a second monitoring point set. The two conditions are as follows:
(1) The human body structural data of the monitoring point has human body structural data with similarity with the human body structural data of the target person being larger than a preset value;
(2) The human body structural data of the monitoring point has human body structural data with similarity larger than a preset value with the human body structural data of the same person.
In implementation, after the server determines the face image data of the downlink person in the above process, the server may acquire the human body structural data in the image of the person to which the server belongs, and determine the human body structural data as the human body structural data of the person of the same line. And a peer character may have multiple pieces of human structured data, and other human structured data of the peer character may be subsequently found based on the multiple pieces of human structured data.
The server may acquire, from the database, human body structured data in a monitoring image captured by monitoring points within a certain range around the monitoring point determined as the track point. The server may calculate the similarity between the obtained human body structured data and the human body structured data of the target person obtained in step 102 according to the similarity algorithm. Then, the server can select the human body structural data with similarity larger than a preset value as other human body structural data of the searched target person.
The server can also calculate the similarity with the human body structural data of the same person in the obtained human body structural data, and can select the human body structural data with the similarity larger than a preset value as the other human body structural data of the searched same person.
If other human body structural data of the target person and the peer person exist in the monitoring image of one monitoring point at the same time, the server can determine the position information and the snapshot time of the corresponding monitoring point as at least one supplementary track point data, and the supplementary track point data is added to the second monitoring point set so as to obtain a more accurate target person action track. The above-mentioned other human body structured data of the simultaneous existence target person and the peer person may be the simultaneous existence of the target person and all the peer persons, or the simultaneous existence of the target person and at least one peer person, which is not limited herein.
It should be noted that all the thresholds involved in the above process may be set by a skilled person according to actual requirements, and are not limited herein.
In step 104, the server determines a trajectory of the action of the target person based on the information of the first set of monitoring points and the information of the second set of monitoring points.
In an implementation, the server may arrange each monitoring point according to the position information and the snapshot time of the monitoring points in the first monitoring point set and the second monitoring point set, and use each monitoring point and the snapshot time as information of a track point, so as to generate an action track of the target person. And the server can feed back the data of each track point in the action track to the user, or display the action track of the target person on the map after rendering, so that the user can analyze and judge. A schematic diagram of the action track of the target person is shown in fig. 3.
In the embodiment of the invention, a server determines a first monitoring point set of a shot target person based on reference face image data of the target person and monitoring images shot by all monitoring points, then acquires human body structural data of the target person in the corresponding monitoring images, determines a second monitoring point set of the shot target person based on the human body structural data of the target person and monitoring images shot by other monitoring points, and determines a movement track of the target person based on the first monitoring point set and the second monitoring point set. Therefore, the target person is searched based on the face image data, the human body structural data of the target person is obtained, the target person is searched based on the human body structural data, more monitoring points for shooting the target person can be searched, the action track of the target person is refined, and the accuracy of determining the action track of the target person is improved.
Based on the same technical concept, the embodiment of the invention also provides a device for determining the action track of the target person, which can be the server in the embodiment of the method. As shown in fig. 4, the apparatus includes:
a determining module 410, configured to determine a first set of monitoring points that capture a target person based on reference face image data of the target person and monitoring images captured by the monitoring points;
an obtaining module 420, configured to obtain the human body structural data of the target person, where the human body structural data of the target person is obtained according to a monitoring image captured by a monitoring point in the first monitoring point set;
the determining module 410 is further configured to determine, based on the human body structured data of the target person and the monitoring images captured by the other monitoring points associated with the monitoring points in the first set of monitoring points, that a second set of monitoring points of the target person is captured; and determining the action track of the target person based on the information of the first monitoring point set and the information of the second monitoring point set.
Optionally, the obtaining module 420 is further configured to:
if the number of the monitoring points in the first monitoring point set is greater than a first preset threshold, acquiring human body structural data of at least one peer person of the target person according to the monitoring images of the monitoring points in the first monitoring point set, wherein the first preset threshold is greater than 1, and the peer person is a person meeting the action track similarity condition with the target person;
The determining module 410 is configured to:
and determining a second monitoring point set shot to the target person based on the human body structural data of the target person, the human body structural data of the at least one peer person and monitoring images shot by other monitoring points associated with the monitoring points in the first monitoring point set.
Optionally, the obtaining module 420 is configured to:
acquiring face image data shot by a plurality of monitoring points in the first monitoring point set and snapshot time corresponding to each piece of face image data;
based on the acquired face image data and the snapshot time corresponding to each piece of face image data, determining face image data meeting the condition that the snapshot time is close to the face image data of the target person;
dividing the face image data with the similarity larger than a second preset threshold value into an image group in the determined face image data;
at least one image group containing the face image data with the number larger than a third preset threshold value is determined, and the human body structural data corresponding to the face image data in the at least one image group is determined to be the human body structural data of at least one peer person of the target person.
Optionally, the determining module 410 is configured to:
acquiring human body structural data of other monitoring points related to the monitoring points in the first monitoring point set;
determining all monitoring points which at least meet the following two conditions in the other associated monitoring points as a second monitoring point set:
the human body structural data of the monitoring point has human body structural data with similarity with the human body structural data of the target person being larger than a preset value;
the human body structural data of the monitoring point has human body structural data with similarity larger than a preset value with the human body structural data of the same-row person.
Optionally, the other monitoring points associated with the monitoring points in the first monitoring point set are monitoring points with a distance between any monitoring point in the first monitoring point set being smaller than a fourth preset threshold.
The specific manner in which the various modules perform the operations in the apparatus of the above embodiments have been described in detail in connection with the embodiments of the method, and will not be described in detail herein.
In the embodiment of the invention, a server determines a first monitoring point set of a shot target person based on reference face image data of the target person and monitoring images shot by all monitoring points, then acquires human body structural data of the target person in the corresponding monitoring images, determines a second monitoring point set of the shot target person based on the human body structural data of the target person and monitoring images shot by other monitoring points, and determines a movement track of the target person based on the first monitoring point set and the second monitoring point set. Therefore, the target person is searched based on the face image data, the human body structural data of the target person is obtained, the target person is searched based on the human body structural data, more monitoring points for shooting the target person can be searched, the action track of the target person is refined, and the accuracy of determining the action track of the target person is improved.
It should be noted that: the device for determining the action track of the target person provided in the above embodiment is only exemplified by the division of the above functional modules when determining the action track of the target person, and in practical application, the above functional allocation may be performed by different functional modules according to needs, i.e. the internal structure of the server is divided into different functional modules to complete all or part of the functions described above. In addition, the device for determining the action track of the target person provided in the above embodiment belongs to the same concept as the method embodiment for determining the action track of the target person, and the detailed implementation process of the device is referred to in the method embodiment and will not be described herein.
Based on the same technical concept, the embodiment of the invention also provides a system for determining the action track of the target person, which can comprise a server and a plurality of monitoring points, wherein:
the monitoring point is used for shooting a monitoring image in a monitoring range and sending the shot monitoring image to the server;
the server is used for determining a first monitoring point set which is shot to the target person based on the reference face imaging data of the target person and the monitoring images shot by each monitoring point; acquiring the human body structural data of the target person, wherein the human body structural data of the target person is acquired according to a monitoring image shot by a monitoring point in the first monitoring point set; determining a second monitoring point set of the target person based on the human body structural data of the target person and monitoring images shot by other monitoring points associated with the monitoring points in the first monitoring point set; and determining the action track of the target person based on the information of the first monitoring point set and the information of the second monitoring point set.
Fig. 5 is a schematic structural diagram of a server according to an embodiment of the present invention, where the server 500 may have a relatively large difference due to different configurations or performances, and may include one or more processors (central processing units, CPU) 501 and one or more memories 502, where at least one instruction is stored in the memories 502, and the at least one instruction is loaded and executed by the processors 501 to implement the following method steps for determining a target person action track:
determining a first monitoring point set which is shot to a target person based on reference face image data of the target person and monitoring images shot by all monitoring points;
acquiring the human body structural data of the target person, wherein the human body structural data of the target person is acquired according to a monitoring image shot by a monitoring point in the first monitoring point set;
determining a second monitoring point set of the target person based on the human body structural data of the target person and monitoring images shot by other monitoring points associated with the monitoring points in the first monitoring point set;
and determining the action track of the target person based on the information of the first monitoring point set and the information of the second monitoring point set.
Optionally, the at least one instruction is loaded and executed by the processor 501 to implement the following method steps:
if the number of the monitoring points in the first monitoring point set is greater than a first preset threshold, acquiring human body structural data of at least one peer person of the target person according to the monitoring images of the monitoring points in the first monitoring point set, wherein the first preset threshold is greater than 1, and the peer person is a person meeting the action track similarity condition with the target person;
and determining a second monitoring point set shot to the target person based on the human body structural data of the target person, the human body structural data of the at least one peer person and monitoring images shot by other monitoring points associated with the monitoring points in the first monitoring point set.
Optionally, the at least one instruction is loaded and executed by the processor 501 to implement the following method steps:
acquiring face image data shot by a plurality of monitoring points in the first monitoring point set and snapshot time corresponding to each piece of face image data;
based on the acquired face image data and the snapshot time corresponding to each piece of face image data, determining face image data meeting the condition that the snapshot time is close to the face image data of the target person;
Dividing the face image data with the similarity larger than a second preset threshold value into an image group in the determined face image data;
at least one image group containing the face image data with the number larger than a third preset threshold value is determined, and the human body structural data corresponding to the face image data in the at least one image group is determined to be the human body structural data of at least one peer person of the target person.
Optionally, the at least one instruction is loaded and executed by the processor 501 to implement the following method steps:
acquiring human body structural data of other monitoring points related to the monitoring points in the first monitoring point set;
determining all monitoring points which at least meet the following two conditions in the other associated monitoring points as a second monitoring point set:
the human body structural data of the monitoring point has human body structural data with similarity with the human body structural data of the target person being larger than a preset value;
the human body structural data of the monitoring point has human body structural data with similarity larger than a preset value with the human body structural data of the same-row person.
Optionally, the other monitoring points associated with the monitoring points in the first monitoring point set are monitoring points with a distance between any monitoring point in the first monitoring point set being smaller than a fourth preset threshold.
In the embodiment of the invention, a server determines a first monitoring point set of a shot target person based on reference face image data of the target person and monitoring images shot by all monitoring points, then acquires human body structural data of the target person in the corresponding monitoring images, determines a second monitoring point set of the shot target person based on the human body structural data of the target person and monitoring images shot by other monitoring points, and determines a movement track of the target person based on the first monitoring point set and the second monitoring point set. Therefore, the target person is searched based on the face image data, the human body structural data of the target person is obtained, the target person is searched based on the human body structural data, more monitoring points for shooting the target person can be searched, the action track of the target person is refined, and the accuracy of determining the action track of the target person is improved.
It will be understood by those skilled in the art that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program for instructing relevant hardware, where the program may be stored in a computer readable storage medium, and the storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.
The foregoing description of the preferred embodiments of the invention is not intended to limit the invention to the precise form disclosed, and any such modifications, equivalents, and alternatives falling within the spirit and scope of the invention are intended to be included within the scope of the invention.
Claims (7)
1. A method of determining a target person action trajectory, the method comprising:
acquiring face image data of each face in a monitoring image shot by each monitoring point; determining the similarity between the reference face image data of the target person and each acquired face image data, and determining target face image data with the similarity larger than a preset threshold value; adding the position information and the snapshot time of the monitoring points corresponding to the target face image data into a first monitoring point set;
acquiring the human body structural data of the target person, wherein the human body structural data of the target person is acquired according to a monitoring image shot by a monitoring point in the first monitoring point set;
if the number of the monitoring points in the first monitoring point set is larger than a first preset threshold value, acquiring face image data shot by a plurality of monitoring points in the first monitoring point set and snapshot time corresponding to each piece of face image data;
Based on the acquired face image data and the snapshot time corresponding to each piece of face image data, determining face image data meeting the condition that the snapshot time is close to the face image data of the target person;
dividing the face image data with the similarity larger than a second preset threshold value into an image group in the determined face image data;
determining at least one image group containing face image data with the number larger than a third preset threshold, and determining human body structural data corresponding to the face image data in the at least one image group as human body structural data of at least one peer person of the target person, wherein the first preset threshold is larger than 1, and the peer person is a person meeting the action track similarity condition with the target person;
acquiring human body structural data of other monitoring points related to the monitoring points in the first monitoring point set;
determining all monitoring points which at least meet the following two conditions in the other associated monitoring points as a second monitoring point set:
the human body structural data of the monitoring point has human body structural data with similarity with the human body structural data of the target person being larger than a preset value;
The human body structural data of the monitoring point has human body structural data with similarity larger than a preset value with the human body structural data of the same-row characters;
according to the position information and the snapshot time of the monitoring points in the first monitoring point set and the second monitoring point set, the monitoring points and the snapshot time in the first monitoring set and the second monitoring set are used as track point information, all the monitoring points in the first monitoring set and the second monitoring set are arranged to generate the action track of the target person, and the action track of the target person is displayed on a map after rendering.
2. The method of claim 1, wherein the other monitoring points associated with the monitoring points in the first set of monitoring points are monitoring points having a distance to any monitoring point in the first set of monitoring points less than a fourth preset threshold.
3. An apparatus for determining a target person action trajectory, the apparatus comprising:
the determining module is used for acquiring the face image data of each monitored image shot by each monitored point; determining the similarity between the reference face image data of the target person and each acquired face image data, and determining target face image data with the similarity larger than a preset threshold value; adding the position information and the snapshot time of the monitoring points corresponding to the target face image data into a first monitoring point set;
The acquisition module is used for acquiring the human body structural data of the target person, and the human body structural data of the target person is acquired according to the monitoring images shot by the monitoring points in the first monitoring point set; if the number of the monitoring points in the first monitoring point set is larger than a first preset threshold value, acquiring face image data shot by a plurality of monitoring points in the first monitoring point set and snapshot time corresponding to each piece of face image data; based on the acquired face image data and the snapshot time corresponding to each piece of face image data, determining face image data meeting the condition that the snapshot time is close to the face image data of the target person; dividing the face image data with the similarity larger than a second preset threshold value into an image group in the determined face image data; determining at least one image group containing face image data with the number larger than a third preset threshold, and determining human body structural data corresponding to the face image data in the at least one image group as human body structural data of at least one peer person of the target person, wherein the first preset threshold is larger than 1, and the peer person is a person meeting the action track similarity condition with the target person;
The determining module is further configured to obtain human body structured data of other monitoring points associated with the monitoring point in the first monitoring point set; determining all monitoring points which at least meet the following two conditions in the other associated monitoring points as a second monitoring point set: the human body structural data of the monitoring point has human body structural data with similarity with the human body structural data of the target person being larger than a preset value; the human body structural data of the monitoring point has human body structural data with similarity larger than a preset value with the human body structural data of the same-row characters; according to the position information and the snapshot time of the monitoring points in the first monitoring point set and the second monitoring point set, the monitoring points and the snapshot time in the first monitoring set and the second monitoring set are used as track point information, all the monitoring points in the first monitoring set and the second monitoring set are arranged to generate the action track of the target person, and the action track of the target person is displayed on a map after rendering.
4. The apparatus of claim 3, wherein the other monitoring points associated with monitoring points in the first set of monitoring points are monitoring points having a distance from any monitoring point in the first set of monitoring points less than a fourth preset threshold.
5. A system for determining a target character action trajectory, the system comprising a server and a plurality of monitoring points, wherein:
the monitoring point shoots a monitoring image in a monitoring range and sends the shot monitoring image to the server;
the server is used for acquiring the face image data of each person in the monitoring image shot by each monitoring point; determining the similarity between the reference face image data of the target person and each acquired face image data, and determining target face image data with the similarity larger than a preset threshold value; adding the position information and the snapshot time of the monitoring points corresponding to the target face image data into a first monitoring point set; acquiring the human body structural data of the target person, wherein the human body structural data of the target person is acquired according to a monitoring image shot by a monitoring point in the first monitoring point set; if the number of the monitoring points in the first monitoring point set is larger than a first preset threshold value, acquiring face image data shot by a plurality of monitoring points in the first monitoring point set and snapshot time corresponding to each piece of face image data; based on the acquired face image data and the snapshot time corresponding to each piece of face image data, determining face image data meeting the condition that the snapshot time is close to the face image data of the target person; dividing the face image data with the similarity larger than a second preset threshold value into an image group in the determined face image data; determining at least one image group containing face image data with the number larger than a third preset threshold, and determining human body structural data corresponding to the face image data in the at least one image group as human body structural data of at least one peer person of the target person, wherein the first preset threshold is larger than 1, and the peer person is a person meeting the action track similarity condition with the target person; acquiring human body structural data of other monitoring points related to the monitoring points in the first monitoring point set; determining all monitoring points which at least meet the following two conditions in the other associated monitoring points as a second monitoring point set: the human body structural data of the monitoring point has human body structural data with similarity with the human body structural data of the target person being larger than a preset value; the human body structural data of the monitoring point has human body structural data with similarity larger than a preset value with the human body structural data of the same-row characters; according to the position information and the snapshot time of the monitoring points in the first monitoring point set and the second monitoring point set, the monitoring points and the snapshot time in the first monitoring set and the second monitoring set are used as track point information, all the monitoring points in the first monitoring set and the second monitoring set are arranged to generate the action track of the target person, and the action track of the target person is displayed on a map after rendering.
6. A server comprising a processor and a memory having stored therein at least one instruction that is loaded and executed by the processor to implement the method of determining a target person action trajectory as claimed in any one of claims 1 to 2.
7. A computer readable storage medium having stored therein at least one instruction loaded and executed by a processor to implement the method of determining a target person action trajectory as claimed in any one of claims 1 to 2.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910020875.0A CN111429476B (en) | 2019-01-09 | 2019-01-09 | Method and device for determining action track of target person |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910020875.0A CN111429476B (en) | 2019-01-09 | 2019-01-09 | Method and device for determining action track of target person |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111429476A CN111429476A (en) | 2020-07-17 |
| CN111429476B true CN111429476B (en) | 2023-10-20 |
Family
ID=71545914
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910020875.0A Active CN111429476B (en) | 2019-01-09 | 2019-01-09 | Method and device for determining action track of target person |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111429476B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112040186B (en) * | 2020-08-28 | 2023-01-31 | 北京市商汤科技开发有限公司 | Method, device and equipment for determining activity area of target object and storage medium |
| CN112380901A (en) * | 2020-10-10 | 2021-02-19 | 杭州翔毅科技有限公司 | Behavior track generation method, behavior track generation equipment, storage medium and device |
| CN117690172A (en) * | 2020-10-19 | 2024-03-12 | 杭州海康威视数字技术股份有限公司 | Associated target identification method and device, electronic equipment and storage medium |
| CN112417977B (en) * | 2020-10-26 | 2023-01-17 | 青岛聚好联科技有限公司 | Target object searching method and terminal |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20150055271A (en) * | 2013-11-13 | 2015-05-21 | 현대모비스 주식회사 | Apparatus for determining motion characteristics of target and device for controlling driving route of vehicle with the said apparatus |
| CN104794458A (en) * | 2015-05-07 | 2015-07-22 | 北京丰华联合科技有限公司 | Fuzzy video person identifying method |
| CN105139040A (en) * | 2015-10-13 | 2015-12-09 | 商汤集团有限公司 | Queuing state information detection method and system thereof |
| CN106874347A (en) * | 2016-12-26 | 2017-06-20 | 深圳市深网视界科技有限公司 | A kind of method and system for matching characteristics of human body and MAC Address |
| CN107016374A (en) * | 2017-04-12 | 2017-08-04 | 电子科技大学 | Intelligent Measurement tracking and the generation method of space-time track towards specific objective |
| CN107292252A (en) * | 2017-06-09 | 2017-10-24 | 南京华捷艾米软件科技有限公司 | A kind of personal identification method of autonomous learning |
| CN107292240A (en) * | 2017-05-24 | 2017-10-24 | 深圳市深网视界科技有限公司 | It is a kind of that people's method and system are looked for based on face and human bioequivalence |
| CN107437075A (en) * | 2017-07-29 | 2017-12-05 | 安徽博威康信息技术有限公司 | A kind of risk alarm system based on daily behavior track |
| CN107480246A (en) * | 2017-08-10 | 2017-12-15 | 北京中航安通科技有限公司 | A kind of recognition methods of associate people and device |
| WO2017219679A1 (en) * | 2016-06-20 | 2017-12-28 | 杭州海康威视数字技术股份有限公司 | Method and device for establishing correspondence between rfid tags and persons, and method and device for trajectory tracking |
| CN108229335A (en) * | 2017-12-12 | 2018-06-29 | 深圳市商汤科技有限公司 | It is associated with face identification method and device, electronic equipment, storage medium, program |
| CN108875548A (en) * | 2018-04-18 | 2018-11-23 | 科大讯飞股份有限公司 | Personage's orbit generation method and device, storage medium, electronic equipment |
| CN108897777A (en) * | 2018-06-01 | 2018-11-27 | 深圳市商汤科技有限公司 | Target object method for tracing and device, electronic equipment and storage medium |
| CN109117714A (en) * | 2018-06-27 | 2019-01-01 | 北京旷视科技有限公司 | A kind of colleague's personal identification method, apparatus, system and computer storage medium |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6276519B2 (en) * | 2013-05-22 | 2018-02-07 | 株式会社 日立産業制御ソリューションズ | Person counting device and human flow line analyzing device |
| AU2013242830B2 (en) * | 2013-10-10 | 2016-11-24 | Canon Kabushiki Kaisha | A method for improving tracking in crowded situations using rival compensation |
| GB2545658A (en) * | 2015-12-18 | 2017-06-28 | Canon Kk | Methods, devices and computer programs for tracking targets using independent tracking modules associated with cameras |
-
2019
- 2019-01-09 CN CN201910020875.0A patent/CN111429476B/en active Active
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20150055271A (en) * | 2013-11-13 | 2015-05-21 | 현대모비스 주식회사 | Apparatus for determining motion characteristics of target and device for controlling driving route of vehicle with the said apparatus |
| CN104794458A (en) * | 2015-05-07 | 2015-07-22 | 北京丰华联合科技有限公司 | Fuzzy video person identifying method |
| CN105139040A (en) * | 2015-10-13 | 2015-12-09 | 商汤集团有限公司 | Queuing state information detection method and system thereof |
| WO2017219679A1 (en) * | 2016-06-20 | 2017-12-28 | 杭州海康威视数字技术股份有限公司 | Method and device for establishing correspondence between rfid tags and persons, and method and device for trajectory tracking |
| CN107527075A (en) * | 2016-06-20 | 2017-12-29 | 杭州海康威视数字技术股份有限公司 | RFID label tag is established with personnel's corresponding relation and trajectory track method and device |
| CN106874347A (en) * | 2016-12-26 | 2017-06-20 | 深圳市深网视界科技有限公司 | A kind of method and system for matching characteristics of human body and MAC Address |
| CN107016374A (en) * | 2017-04-12 | 2017-08-04 | 电子科技大学 | Intelligent Measurement tracking and the generation method of space-time track towards specific objective |
| CN107292240A (en) * | 2017-05-24 | 2017-10-24 | 深圳市深网视界科技有限公司 | It is a kind of that people's method and system are looked for based on face and human bioequivalence |
| CN107292252A (en) * | 2017-06-09 | 2017-10-24 | 南京华捷艾米软件科技有限公司 | A kind of personal identification method of autonomous learning |
| CN107437075A (en) * | 2017-07-29 | 2017-12-05 | 安徽博威康信息技术有限公司 | A kind of risk alarm system based on daily behavior track |
| CN107480246A (en) * | 2017-08-10 | 2017-12-15 | 北京中航安通科技有限公司 | A kind of recognition methods of associate people and device |
| CN108229335A (en) * | 2017-12-12 | 2018-06-29 | 深圳市商汤科技有限公司 | It is associated with face identification method and device, electronic equipment, storage medium, program |
| CN108875548A (en) * | 2018-04-18 | 2018-11-23 | 科大讯飞股份有限公司 | Personage's orbit generation method and device, storage medium, electronic equipment |
| CN108897777A (en) * | 2018-06-01 | 2018-11-27 | 深圳市商汤科技有限公司 | Target object method for tracing and device, electronic equipment and storage medium |
| CN109117714A (en) * | 2018-06-27 | 2019-01-01 | 北京旷视科技有限公司 | A kind of colleague's personal identification method, apparatus, system and computer storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111429476A (en) | 2020-07-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111429476B (en) | Method and device for determining action track of target person | |
| US11354901B2 (en) | Activity recognition method and system | |
| US10755086B2 (en) | Picture ranking method, and terminal | |
| CN109426785B (en) | Human body target identity recognition method and device | |
| US20180061076A1 (en) | Fast multi-object detection and tracking system | |
| CN111950321B (en) | Gait recognition method, device, computer equipment and storage medium | |
| CN108875507B (en) | Pedestrian tracking method, apparatus, system, and computer-readable storage medium | |
| WO2016199662A1 (en) | Image information processing system | |
| CN111814655B (en) | Target re-identification method, network training method thereof and related device | |
| CN109426787A (en) | A kind of human body target track determines method and device | |
| WO2019033567A1 (en) | Method for capturing eyeball movement, device and storage medium | |
| CN109033985A (en) | Processing method, device, equipment, system and the storage medium of commodity identification | |
| CN111814690A (en) | Target re-identification method and device and computer readable storage medium | |
| CN114333039B (en) | Method, device and medium for clustering human images | |
| US20220319232A1 (en) | Apparatus and method for providing missing child search service based on face recognition using deep-learning | |
| US20190355130A1 (en) | Image processing device, image processing method, and recording medium storing program | |
| CN115391596A (en) | Video archive generation method and device and storage medium | |
| CN114387296A (en) | Target track tracking method and device, computer equipment and storage medium | |
| CN111666786B (en) | Image processing method, device, electronic equipment and storage medium | |
| WO2014062969A1 (en) | Optimizing photos based on an attention map acquired by eye-tracking | |
| CN113887384A (en) | Pedestrian trajectory analysis method, device, equipment and medium based on multi-trajectory fusion | |
| CN113313062A (en) | Path acquisition method, device, system, electronic equipment and storage medium | |
| CN113591620A (en) | Early warning method, device and system based on integrated mobile acquisition equipment | |
| CN113111689A (en) | Sample mining method, device, equipment and storage medium | |
| CN116433939B (en) | Sample image generation method, training method, recognition method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |