CN109145844A - Archive management method, device and electronic equipment for city safety monitoring - Google Patents
Archive management method, device and electronic equipment for city safety monitoring Download PDFInfo
- Publication number
- CN109145844A CN109145844A CN201810999452.3A CN201810999452A CN109145844A CN 109145844 A CN109145844 A CN 109145844A CN 201810999452 A CN201810999452 A CN 201810999452A CN 109145844 A CN109145844 A CN 109145844A
- Authority
- CN
- China
- Prior art keywords
- cluster
- similarity
- group
- facial image
- result
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
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
- G06V40/172—Classification, e.g. identification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Human Computer Interaction (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of archive management methods, device and electronic equipment for city safety monitoring, it is related to technical field of data administration, archive management method for city safety monitoring includes: to carry out recognition of face to multiple facial images, obtains multiple feature vectors；Calculate the vector distance between the multiple feature vector；The similarity between the multiple facial image is determined according to the vector distance；Cluster is merged to the multiple facial image according to the similarity, obtains agglomerative clustering result；Several Profiles are established according to the agglomerative clustering result, the data precision for solving Profile existing in the prior art is lower to influence the technical issues of city safety monitoring works.
Description
Technical field
The present invention relates to technical field of data administration, more particularly, to a kind of file administration side for city safety monitoring
Method, device and electronic equipment.
Background technique
File administration is also known as archives work, be archives (room) directly to profile entity data, archives electronic data etc. into
Row update etc. manages the general name of every vocational work.It is multi-party that Profile recites personally identifiable information, personal images information etc.
The documents and electronic data in face are the important evidences for recording personal considerations.It may also be said that Profile is personal information
Repository, what it summarized reflects personal overall picture.
Furthermore safety defense monitoring system is that transmission video is believed in the loop of its closure using optical fiber, coaxial cable or microwave
Number, and the system for constituting independent completion is shown and recorded from camera shooting to image.It can in real time, image, be truly reflected it is monitored
Object, not only greatly extends the viewing distance of human eye, and expands the function of human eye, it can be under rugged environment
Instead of manually carrying out long-time monitoring, allow people it can be seen that being monitored all situations actually occurred at scene, and pass through video recording
Machine is recorded.Alarm system equipment alarms to illegal invasion simultaneously, and the alarm signal of generation inputs alarm host machine, alarm
Host triggering monitoring system is recorded a video and is recorded.
But in the case where the extensive Profile of processing City-level is established, portrait quantity is excessively huge, such as 10,000,000,000 grades
Quantity, therefore, it is larger that the Profile for city safety monitoring establishes work difficulty, it is easy to occur the same person
There is situations such as image information of different people in people's archives, to cause the data corruption between multiple Profiles, cause a
The data precision of people's archives is lower, to influence the work of city safety monitoring.
Summary of the invention
In view of this, the purpose of the present invention is to provide a kind of archive management methods for city safety monitoring, device
And electronic equipment, the data precision to solve Profile existing in the prior art are lower to influence city security protection prison
The technical issues of controlling work.
In a first aspect, the embodiment of the invention provides a kind of archive management methods for city safety monitoring, comprising:
Recognition of face is carried out to multiple facial images, obtains multiple feature vectors；
Calculate the vector distance between the multiple feature vector；
The similarity between the multiple facial image is determined according to the vector distance；
Cluster is merged to the multiple facial image according to the similarity, obtains agglomerative clustering result；
Several Profiles are established according to the agglomerative clustering result.
With reference to first aspect, the embodiment of the invention provides the first possible embodiments of first aspect, wherein institute
It states and recognition of face is carried out to multiple facial images, before obtaining multiple feature vectors, further includes:
Structuring parsing is carried out to monitor video, obtains multiple facial images.
With reference to first aspect, the embodiment of the invention provides second of possible embodiments of first aspect, wherein institute
It states and cluster is merged to the multiple facial image according to the similarity, obtain agglomerative clustering result, comprising:
Greedy algorithm is based on according to the similarity, cluster is merged to the multiple facial image, obtain by multiple poly-
The preliminary clusters result that class group is constituted；
Cluster is merged to the facial image between the multiple phylogenetic group, obtains cluster result between group, and will be described
Cluster result is determined as agglomerative clustering result between group.
With reference to first aspect, the embodiment of the invention provides the third possible embodiments of first aspect, wherein institute
It states and cluster is merged to the facial image between the multiple phylogenetic group, obtain cluster result between group, comprising:
In each phylogenetic group in the preliminary clusters result, several representative character images are chosen；
The similarity for calculating the representative character image between every two phylogenetic group, obtains several representative image phases
Like degree；
The average value for calculating several representative image similarities, obtains similarity between group；
Cluster is merged to the facial image between phylogenetic group according to similarity between described group, obtains clustering knot between group
Fruit.
With reference to first aspect, the embodiment of the invention provides the 4th kind of possible embodiments of first aspect, wherein institute
It states in each phylogenetic group in the preliminary clusters result, chooses several representative character images, comprising:
In each phylogenetic group in the preliminary clusters result, according to picture quality, temporal information, location information, people
At least one of face angle, personage's posture, using electing several representative character images of algorithm picks.
With reference to first aspect, the embodiment of the invention provides the 5th kind of possible embodiments of first aspect, wherein institute
It states and cluster is merged to the facial image between phylogenetic group according to similarity between described group, obtain cluster result between group, comprising:
According to the temporal information and/or location information of the facial image, judge whether the face figure between phylogenetic group
As merging, auxiliary judgment result is obtained；
According to similarity between described group, the facial image between the phylogenetic group is carried out based on the auxiliary judgment result
Agglomerative clustering obtains cluster result between group.
With reference to first aspect, the embodiment of the invention provides the 6th kind of possible embodiments of first aspect, wherein institute
The temporal information and/or location information according to the facial image are stated, judges whether to carry out the facial image between phylogenetic group
Merge, obtain auxiliary judgment result, comprising:
According to the phase recency of image acquisition time, same collecting location the different acquisition time facial image, it is close when
Between personage there is at least one of distance location, same collecting location personage frequency of occurrences, judging whether will be between phylogenetic group
Facial image merge, obtain auxiliary judgment result.
With reference to first aspect, the embodiment of the invention provides the 7th kind of possible embodiments of first aspect, wherein institute
It states and cluster is merged to the multiple facial image according to the similarity, after obtaining agglomerative clustering result, further includes:
Corresponding prompt information of examining oneself is exported according to the agglomerative clustering result；
User is obtained according to the prompt information of examining oneself to the artificial annotation results returned after facial image mark；
The agglomerative clustering result is updated according to the artificial annotation results.
Second aspect, the embodiment of the present invention also provide a kind of archive management device for city safety monitoring, comprising:
Identification module obtains multiple feature vectors for carrying out recognition of face to multiple facial images；
Computing module, for calculating the vector distance between the multiple feature vector；
Determining module, for determining the similarity between the multiple facial image according to the vector distance；
Cluster module obtains merging poly- for merging cluster to the multiple facial image according to the similarity
Class result；
Module is established, for establishing several Profiles according to the agglomerative clustering result.
The third aspect, the embodiment of the present invention also provide a kind of electronic equipment, including memory, processor, the memory
In be stored with the computer program that can be run on the processor, the processor is realized when executing the computer program
The step of stating method as described in relation to the first aspect.
Fourth aspect, the embodiment of the present invention also provide a kind of meter of non-volatile program code that can be performed with processor
Calculation machine readable medium, said program code make the method for the processor execution as described in relation to the first aspect.
Technical solution provided in an embodiment of the present invention brings following the utility model has the advantages that provided in an embodiment of the present invention for city
Archive management method, device and the electronic equipment of city's safety monitoring, firstly, to multiple facial images carry out recognition of face to
Multiple feature vectors are obtained, then, calculate the vector distance between these feature vectors, later, are determined according to the vector distance
Then similarity between these facial images merges cluster to these facial images according to the similarity to obtain
Agglomerative clustering is as a result, finally, establish several Profiles according to the agglomerative clustering result, therefore, by by recognition of face, poly-
Class algorithm and personal filing combine, and the image of same people can be made accurately to be grouped into same for city safety monitoring
Profile in, that is, make the quantity of people more, by facial image carry out recognition of face, vector distance calculate, similarity
The analysis and excavation of the processes such as determination, agglomerative clustering are improved the accuracy to the Put on file of people, realize and be used for
The raising of the Profile the data precision of city safety monitoring, to solve the number of Profile existing in the prior art
It is lower to influence the technical issues of city safety monitoring works according to accuracy.
Other features and advantages of the present invention will illustrate in the following description, also, partly become from specification
It obtains it is clear that understand through the implementation of the invention.The objectives and other advantages of the invention are in specification and attached drawing
Specifically noted structure is achieved and obtained.
To enable the above objects, features and advantages of the present invention to be clearer and more comprehensible, preferred embodiment is cited below particularly, and cooperate
Appended attached drawing, is described in detail below.
Detailed description of the invention
It, below will be to specific in order to illustrate more clearly of the specific embodiment of the invention or technical solution in the prior art
Embodiment or attached drawing needed to be used in the description of the prior art be briefly described, it should be apparent that, it is described below
Attached drawing is some embodiments of the present invention, for those of ordinary skill in the art, before not making the creative labor
It puts, is also possible to obtain other drawings based on these drawings.
Fig. 1 shows the process provided by the embodiment of the present invention one for the archive management method of city safety monitoring
Figure；
Fig. 2 shows the processes for the archive management method that city safety monitoring is used for provided by the embodiment of the present invention two
Figure；
Fig. 3 shows a kind of knot of the archive management device for city safety monitoring provided by the embodiment of the present invention three
Structure schematic diagram；
Fig. 4 shows the structural schematic diagram of a kind of electronic equipment provided by the embodiment of the present invention four.
Icon: 3- archive management device；31- identification module；32- computing module；33- determining module；34- cluster module；
35- establishes module；4- electronic equipment；41- memory；42- processor；43- bus；44- communication interface.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with attached drawing to the present invention
Technical solution be clearly and completely described, it is clear that described embodiments are some of the embodiments of the present invention, rather than
Whole embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art are not making creative work premise
Under every other embodiment obtained, shall fall within the protection scope of the present invention.
Currently, a large amount of Profile for city safety monitoring establishes work difficulty in the case where number is more
It is larger, it is easy to make the image information etc. in the Profile of the same person with different people, to cause multiple Profiles
Between data corruption, cause the data precision of Profile lower, thus influence city safety monitoring work.It is based on
This, a kind of archive management method, device and electronic equipment for city safety monitoring provided in an embodiment of the present invention can be with
The lower technology to influence the work of city safety monitoring of the data precision for solving Profile existing in the prior art is asked
Topic.For convenient for understanding the present embodiment, first to a kind of archive management method, device disclosed in the embodiment of the present invention with
And electronic equipment describes in detail.
Embodiment one:
A kind of archive management method for city safety monitoring provided in an embodiment of the present invention, as a kind of supervision of the cities
The method that Profile is established, as shown in Figure 1, this method comprises:
S11: recognition of face is carried out to multiple facial images, obtains multiple feature vectors.
Wherein, which can be the video image of multiframe, or the picture of single frames.In this step, respectively
Recognition of face is carried out to multiple facial images, to respectively obtain multiple face feature vectors (i.e. feature vector).As this reality
The preferred embodiment of example is applied, facial image can be obtained from the character image in monitor video.
S12: the vector distance between multiple feature vectors is calculated.
In this step, the vector distance between multiple face feature vectors (i.e. feature vector) in step S11 is calculated.Its
In, vector distance may include Euclidean distance, manhatton distance, Chebyshev's distance, Minkowski Distance, mahalanobis distance,
At least one of COS distance, Hamming distance, Jie Kade distance.
S13: the similarity between multiple facial images is determined according to vector distance.
It is calculated as a preferred embodiment according to the vector distance between multiple face feature vectors (i.e. feature vector)
Similarity between multiple facial images out.Wherein, so-called similarity, refer to the more close then corresponding facial image of vector distance more
It is similar.
S14: cluster is merged to multiple facial images according to similarity, obtains agglomerative clustering result.
Specifically, according to similarity calculated in step S13, by clustering algorithm by these facial images, more
Similar (i.e. similarity is higher) facial image first carries out agglomerative clustering, to obtain agglomerative clustering result.For example, can incite somebody to action
The facial image that similarity is greater than preset value first carries out agglomerative clustering.
S15: several Profiles are established according to agglomerative clustering result.
It wherein, include: various documents and the electricity such as personally identifiable information, personal images information in Profile
Subdata.It is also understood that Profile is the repository for the personal information of city safety monitoring, the reflection that it summarizes
Everyone personal overall picture in city.
In practical applications, who is had according to the confirmation of agglomerative clustering result, come further according to the people of confirmation to these people
Profile established, to manage the Profile of a large amount of numbers.It should be noted that the Profile established is interior
Appearance may include: personal images information, personal appearance change information, personal regional information, individual sports trace information often occurs
Etc..
In the present embodiment, Profile can not only be established, existing Profile can also be updated.
For example, establishing the Profile of the people in file store if the Profile of the people is not present in archives；If in shelves
The Profile of the people is had existed in case, then the Profile of the people is carried out further perfect.
For the prior art, the foundation of the city safety monitoring Profile based on recognition of face is blank field,
Moreover, existing recognition of face is established in the processing extensive quantity face of City-level not can guarantee precision on archives, while big number
According to magnitude but also existing face recognition technology is difficult to meet.
Monitoring camera acquisition is utilized in the management that Profile is carried out by the data got according to monitoring camera
The video arrived clusters face by clustering algorithm by carrying out recognition of face to the people occurred in monitor video, from
And the people clustered out is filed or archives are established, such as the image of same people is grouped into the same Profile or root
The Profile of the people is established according to someone emerging image, to manage everyone Profile, so as to obtain somewhere
The proprietary Profile occurred in area.
Therefore, it by carrying out analysis mining to the collected data of monitoring camera in somewhere, can be taken the photograph based on monitoring
Take as head everyone form an independent Profile for city safety monitoring.Even so in people
The large number of area of number, is combined by the way that recognition of face, face cluster and individual file, can also guarantee this area
The file store for city safety monitoring accuracy, and can the significantly more efficient file administration for carrying out big quantity size,
To facilitate the raising of city safety monitoring working efficiency.
Embodiment two:
A kind of archive management method for city safety monitoring provided in an embodiment of the present invention, as shown in Figure 2, comprising:
S21: structuring parsing is carried out to monitor video, obtains multiple facial images.
As the preferred embodiment of the present embodiment, by carrying out structuring parsing to collected video image, thus
Obtain multiple facial images.
S22: recognition of face is carried out to multiple facial images, obtains multiple feature vectors.
In this step, firstly, calculating a feature vector for each face by recognition of face.Wherein, feature
The calculating process of vector can be carried out by human face recognition model, i.e., facial image, face are inputted into human face recognition model
Identification model will export the feature vector of the face.Specifically, by carrying out Face datection, face label etc. to facial image
The process of recognition of face, and then obtain the feature vector of face.
S23: the vector distance between multiple feature vectors is calculated.
Wherein, vector distance may include: Euclidean distance, manhatton distance, Chebyshev distance, Minkowski away from
From, at least one of with a distance from mahalanobis distance, COS distance, Hamming distance, Jie Kade.
It should be noted that Euclidean distance (Euclidean Distance) is between the most common two o'clock or between multiple spot
Apart from representation, also referred to as euclidean metric, it is defined in Euclidean space, be easiest to understand it is a kind of away from
From calculation method.For COS distance, alternatively referred to as included angle cosine, included angle cosine can be used to measure two vector directions in geometry
Difference, therefore, it is also possible to included angle cosine indicate vector between vector distance.Jie Kade distance (Jaccard Distance)
It is a kind of index for measuring two set difference opposite sex, it is the supplementary set of Jie Kade similarity factor, and Jie Kade similarity factor
(Jaccard similarity coefficient) is a kind of index for measuring two set similarities.
S24: the similarity between multiple facial images is determined according to vector distance.
In the present embodiment, similarity refers to that the more close then corresponding facial image of vector distance is more similar, therefore, can basis
Vector distance calculates the similarity between multiple facial images by similarity measurements quantity algorithm.Specifically, based on face characteristic to
The distance of any modes such as Euclidean distance, manhatton distance, Chebyshev's distance between amount, is calculated using similarity measurement
Method calculates the similarity between face.
It should be noted that needing to estimate between different samples (i.e. facial image) before classifying doing and (clustering)
Similarity measurement (Similarity Measurement), in the present embodiment, the method for use is to calculate the feature vector of sample
The distance between (i.e. vector distance), be then based on these distance (i.e. vector distance), carry out similarity measurement, to calculate
Similarity between sample (i.e. facial image).
Wherein, similarity measurement refers to a kind of measurement of close degree between two things (i.e. facial image) of Comprehensive Assessment.
The feature vector of two things is apart from closer, then the similarity measurement of the two things is also bigger, and the spy of two things
Sign vector distance is more become estranged, then the similarity measurement of the two things is also just smaller.Similitude may include: related coefficient (weighing apparatus
Degree of closeness between quantitative change amount), similarity factor (measure sample between degree of closeness) etc..
S25: greedy algorithm is based on according to similarity, cluster is merged to multiple facial images, obtained by multiple phylogenetic groups
The preliminary clusters result of composition.
It should be noted that greedy algorithm refers to: when to problem solving, always making and currently appearing to be best choosing
It selects.That is, not taking in from total optimization, what he was made is locally optimal solution in some sense.Specifically
, greedy algorithm is not that can obtain total optimization solution to all problems, it is important to the selection of Greedy strategy, the greedy plan of selection
Must slightly have markov property, i.e. the pervious process of some state will not influence later state, only related with current state.
As a preferred embodiment, in this step, by making most like preferential combination, similarity is low to be considered rearward
(i.e. similarity is higher) facial image more similar in multiple faces is first carried out preliminary agglomerative clustering by mode, thus
Obtain the preliminary clusters result being made of multiple phylogenetic groups.For example, in multiple facial images, based on the phase between face two-by-two
Like degree, the face that similarity is higher than preset value is first carried out by preliminary agglomerative clustering by greedy algorithm, to obtain preliminary clusters
As a result.
S26: merging cluster to the facial image between multiple phylogenetic groups, obtains cluster result between group, and will be between group
Cluster result is determined as agglomerative clustering result.
It can be that have passed through the iteration cluster process repeatedly merged, i.e., to multiple phylogenetic groups for the process of agglomerative clustering
Between facial image carry out successive ignition cluster, thus cluster result between group after repeatedly being merged.
If only being clustered (the i.e. preliminary clusters mistake based on similarity of step S22 to S25 according to image similarity
Journey), it is likely that of a sort people can be divided into different classes, lead to an identical people originally, but due to image similarity
Different several individuals not enough are divided into, to cause to cluster not accurate enough problem.And it is clustered between the group for passing through step S26
Process, the situation can be made further to be handled, i.e., will be divided into inhomogeneous same class people merge again it is poly-
Class (carries out clustering between group), and then solves this problem, so that the accuracy of agglomerative clustering result be made to be improved.
In this step, process is obtained for cluster result between group, it can specifically includes the following steps: firstly, preliminary
In each phylogenetic group in cluster result, several representative character images are chosen；Then, it calculates between every two phylogenetic group
The similarity of representative character image obtains several representative image similarities；Later, several representative image similarities are calculated
Average value, obtain similarity between group；Finally, being merged according to similarity between group to the facial image between phylogenetic group poly-
Class obtains cluster result between group.
It, can be with specifically, for the selection process of representative character image are as follows: in any phylogenetic group, calculated using electing
Method chooses several representative character images, i.e., elects algorithm by face, selecting in some phylogenetic group most has reference
The facial image (such as recent images, same position image, the close image of posture) of value.For example, in preliminary clusters result
Each phylogenetic group in, according at least one of picture quality, temporal information, location information, facial angle, personage's posture,
Using electing several representative character images of algorithm picks.
In practical applications, for electing algorithm, concrete implementation strategy may include: that quality is elected, that is, elect pledge
Measure representative of the best several pictures as face；Different dimensions are elected, i.e., when there are many historical data that someone accumulates,
Can different dimensions elect, such as time, place, facial angle etc. can find same time period (such as under being all in this way
Noon two o'clock is between 4 points), identical place, identical facial angle and identical posture etc. image, in this, as the generation of the people
Table image.
Later, similarity calculation just is carried out to these representative character images, the present embodiment is between certain two phylogenetic group
Group between be illustrated for similarity calculation process, if the two phylogenetic groups are respectively internal all to have 10 facial images, lead to
It crosses above-mentioned selection process and elects 3 representative character images in 10 facial images in each phylogenetic group, i.e., every group has 3
It is a to elect image (i.e. representative character image), then image is elected for 3 in every group, is carried out between the two phylogenetic groups
Then the similarity calculation of image two-by-two, is averaged to this 9 similarity values to obtain 9 similarity values of 3 × 3 matrixes
Algorithm, to obtain similarity between the group between the two phylogenetic groups.
Therefore, for each phylogenetic group, the image of preset quantity, such as 3 are generally therefrom elected, if in the phylogenetic group
Image opened less than 3, as a preferred embodiment, all images in the phylogenetic group are to elect image, that is, are both participated in similar
The calculating of degree.
As the another embodiment of the present embodiment, the facial image between phylogenetic group is carried out according to similarity between group
The process of agglomerative clustering may include: to judge whether to gather firstly, according to the temporal information and/or location information of facial image
Facial image between class group merges, and obtains auxiliary judgment result；Then, according to similarity between group, it is based on auxiliary judgment
As a result cluster is merged to the facial image between phylogenetic group, obtains cluster result between group.For example, big for similarity between group
In two phylogenetic groups of preset value, according to the temporal information of facial image and/or location information to determine whether the two are gathered
Facial image between class group merges, and sentences in this, as the auxiliary whether merged to facial image between phylogenetic group
It is disconnected.
Wherein, temporal information can be the time of acquisition facial image, and location information can be the ground of acquisition facial image
Point.Specifically, the acquisition process of auxiliary judgment result may include: according to the phase recency of image acquisition time, it is same locality
There is distance location, same collecting location personage appearance frequency in the facial image of different acquisition time, the personage of similar time in point
At least one of rate judges whether to merge the facial image between phylogenetic group, obtains auxiliary judgment result.For example, can
First to obtain temporal information and location information (the i.e. time of acquisition facial image and acquisition facial image of shooting character image
Place), it is then based on the temporal information and/or the location information, (Person Re- is identified by pedestrian again
Identification, abbreviation ReID) algorithm obtains auxiliary judgment as a result, whether will be between certain two phylogenetic group with auxiliary judgment
Facial image merges.
There is the judgment mode of distance location for the above-mentioned personage according to similar time, it should be noted that shorter
Time within, if the collecting location of two character images apart from far, such as spans provinces and cities, even if then the two personages
The human face similarity degree of image is high again, also judges that the two character images are not the same persons, in this, as being auxiliary judgment as a result,
To judge whether to merge facial image between two phylogenetic groups.
For the judgment mode of the above-mentioned phase recency according to image acquisition time, specifically, by character image on the same day
High priority data is integrated into together, because fat or thin etc. a possibility that changing of the hair style of people on the same day, people, is very low,
A possibility that dressing changes is relatively low, and this characteristic is applicable on ReID algorithm.Therefore, same stature hair
Type clothing people ReID distance be it is close, can merge；And the different people ReID such as stature, hair style, clothing away from
From be it is farther away, just without merge, so as to carry out auxiliary judgment by ReID.In this way, cluster result between obtained group
The similarity of the face considered not only, it is also contemplated that many factors such as the stature of people, hair style, clothing.
In addition, being needed for the judgment mode of the above-mentioned facial image according to same collecting location in the different acquisition time
Illustrate, since the camera imaging feature of different location has difference, by the imaging of the history in the same place come with it is new
Imaging compares, so as to pass through the error of agglomerative clustering process between supplementary mode reduction group.
For the above-mentioned judgment mode according to the same collecting location personage frequency of occurrences, it should be noted that due to people's
Activity is usually to compare concentration, can be by history on this ground when sufficiently long to Profile integration time
The frequency that point the people occurs is as auxiliary reference, to judge whether to agglomerative clustering between group.
Finally, cluster result is determined as agglomerative clustering result between the group that these are obtained by step S21 to step S26.
S27: corresponding prompt information of examining oneself is exported according to agglomerative clustering result.
According to the obtained agglomerative clustering of step S26 as a result, prompt letter of examining oneself corresponding to output and the agglomerative clustering result
Breath.Wherein, prompt information of examining oneself is for characterizing the problem of cluster result is likely to occur, such as error message either exception information
Deng.Certainly, prompt information of examining oneself can also all show that is, each cluster can examine oneself one after each cluster process
The similar prompt information of " it may be poly- wrong class that I, which has much, ", all prompt informations of examining oneself are summarized, sorted and exported,
For example, the image of the poly- wrong class of most probable can be pushed to artificial mark interface, the prompt information that can also will examine oneself is sent to use
Family end, in order to which user (i.e. mark personnel) is labeled judgement.
S28: user is obtained according to prompt information of examining oneself to the artificial annotation results returned after facial image mark.
User can check Profile belonging to each image, can also look at individual when carrying out judgement mark
The historical information of archives, the certificate photo of Profile, the shooting time of character image and shooting location etc. are passed through artificial with this
Whether the judgement of various dimensions information correctly judges to make clustering.
In addition, can also have oneself of " certain two class may be a people " in addition to whether other than the information of examining oneself of poly- wrong class
It saves, then needs to mark personnel and judge whether the character image inside two classes is the same person.Mark personnel can be according to personage
The hair style of people, clothing, stature etc. judge whether it is same people in image, because the hair style of short time (such as one day) interior people, people
A possibility that fat or thin etc. a possibility that changing is very low, and dressing changes is relatively low, and this characteristic energy
It is enough applied in manual intervention, can judge whether it is same people with indirect labor.
Preferably, user can be labeled facial image according to prompt information of examining oneself on mark interface, wherein right
In mark interface, interface can be provided in order to artificial carry out data mark by referring to, the process of data mark can be in no prison
In the case where superintending and directing or having a small amount of mark, human face data mark speed is greatly speeded up, data is reduced and marks cost.
Then, the judging result for marking personnel can be carried out with the format of (img_a, img_b, if_is_same_person)
Save, clustering algorithm can absorb then adjustment that such artificial markup information carries out archives, merging including two archives, certain
Some character images of one archives are removed out.
S29: it is updated according to artificial annotation results to cluster result is merged.
Therefore, examining oneself for algorithm is carried out and having a cluster of human intervention, such as " most probable, which gathers wrong photo, in such is
Which is opened ", " which two photo is most unlikely a same person in such ", " such and which other class are most likely same
People " etc. examines oneself information, mode of examining oneself in this way, when can elect the candid photograph associated therewith of picture that algorithm can not determine
Between, capture the information such as date, shown on manually mark interface, and personnel to be marked is waited to carry out naked eyes judgement, making " is
The selection of same people " or " not being same people ".Then, based on mark personnel feed back whether be same people selection letter
It ceases (i.e. artificial annotation results), can concentrate and be diffused in character data, for example, A1 and A are same if A and B are same people
One people, B1 and B are same people, then increase A1 and B1 is the probability of same people, be updated with this to cluster result is merged.
S30: several Profiles are established according to agglomerative clustering result.
Specifically, who object first can be determined according to step S29 updated agglomerative clustering result, then, according to true
The who object made establishes Profile.Therefore, the people each identified can accurately correspond to the use of its own
In the Profile of city safety monitoring.
Further, in Profile in addition to may include the people character image etc., can also include the people row
For track, high frequency region etc..Specifically, firstly, according to collect someone image location information and temporal information (i.e. certain
People is monitored the position occurred when camera is captured to photo and time), can analyze the people when arrived where
Information, the action trail of the people can be depicted with this；The height of the people can also be obtained according to the action trail of the people later
Frequency region etc.；Then, the action trail of the people, high frequency region etc. just can be stored into the Profile of the people
In.It therefore, can be to occur in somewhere by making a people that there is a corresponding Profile in security protection scene setting
Cross everyone create personal archives, the behavior record of the people is established by image data, and can realize trajectory track,
The application such as zone of action thermodynamic chart.
In the present embodiment, by the way that clustering algorithm, ReID algorithm etc. are optimized and combined, make polyalgorithm can highly simultaneously
Rowization can not only be such that the precision for calculating identification is improved with accuracy, additionally it is possible to reduce computing cost and memory overhead, make
Can realize under occupation condition low as far as possible to the cluster and filing of greater number character image, to improve city
The efficiency of city's safety monitoring work.
Embodiment three:
A kind of archive management device for city safety monitoring provided in an embodiment of the present invention, as shown in figure 3, archives pipe
Reason device 3 includes: identification module 31, computing module 32, determining module 33, cluster module 34 and establishes module 35.
Preferably, identification module is used to carry out recognition of face to multiple facial images, obtains multiple feature vectors.Calculate mould
Block is used to calculate the vector distance between multiple feature vectors.Determining module is used to determine multiple facial images according to vector distance
Between similarity.
As the preferred embodiment of the present embodiment, cluster module is for closing multiple facial images according to similarity
And cluster, obtain agglomerative clustering result.Module is established for establishing several Profiles according to agglomerative clustering result.
Archive management device provided in an embodiment of the present invention for city safety monitoring, with use provided by the above embodiment
It is reached in the archive management method of city safety monitoring technical characteristic having the same so also can solve identical technical problem
To identical technical effect.
Example IV:
A kind of electronic equipment provided in an embodiment of the present invention, as shown in figure 4, electronic equipment 4 includes memory 41, processor
42, the computer program that can be run on the processor is stored in the memory, the processor executes the calculating
The step of method that above-described embodiment one or embodiment two provide is realized when machine program.
Referring to fig. 4, electronic equipment further include: bus 43 and communication interface 44, processor 42, communication interface 44 and memory
41 are connected by bus 43；Processor 42 is for executing the executable module stored in memory 41, such as computer program.
Wherein, memory 41 may include high-speed random access memory (RAM, Random Access Memory),
It may further include nonvolatile memory (non-volatile memory), for example, at least a magnetic disk storage.By at least
One communication interface 44 (can be wired or wireless) realizes the communication between the system network element and at least one other network element
Connection, can be used internet, wide area network, local network, Metropolitan Area Network (MAN) etc..
Bus 43 can be isa bus, pci bus or eisa bus etc..The bus can be divided into address bus, data
Bus, control bus etc..Only to be indicated with a four-headed arrow convenient for indicating, in Fig. 4, it is not intended that an only bus or
A type of bus.
Wherein, memory 41 is for storing program, and the processor 42 executes the journey after receiving and executing instruction
Sequence, method performed by the device that the stream process that aforementioned any embodiment of the embodiment of the present invention discloses defines can be applied to handle
In device 42, or realized by processor 42.
Processor 42 may be a kind of IC chip, the processing capacity with signal.During realization, above-mentioned side
Each step of method can be completed by the integrated logic circuit of the hardware in processor 42 or the instruction of software form.Above-mentioned
Processor 42 can be general processor, including central processing unit (Central Processing Unit, abbreviation CPU), network
Processor (Network Processor, abbreviation NP) etc.；It can also be digital signal processor (Digital Signal
Processing, abbreviation DSP), specific integrated circuit (Application Specific Integrated Circuit, referred to as
ASIC), ready-made programmable gate array (Field-Programmable Gate Array, abbreviation FPGA) or other are programmable
Logical device, discrete gate or transistor logic, discrete hardware components.It may be implemented or execute in the embodiment of the present invention
Disclosed each method, step and logic diagram.General processor can be microprocessor or the processor is also possible to appoint
What conventional processor etc..The step of method in conjunction with disclosed in the embodiment of the present invention, can be embodied directly in hardware decoding processing
Device executes completion, or in decoding processor hardware and software module combination execute completion.Software module can be located at
Machine memory, flash memory, read-only memory, programmable read only memory or electrically erasable programmable memory, register etc. are originally
In the storage medium of field maturation.The storage medium is located at memory 41, and processor 42 reads the information in memory 41, in conjunction with
Its hardware completes the step of above method.
Embodiment five:
It is provided in an embodiment of the present invention it is a kind of with processor can be performed non-volatile program code it is computer-readable
Medium, said program code make the method that the processor executes above-described embodiment one or embodiment two provides.
Unless specifically stated otherwise, the opposite step of the component and step that otherwise illustrate in these embodiments, digital table
It is not limit the scope of the invention up to formula and numerical value.
It is apparent to those skilled in the art that for convenience and simplicity of description, the system of foregoing description
It with the specific work process of device, can refer to corresponding processes in the foregoing method embodiment, details are not described herein.
In all examples being illustrated and described herein, any occurrence should be construed as merely illustratively, without
It is as limitation, therefore, other examples of exemplary embodiment can have different values.
It should also be noted that similar label and letter indicate similar terms in following attached drawing, therefore, once a certain Xiang Yi
It is defined in a attached drawing, does not then need that it is further defined and explained in subsequent attached drawing.
The flow chart and block diagram in the drawings show the system of multiple embodiments according to the present invention, method and computer journeys
The architecture, function and operation in the cards of sequence product.In this regard, each box in flowchart or block diagram can generation
A part of one module, section or code of table, a part of the module, section or code include one or more use
The executable instruction of the logic function as defined in realizing.It should also be noted that in some implementations as replacements, being marked in box
The function of note can also occur in a different order than that indicated in the drawings.For example, two continuous boxes can actually base
Originally it is performed in parallel, they can also be executed in the opposite order sometimes, and this depends on the function involved.It is also noted that
It is the combination of each box in block diagram and or flow chart and the box in block diagram and or flow chart, can uses and execute rule
The dedicated hardware based system of fixed function or movement is realized, or can use the group of specialized hardware and computer instruction
It closes to realize.
The computer-readable medium of the non-volatile program code provided in an embodiment of the present invention that can be performed with processor,
With provided by the above embodiment for the archive management method of city safety monitoring, device and electronic equipment skill having the same
Art feature reaches identical technical effect so also can solve identical technical problem.
It carries out producing for the computer program of the archive management method of city safety monitoring provided by the embodiment of the present invention
Product, the computer readable storage medium including storing the executable non-volatile program code of processor, said program code
Including instruction can be used for executing previous methods method as described in the examples, specific implementation can be found in embodiment of the method, herein
It repeats no more.
In several embodiments provided herein, it should be understood that disclosed systems, devices and methods, it can be with
It realizes by another way.The apparatus embodiments described above are merely exemplary, for example, the division of the unit,
Only a kind of logical function partition, there may be another division manner in actual implementation, in another example, multiple units or components can
To combine or be desirably integrated into another system, or some features can be ignored or not executed.Another point, it is shown or beg for
The mutual coupling, direct-coupling or communication connection of opinion can be through some communication interfaces, device or unit it is indirect
Coupling or communication connection can be electrical property, mechanical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.
It, can be with if the function is realized in the form of SFU software functional unit and when sold or used as an independent product
It is stored in a computer readable storage medium.Based on this understanding, technical solution of the present invention is substantially in other words
The part of the part that contributes to existing technology or the technical solution can be embodied in the form of software products, the meter
Calculation machine software product is stored in a storage medium, including some instructions are used so that a computer equipment (can be a
People's computer, server or network equipment etc.) it performs all or part of the steps of the method described in the various embodiments of the present invention.
And storage medium above-mentioned includes: that USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), arbitrary access are deposited
The various media that can store program code such as reservoir (RAM, Random Access Memory), magnetic or disk.
Finally, it should be noted that embodiment described above, only a specific embodiment of the invention, to illustrate the present invention
Technical solution, rather than its limitations, scope of protection of the present invention is not limited thereto, although with reference to the foregoing embodiments to this hair
It is bright to be described in detail, those skilled in the art should understand that: anyone skilled in the art
In the technical scope disclosed by the present invention, it can still modify to technical solution documented by previous embodiment or can be light
It is readily conceivable that variation or equivalent replacement of some of the technical features；And these modifications, variation or replacement, do not make
The essence of corresponding technical solution is detached from the spirit and scope of technical solution of the embodiment of the present invention, should all cover in protection of the invention
Within the scope of.Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (11)
1. a kind of archive management method for city safety monitoring characterized by comprising
Recognition of face is carried out to multiple facial images, obtains multiple feature vectors；
Calculate the vector distance between the multiple feature vector；
The similarity between the multiple facial image is determined according to the vector distance；
Cluster is merged to the multiple facial image according to the similarity, obtains agglomerative clustering result；
Several Profiles are established according to the agglomerative clustering result.
2. archive management method according to claim 1, which is characterized in that described to carry out face knowledge to multiple facial images
Not, before obtaining multiple feature vectors, further includes:
Structuring parsing is carried out to monitor video, obtains multiple facial images.
3. archive management method according to claim 1, which is characterized in that it is described according to the similarity to the multiple
Facial image merges cluster, obtains agglomerative clustering result, comprising:
Greedy algorithm is based on according to the similarity, cluster is merged to the multiple facial image, obtain by multiple phylogenetic groups
The preliminary clusters result of composition；
Cluster is merged to the facial image between the multiple phylogenetic group, obtains cluster result between group, and will be between described group
Cluster result is determined as agglomerative clustering result.
4. archive management method according to claim 3, which is characterized in that the people between the multiple phylogenetic group
Face image merges cluster, obtains cluster result between group, comprising:
In each phylogenetic group in the preliminary clusters result, several representative character images are chosen；
The similarity for calculating the representative character image between every two phylogenetic group, it is similar to obtain several representative images
Degree；
The average value for calculating several representative image similarities, obtains similarity between group；
Cluster is merged to the facial image between phylogenetic group according to similarity between described group, obtains cluster result between group.
5. archive management method according to claim 4, which is characterized in that described every in the preliminary clusters result
In one phylogenetic group, several representative character images are chosen, comprising:
In each phylogenetic group in the preliminary clusters result, according to picture quality, temporal information, location information, face angle
Degree, at least one of personage's posture, using electing several representative character images of algorithm picks.
6. archive management method according to claim 4, which is characterized in that it is described according to similarity between described group to cluster
Facial image between group merges cluster, obtains cluster result between group, comprising:
According to the temporal information and/or location information of the facial image, judge whether by the facial image between phylogenetic group into
Row merges, and obtains auxiliary judgment result；
According to similarity between described group, the facial image between the phylogenetic group is merged based on the auxiliary judgment result
Cluster, obtains cluster result between group.
7. archive management method according to claim 6, which is characterized in that described to be believed according to the time of the facial image
Breath and/or location information, judge whether to merge the facial image between phylogenetic group, obtain auxiliary judgment result, comprising:
According to the phase recency of image acquisition time, same collecting location in the facial image of different acquisition time, similar time
There is at least one of distance location, same collecting location personage frequency of occurrences in personage, judges whether the people between phylogenetic group
Face image merges, and obtains auxiliary judgment result.
8. archive management method according to claim 1, which is characterized in that it is described according to the similarity to the multiple
Facial image merges cluster, after obtaining agglomerative clustering result, further includes:
Corresponding prompt information of examining oneself is exported according to the agglomerative clustering result；
User is obtained according to the prompt information of examining oneself to the artificial annotation results returned after facial image mark；
The agglomerative clustering result is updated according to the artificial annotation results.
9. a kind of archive management device for city safety monitoring characterized by comprising
Identification module obtains multiple feature vectors for carrying out recognition of face to multiple facial images；
Computing module, for calculating the vector distance between the multiple feature vector；
Determining module, for determining the similarity between the multiple facial image according to the vector distance；
Cluster module obtains agglomerative clustering knot for merging cluster to the multiple facial image according to the similarity
Fruit；
Module is established, for establishing several Profiles according to the agglomerative clustering result.
10. a kind of electronic equipment, including memory, processor, it is stored with and can runs on the processor in the memory
Computer program, which is characterized in that the processor realizes the claims 1 to 8 when executing the computer program
The step of method described in one.
11. a kind of computer-readable medium for the non-volatile program code that can be performed with processor, which is characterized in that described
Program code makes the processor execute described any the method for claim 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810999452.3A CN109145844A (en) | 2018-08-29 | 2018-08-29 | Archive management method, device and electronic equipment for city safety monitoring |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810999452.3A CN109145844A (en) | 2018-08-29 | 2018-08-29 | Archive management method, device and electronic equipment for city safety monitoring |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109145844A true CN109145844A (en) | 2019-01-04 |
Family
ID=64829126
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810999452.3A Pending CN109145844A (en) | 2018-08-29 | 2018-08-29 | Archive management method, device and electronic equipment for city safety monitoring |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109145844A (en) |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110059657A (en) * | 2019-04-25 | 2019-07-26 | 北京旷视科技有限公司 | Records handling method, apparatus, electronic equipment and computer readable storage medium |
| CN110163135A (en) * | 2019-05-10 | 2019-08-23 | 杭州商警云智能科技有限公司 | A kind of method and system of one grade of face cluster of people based on dynamic algorithm |
| CN110390031A (en) * | 2019-06-28 | 2019-10-29 | 深圳市商汤科技有限公司 | Information processing method and device, vision facilities and storage medium |
| CN110413811A (en) * | 2019-08-05 | 2019-11-05 | 北京深醒科技有限公司 | A kind of across channel polymerization in passerby library based on recognition of face |
| CN110750670A (en) * | 2019-09-05 | 2020-02-04 | 北京旷视科技有限公司 | Stranger monitoring method, device and system and storage medium |
| CN110929679A (en) * | 2019-12-05 | 2020-03-27 | 杭州电子科技大学 | Non-supervision self-adaptive pedestrian re-identification method based on GAN |
| CN111026920A (en) * | 2019-12-17 | 2020-04-17 | 深圳云天励飞技术有限公司 | File merging method and device, electronic equipment and storage medium |
| CN111061899A (en) * | 2019-12-18 | 2020-04-24 | 深圳云天励飞技术有限公司 | Archive representative picture generation method and device and electronic equipment |
| CN111062294A (en) * | 2019-12-10 | 2020-04-24 | 北京文安智能技术股份有限公司 | Method, device and system for detecting passenger flow queuing time |
| CN111242040A (en) * | 2020-01-15 | 2020-06-05 | 佳都新太科技股份有限公司 | Dynamic face clustering method, device, equipment and storage medium |
| CN111310834A (en) * | 2020-02-19 | 2020-06-19 | 深圳市商汤科技有限公司 | Data processing method and device, processor, electronic equipment and storage medium |
| WO2020151507A1 (en) * | 2019-01-25 | 2020-07-30 | 华为技术有限公司 | File merging method and apparatus |
| CN112084812A (en) * | 2019-06-12 | 2020-12-15 | 腾讯科技（深圳）有限公司 | Image processing method, image processing device, computer equipment and storage medium |
| CN112131419A (en) * | 2020-08-17 | 2020-12-25 | 浙江大华技术股份有限公司 | Image archive merging method and device, electronic equipment and storage medium |
| CN112487221A (en) * | 2020-11-04 | 2021-03-12 | 北京澎思科技有限公司 | Image filing method and device, electronic device and storage medium |
| CN112651321A (en) * | 2020-12-21 | 2021-04-13 | 浙江商汤科技开发有限公司 | File processing method and device and server |
| CN112699328A (en) * | 2020-12-28 | 2021-04-23 | 建信金融科技有限责任公司 | Network point service data processing method, device, system, equipment and storage medium |
| CN112989887A (en) * | 2019-12-16 | 2021-06-18 | 深圳云天励飞技术有限公司 | File merging method and device and electronic equipment |
| CN113052079A (en) * | 2021-03-26 | 2021-06-29 | 重庆紫光华山智安科技有限公司 | Regional passenger flow statistical method, system, equipment and medium based on face clustering |
| CN113515597A (en) * | 2021-06-21 | 2021-10-19 | 中盾创新档案管理(北京)有限公司 | File processing method based on association rule mining |
| CN113673550A (en) * | 2021-06-30 | 2021-11-19 | 浙江大华技术股份有限公司 | Clustering method, clustering device, electronic equipment and computer-readable storage medium |
| CN114332940A (en) * | 2021-12-30 | 2022-04-12 | 北京爱奇艺科技有限公司 | Model training method, clothing recognition processing method, related device and terminal |
| CN114898420A (en) * | 2022-03-29 | 2022-08-12 | 浙江大华技术股份有限公司 | Abnormal face archive identification method and device, electronic device and storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130142402A1 (en) * | 2005-09-28 | 2013-06-06 | Facedouble, Inc. | Method And System For Attaching A Metatag To A Digital Image |
| CN104252616A (en) * | 2013-06-28 | 2014-12-31 | 广州华多网络科技有限公司 | Human face marking method, device and equipment |
| CN104391879A (en) * | 2014-10-31 | 2015-03-04 | 小米科技有限责任公司 | Method and device for hierarchical clustering |
| CN104408402A (en) * | 2014-10-29 | 2015-03-11 | 小米科技有限责任公司 | Face identification method and apparatus |
| CN107392222A (en) * | 2017-06-07 | 2017-11-24 | 深圳市深网视界科技有限公司 | A kind of face cluster method, apparatus and storage medium |
| CN107609466A (en) * | 2017-07-26 | 2018-01-19 | 百度在线网络技术（北京）有限公司 | Face cluster method, apparatus, equipment and storage medium |
| CN108229419A (en) * | 2018-01-22 | 2018-06-29 | 百度在线网络技术（北京）有限公司 | For clustering the method and apparatus of image |
-
2018
- 2018-08-29 CN CN201810999452.3A patent/CN109145844A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130142402A1 (en) * | 2005-09-28 | 2013-06-06 | Facedouble, Inc. | Method And System For Attaching A Metatag To A Digital Image |
| CN104252616A (en) * | 2013-06-28 | 2014-12-31 | 广州华多网络科技有限公司 | Human face marking method, device and equipment |
| CN104408402A (en) * | 2014-10-29 | 2015-03-11 | 小米科技有限责任公司 | Face identification method and apparatus |
| CN104391879A (en) * | 2014-10-31 | 2015-03-04 | 小米科技有限责任公司 | Method and device for hierarchical clustering |
| CN107392222A (en) * | 2017-06-07 | 2017-11-24 | 深圳市深网视界科技有限公司 | A kind of face cluster method, apparatus and storage medium |
| CN107609466A (en) * | 2017-07-26 | 2018-01-19 | 百度在线网络技术（北京）有限公司 | Face cluster method, apparatus, equipment and storage medium |
| CN108229419A (en) * | 2018-01-22 | 2018-06-29 | 百度在线网络技术（北京）有限公司 | For clustering the method and apparatus of image |
Cited By (32)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020151507A1 (en) * | 2019-01-25 | 2020-07-30 | 华为技术有限公司 | File merging method and apparatus |
| CN111488894A (en) * | 2019-01-25 | 2020-08-04 | 华为技术有限公司 | File merging method and device |
| CN110059657A (en) * | 2019-04-25 | 2019-07-26 | 北京旷视科技有限公司 | Records handling method, apparatus, electronic equipment and computer readable storage medium |
| CN110163135A (en) * | 2019-05-10 | 2019-08-23 | 杭州商警云智能科技有限公司 | A kind of method and system of one grade of face cluster of people based on dynamic algorithm |
| CN112084812B (en) * | 2019-06-12 | 2023-08-01 | 腾讯科技（深圳）有限公司 | Image processing method, device, computer equipment and storage medium |
| CN112084812A (en) * | 2019-06-12 | 2020-12-15 | 腾讯科技（深圳）有限公司 | Image processing method, image processing device, computer equipment and storage medium |
| CN110390031A (en) * | 2019-06-28 | 2019-10-29 | 深圳市商汤科技有限公司 | Information processing method and device, vision facilities and storage medium |
| CN110413811A (en) * | 2019-08-05 | 2019-11-05 | 北京深醒科技有限公司 | A kind of across channel polymerization in passerby library based on recognition of face |
| CN110750670A (en) * | 2019-09-05 | 2020-02-04 | 北京旷视科技有限公司 | Stranger monitoring method, device and system and storage medium |
| CN110750670B (en) * | 2019-09-05 | 2022-04-19 | 北京旷视科技有限公司 | Stranger monitoring method, device and system and storage medium |
| CN110929679A (en) * | 2019-12-05 | 2020-03-27 | 杭州电子科技大学 | Non-supervision self-adaptive pedestrian re-identification method based on GAN |
| CN111062294A (en) * | 2019-12-10 | 2020-04-24 | 北京文安智能技术股份有限公司 | Method, device and system for detecting passenger flow queuing time |
| CN111062294B (en) * | 2019-12-10 | 2024-03-22 | 北京文安智能技术股份有限公司 | Passenger flow queuing time detection method, device and system |
| CN112989887A (en) * | 2019-12-16 | 2021-06-18 | 深圳云天励飞技术有限公司 | File merging method and device and electronic equipment |
| CN112989887B (en) * | 2019-12-16 | 2024-06-11 | 深圳云天励飞技术有限公司 | File merging method and device and electronic equipment |
| CN111026920A (en) * | 2019-12-17 | 2020-04-17 | 深圳云天励飞技术有限公司 | File merging method and device, electronic equipment and storage medium |
| CN111061899A (en) * | 2019-12-18 | 2020-04-24 | 深圳云天励飞技术有限公司 | Archive representative picture generation method and device and electronic equipment |
| WO2021143237A1 (en) * | 2020-01-15 | 2021-07-22 | 佳都新太科技股份有限公司 | Dynamic human face clustering method and apparatus, device, and storage medium |
| CN111242040A (en) * | 2020-01-15 | 2020-06-05 | 佳都新太科技股份有限公司 | Dynamic face clustering method, device, equipment and storage medium |
| CN111242040B (en) * | 2020-01-15 | 2022-08-02 | 佳都科技集团股份有限公司 | Dynamic face clustering method, device, equipment and storage medium |
| CN111310834B (en) * | 2020-02-19 | 2024-05-28 | 深圳市商汤科技有限公司 | Data processing method and device, processor, electronic equipment and storage medium |
| CN111310834A (en) * | 2020-02-19 | 2020-06-19 | 深圳市商汤科技有限公司 | Data processing method and device, processor, electronic equipment and storage medium |
| CN112131419A (en) * | 2020-08-17 | 2020-12-25 | 浙江大华技术股份有限公司 | Image archive merging method and device, electronic equipment and storage medium |
| CN112487221A (en) * | 2020-11-04 | 2021-03-12 | 北京澎思科技有限公司 | Image filing method and device, electronic device and storage medium |
| CN112651321A (en) * | 2020-12-21 | 2021-04-13 | 浙江商汤科技开发有限公司 | File processing method and device and server |
| CN112699328A (en) * | 2020-12-28 | 2021-04-23 | 建信金融科技有限责任公司 | Network point service data processing method, device, system, equipment and storage medium |
| CN113052079A (en) * | 2021-03-26 | 2021-06-29 | 重庆紫光华山智安科技有限公司 | Regional passenger flow statistical method, system, equipment and medium based on face clustering |
| CN113515597A (en) * | 2021-06-21 | 2021-10-19 | 中盾创新档案管理(北京)有限公司 | File processing method based on association rule mining |
| CN113673550A (en) * | 2021-06-30 | 2021-11-19 | 浙江大华技术股份有限公司 | Clustering method, clustering device, electronic equipment and computer-readable storage medium |
| CN114332940A (en) * | 2021-12-30 | 2022-04-12 | 北京爱奇艺科技有限公司 | Model training method, clothing recognition processing method, related device and terminal |
| CN114332940B (en) * | 2021-12-30 | 2024-06-07 | 北京爱奇艺科技有限公司 | Model training method, clothing recognition processing method, related device and terminal |
| CN114898420A (en) * | 2022-03-29 | 2022-08-12 | 浙江大华技术股份有限公司 | Abnormal face archive identification method and device, electronic device and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109145844A (en) | Archive management method, device and electronic equipment for city safety monitoring | |
| WO2021217934A1 (en) | Method and apparatus for monitoring number of livestock, and computer device and storage medium | |
| CN109344271A (en) | Video portrait records handling method and its system | |
| CN102187336B (en) | Method for identifying an object in a video archive | |
| CN109086739A (en) | A kind of face identification method and system of no human face data training | |
| CN109614882A (en) | A kind of act of violence detection system and method based on human body attitude estimation | |
| Ksibi et al. | Overview of lifelogging: current challenges and advances | |
| CN102170563A (en) | Intelligent person capture system and person monitoring management method | |
| CN101778260A (en) | Method and system for monitoring and managing videos on basis of structured description | |
| WO2012013706A1 (en) | Facilitating people search in video surveillance | |
| CN111327949B (en) | Video time sequence action detection method, device, equipment and storage medium | |
| Han et al. | Personalized video summarization with human in the loop | |
| Marcos-Ramiro et al. | Body communicative cue extraction for conversational analysis | |
| CN108986137A (en) | Human body tracing method, device and equipment | |
| CN110569918A (en) | sample classification method and related device | |
| CN113762106A (en) | Face recognition method and device, electronic equipment and storage medium | |
| CN107168536A (en) | Test question searching method, test question searching device and electronic terminal | |
| CN113538513A (en) | Method, device and equipment for controlling access of monitored object and storage medium | |
| CN202121706U (en) | Intelligent personnel monitoring system | |
| CN111767880B (en) | Living body identity recognition method and device based on facial features and storage medium | |
| CN116959099A (en) | Abnormal behavior identification method based on space-time diagram convolutional neural network | |
| CN106960183A (en) | A kind of image pedestrian's detection algorithm that decision tree is lifted based on gradient | |
| CN115374843A (en) | Internet of things intrusion detection model training method, device, equipment and storage medium | |
| CN111582031B (en) | Multi-model collaborative violence detection method and system based on neural network | |
| JP4232388B2 (en) | Video search apparatus, video search method used therefor, and program thereof |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190104 |