CN109409962A - Image processing method, device, electronic equipment, computer readable storage medium - Google Patents
Image processing method, device, electronic equipment, computer readable storage medium Download PDFInfo
- Publication number
- CN109409962A CN109409962A CN201811328072.3A CN201811328072A CN109409962A CN 109409962 A CN109409962 A CN 109409962A CN 201811328072 A CN201811328072 A CN 201811328072A CN 109409962 A CN109409962 A CN 109409962A
- Authority
- CN
- China
- Prior art keywords
- face
- customer
- image
- similarity
- score
- 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
- 238000003672 processing method Methods 0.000 title claims abstract description 19
- 230000001815 facial effect Effects 0.000 claims abstract description 102
- 238000000034 method Methods 0.000 claims abstract description 41
- 238000004590 computer program Methods 0.000 claims description 19
- 238000012360 testing method Methods 0.000 claims description 15
- 238000012545 processing Methods 0.000 claims description 12
- 238000001514 detection method Methods 0.000 claims description 8
- 239000000284 extract Substances 0.000 claims description 7
- 238000012512 characterization method Methods 0.000 claims description 4
- 238000010586 diagram Methods 0.000 description 12
- 238000004364 calculation method Methods 0.000 description 11
- 230000006870 function Effects 0.000 description 10
- 210000003128 head Anatomy 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 239000013598 vector Substances 0.000 description 5
- 238000000605 extraction Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 238000013527 convolutional neural network Methods 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 210000004709 eyebrow Anatomy 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 210000005252 bulbus oculi Anatomy 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 210000004209 hair Anatomy 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000009434 installation Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000000429 assembly Methods 0.000 description 1
- 230000000712 assembly Effects 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 239000011449 brick Substances 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005538 encapsulation Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 210000001508 eye Anatomy 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000004570 mortar (masonry) Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0281—Customer communication at a business location, e.g. providing product or service information, consulting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Finance (AREA)
- General Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Multimedia (AREA)
- Human Computer Interaction (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Health & Medical Sciences (AREA)
- Development Economics (AREA)
- Accounting & Taxation (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Economics (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of image processing method, device, electronic equipment, computer readable storage mediums, this method comprises: obtaining the facial image of customer to be identified, the similarity between the facial image and the custom image pre-saved is calculated, at least one similarity is obtained；Then judge the similarity threshold whether at least one described similarity all not up to pre-saves；When to be, determine that the customer is and to save the facial image as new custom image for the first time to shop customer；Otherwise determine that the customer is to repeat to shop customer.In this way, the customer that can be automatically identified to shop is new customer or repeats to shop customer, hotel owner or sales force do not have to the memory further according to itself to identify to customer.
Description
Technical field
The present invention relates to field of image processings, in particular to a kind of image processing method, device, electronic equipment, meter
Calculation machine readable storage medium storing program for executing.
Background technique
As new retail and the generation and popularization of intelligence retail concept are increasingly intended to user for solid shop/brick and mortar store
Bring better user experience, for example, hotel owner or sales force wish to know that shop customer be for the first time to shop new customer still
The patron for repeatedly arriving shop, targetedly services consequently facilitating providing for customer.
But since hotel owner or sales force's manpower and memory are limited, be difficult to it is all repeatedly to customer family into
Row effectively memory, therefore, so that hotel owner or sales force know that customer's is more difficult to shop number.
Summary of the invention
In view of this, the embodiment of the present invention is designed to provide a kind of image processing method, device, electronic equipment, meter
Calculation machine readable storage medium storing program for executing, to alleviate the above problem.
In a first aspect, the embodiment of the invention provides a kind of image processing methods, which comprises obtain Gu to be identified
The facial image of visitor；The similarity between the facial image and the custom image pre-saved is calculated, at least one phase is obtained
Like degree；Judge the similarity threshold whether all not up to pre-saved at least one described similarity；When to be, institute is determined
Stating customer is and to save the facial image as new custom image for the first time to shop customer；When to be no, institute is determined
Stating customer is to repeat to shop customer.
A kind of embodiment with reference to first aspect, the method also includes: extract the age characteristics of the facial image；
When judging that the age characteristics characterization customer is in preset age level, the first threshold pre-saved is determined as
The similarity threshold；Otherwise, the second threshold pre-saved is determined as the similarity threshold；Wherein, first threshold
Value is greater than the second threshold.
A kind of embodiment with reference to first aspect, the method also includes: obtain the test including multiple samples pictures
Collection；Test accuracy rate of the test set under different threshold values；Using the threshold value under obtained highest accuracy rate as the phase
Like degree threshold value.
A kind of embodiment with reference to first aspect, the method also includes: extract the characteristic information of the facial image；
Fraction assessment is carried out based on the characteristic information, obtains face mass fraction；Judge whether the face mass fraction reaches pre-
The quality score thresholds first saved；Correspondingly, the similarity between the facial image and the custom image pre-saved is calculated,
It include: to calculate the facial image when judging that the face mass fraction reaches the quality score thresholds and pre-save
Custom image between similarity.
A kind of embodiment with reference to first aspect, it is each described to repeat to shop customer and be corresponding with customer's figure
Picture, each corresponding original face mass fraction pre-saved of the custom image, the method also includes: judging
Face mass fraction is stated to be greater than and the original face mass fraction for repeating to the corresponding custom image of shop customer
When, the custom image is replaced with the facial image, wherein the similarity between the facial image and the custom image
Reach the similarity threshold.
A kind of embodiment with reference to first aspect, it is each described to repeat to shop customer and be corresponding with multiple customers' figures
Picture, each custom image are corresponding with the original face mass fraction pre-saved, the method also includes: judging
The face mass fraction be greater than with it is described repeat to shop customer corresponding multiple original face mass fractions when, will be the multiple
The corresponding custom image of the smallest original face mass fraction is determined as target customers's image in original face mass fraction；With institute
It states facial image and replaces target customers's image；Wherein, similar between the facial image and target customers's image
Degree reaches the similarity threshold.
A kind of embodiment with reference to first aspect, the characteristic information include: that face blocks characteristic information, face obscures
Characteristic information and human face posture characteristic information are spent, it is described that fraction assessment is carried out based on the characteristic information, obtain face quality
Score, comprising: posture score is obtained based on the human face posture characteristic information；It is obtained based on the face fuzziness characteristic information
Face values of ambiguity；Characteristic information is blocked based on the face obtain face block score；Based on formula face-quality-
Score=1/ (f1 × pose+f2 × blurness+f3 × occlusion), is calculated the face mass fraction, wherein
Face-quality-score is the face mass fraction, and pose is the posture score, and blurness is the face
Values of ambiguity, occlusion are that the face blocks score, and f1, f2 and f3 are respectively the weighted value pre-set.
A kind of embodiment with reference to first aspect, it is described that posture score is obtained based on the human face posture characteristic information,
It include: that the human face posture characteristic information, the human face posture characteristic information are obtained according to the human face posture model pre-saved
Including three Eulerian angles angle value corresponding with the facial image；Based on formula pose=p1 × yaw+p2 × row+p3 ×
The posture score is calculated in pitch, wherein pose is the posture score, and pitch is the pitching around X-axis rotation
Angle, yaw are the yaw angles around Y-axis rotation, and roll is around the roll angle of Z axis rotation, and p1, p2, p3 are pre-set
Weighted value；And described face values of ambiguity is obtained based on the face fuzziness characteristic information, comprising: according to pre-saving
Face fuzziness model, obtain include blurness the face fuzziness characteristic information, blurness be the face
Values of ambiguity；And it is described block characteristic information based on the face and obtain face block score, comprising: according to pre-saving
Face critical point detection model, obtain the area of corresponding with the facial image multiple key areas；According to pre-saving
Face block model, obtain the face and block characteristic information, the face blocks characteristic information and includes and each pass
The corresponding area that is blocked in key range；For each key area, by the area of the key area and the key area pair
The overlapping region for the area that is blocked answered obtains blocking point for the key area divided by the corresponding area that is blocked of the key area
Number；Based on formula occlusion=k1 × A1_occlusion+k2 × A2_occlusion+ ...+kn × An_occlusion,
The face is calculated and blocks score, wherein occlusion is that the face blocks score, A1_occlusion, A2_
Occlusion ... An_occlusion is that each key area blocks score, k1, k2 ... kn is the score pre-set
Weighted value.
A kind of embodiment with reference to first aspect described repeats to the corresponding custom image of shop customer with same
Common corresponding one pre-save to shop number, the method also includes: by the customer be confirmed as described in repeat to
The described of shop customer adds one to shop number；Alternatively, by the customer be confirmed as described in for the first time to shop customer to shop number
Set one.
Second aspect, the embodiment of the invention provides a kind of image processing apparatus, described device includes: acquisition module, is used
In the facial image for obtaining customer to be identified；Computing module, the custom image for calculating the facial image Yu pre-saving
Between similarity, obtain at least one similarity；Judge execution module, for judge at least one described similarity whether
The similarity threshold all not up to pre-saved；And when the judgment result is yes, determine the customer be for the first time to shop customer, and
It is saved the facial image as new custom image；Also when the judgment result is No, the customer is determined to repeat
To shop customer.
The third aspect the embodiment of the invention provides a kind of electronic equipment, including processor and connects with the processor
The memory connect, the memory is interior to store computer program, when the computer program is executed by the processor, so that
The electronic equipment executes method described in any one of first aspect embodiment.
Fourth aspect, the embodiment of the invention provides a kind of computer readable storage medium, the computer-readable storage
Computer program is stored in medium, when the computer program is run on computers, so that the computer executes the
Method described in any one of one side.
Compared with prior art, a kind of image processing method, the device, electronic equipment, meter of various embodiments of the present invention proposition
Calculation machine readable storage medium storing program for executing, electronic equipment first calculate the facial image and pre- after the facial image for obtaining customer to be identified
The similarity between custom image first saved, obtains at least one similarity；Then judging at least one described similarity is
The no similarity threshold all not up to pre-saved；When to be, determine that the customer is for the first time to shop customer, and by the people
Face image is saved as new custom image；Otherwise determine that the customer is to repeat to shop customer.In this way, can be with
The customer automatically identified to shop is new customer or repeats to shop customer, hotel owner or sales force do not have to further according to itself
Memory identifies customer.
Other feature and advantage disclosed by the embodiments of the present invention will illustrate in the following description, alternatively, Partial Feature
It can deduce from specification with advantage or unambiguously determine, or by implementing above-mentioned technology disclosed by the embodiments of the present invention
It can be learnt that.
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
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below will be to needed in the embodiment attached
Figure is briefly described, it should be understood that the following drawings illustrates only certain embodiments of the present invention, therefore is not construed as pair
The restriction of range for those of ordinary skill in the art without creative efforts, can also be according to this
A little attached drawings obtain other relevant attached drawings.
Fig. 1 is the schematic diagram of electronic equipment provided in an embodiment of the present invention；
Fig. 2 is one of the flow chart of image processing method provided in an embodiment of the present invention；
Fig. 3 is the two of the flow chart of image processing method provided in an embodiment of the present invention；
Fig. 4 a is face critical point detection result schematic diagram provided in an embodiment of the present invention；
Fig. 4 b is face occlusion area testing result schematic diagram provided in an embodiment of the present invention；
Fig. 4 c blocks score calculation schematic diagram to be provided in an embodiment of the present invention；
Fig. 5 is three axis Eulerian angles schematic diagram provided in an embodiment of the present invention；
Fig. 6 is the structural block diagram of image processing apparatus provided in an embodiment of the present invention.
Specific embodiment
Below in conjunction with attached drawing in the embodiment of the present invention, technical solution in the embodiment of the present invention carries out clear, complete
Ground description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.Based on this
The embodiment of invention, those skilled in the art's every other implementation obtained without making creative work
Example, shall fall within the protection scope of the present invention.
In traditional scheme, artificial memory is generally relied on to recognize shop customer to be still repeated as many times to shop for the first time and arrive
Shop.But since hotel owner or sales force's manpower and memory are limited, it is difficult repeatedly to carry out effectively all to customer family
Memory, therefore, so that hotel owner or sales force know that customer's is more difficult to shop number.
In order to improve the above problem, the embodiment of the invention provides a kind of image processing method, device, electronic equipment, meters
Calculation machine readable storage medium storing program for executing, the mode which can be used corresponding software, hardware and soft or hard combination are realized.Below to this hair
Bright embodiment describes in detail.
Firstly, describing the electronic equipment of image processing method for realizing the embodiment of the present invention, device referring to Fig.1
100。
Electronic equipment 100 may include processor 110, memory 120 and image processing apparatus.
Processor 110, memory 120 these components can be by the bindiny mechanisms of bus system and/or other forms (not
Show) interconnection.It should be noted that the component and structure of electronic equipment 100 shown in FIG. 1 be it is illustrative, and not restrictive,
As needed, the electronic equipment 100 also can have other assemblies and structure.Described image processing unit includes at least one
The operation of electronic equipment 100 can be stored in the memory 120 or is solidificated in the form of software or firmware (firmware)
Software function module in system (operating system, OS).The processor 110 is deposited in memory 120 for executing
The executable module of storage, such as software function module or computer program that described image processing unit includes.
The memory 120 may include one or more computer program products, and the computer program product can be with
Including various forms of computer readable storage mediums, such as volatile memory and/or nonvolatile memory.It is described volatile
Property memory for example may include random access memory (RAM) and/or cache memory (cache) etc..It is described non-easy
The property lost memory for example may include read-only memory (ROM), hard disk, flash memory etc..On the computer readable storage medium
It can store one or more computer program instructions, processor 110 can run described program instruction, described below to realize
The embodiment of the present invention in it is expected realize function.It can also be stored in the computer readable storage medium various using journey
Sequence and various data, such as application program use and/or the various data generated etc..
Processor 110 may be a kind of IC chip, the processing capacity with signal.Above-mentioned processor 110 can
To be general processor, including central processing unit (Central Processing Unit, abbreviation CPU), network processing unit
(Network Processor, abbreviation NP) etc.；Can also be digital signal processor (DSP), specific integrated circuit (ASIC),
Field programmable gate array (FPGA) either other programmable logic device, discrete gate or transistor logic, discrete hard
Part component.Processor 110 may be implemented or execute disclosed each method, step and logic diagram in the embodiment of the present invention.
Referring to figure 2., Fig. 2 is a kind of flow chart of image processing method provided in an embodiment of the present invention.It below will be to Fig. 2
Shown in process be described in detail, which comprises
Step S110: the facial image of customer to be identified is obtained.
As an alternative embodiment, can be in enter place or the cashier's installation encapsulation in advance of new retail location
There is the intelligent video camera head of face detection module to collect the facial image of shop customer.Face figure has been detected in intelligent video camera head
As and after collecting facial image, facial image can be sent to electronic equipment 100, so that electronic equipment 100 is available
The facial image is identified.
Wherein, electronic equipment 100 can be carried out wired or is wirelessly connected with intelligent video camera head, to carry out the logical of data
Letter.Certainly, intelligent video camera head can also be integrated in electronic equipment 100.Since the angle of intelligent video camera head and face is bigger, people
The characteristic present of face is poorer, therefore, the position of intelligent video camera head can be made horizontal with the face holding of customer as much as possible,
To guarantee the quality of collected facial image.
Step S120: calculating the similarity between the facial image and the custom image pre-saved, obtains at least one
A similarity.
Optionally, multiple customers can be pre-established in the local data base of electronic equipment 100 or cloud database
ID preserves custom image corresponding with the customer ID at each customer ID.Wherein, one can be saved under a customer ID
The custom image of the customer can also save the custom image that multiple customers are photographed under different shooting angles, may be used also
To save the custom image that multiple customers are photographed under different shooting times.
Electronic equipment 100 after getting facial image, can by facial image be pre-stored in it is every in database
It opens custom image and carries out similarity calculation.
Since feature vector included between identical two facial images is similar, carrying out similarity calculation
When, optionally, for each custom image, convolutional neural networks can be first passed through and extract the facial image and the Gu respectively
The multidimensional characteristic vectors (such as 6400 dimensional feature vectors) of objective image, the multidimensional characteristic vectors for then calculating the facial image arrive
It corresponding Euclidean distance and is normalized between the multidimensional characteristic vectors of the custom image, obtains multiple distance values；Then will
The multiple distance value is filled into the M × N similarity matrix constructed in advance, and the calculated result of similarity matrix is the people
Similarity between face image and the custom image, wherein M and N is positive integer.
Step S130: judge the similarity threshold whether at least one described similarity all not up to pre-saves.
When determining similarity threshold, optionally, it can first establish one and include multiple samples pictures (such as Wan Zhangtu
Piece) test set, accuracy rate of the test set under different threshold values is then tested by experiment, it is finally that obtained highest is accurate
Threshold value under rate is as similarity threshold.
It is of course also possible to use other modes determine similarity threshold, such as directly acquire developer according to individual
Empirically determined empirical value out.
Further, since when carrying out similarity judgement, the judging result error of the crowd very big to age very little and age
It is larger, so that judging result less robust.It therefore, can also be by age characteristics come to similar before carrying out similarity judgement
Degree threshold value optimizes.As an alternative embodiment, electronic equipment 100 can be first before carrying out similarity judgement
The age characteristics that model extraction facial image is extracted by the age is preset judging that the age characteristics characterization customer is in
Age level when, the first threshold pre-saved is determined as the similarity threshold；Otherwise, the second threshold that will be pre-saved
Value is determined as the similarity threshold.Judge that the age level of customer corresponding with facial image is in preset age level
When, using the bigger similarity threshold of numerical value.
Optionally, the preset age level may include an age level, for example, 0-12 years old or 60 years old with
On, it also may include multiple age levels, for example, 0-12 years old and 60 years old or more.
Wherein, it is worth noting that, the first threshold is greater than the second threshold, and second threshold can be using above
The similarity threshold determined by test set, first threshold can be on the basis of second threshold, and increase by one is normal less than 1
Number, certainly, first threshold is again smaller than 1.
When sport career age extracts model, a large amount of face sample data can be obtained in advance, wherein each face sample
It all include the real age data label corresponding with the face sample artificially marked in data, then by a large amount of face sample
Notebook data is input in convolutional neural networks and is trained, so that training, which obtains the age, extracts model.
Step S140: when to be, determine the customer be for the first time to shop customer, and using the facial image as newly
Custom image is saved.
Certainly, when facial image is saved in database as new custom image, it is also necessary to which being in database should
Shop customer is arrived for the first time and establishes a new customer ID, then facial image save and carries out pass in the new customer ID
Connection.
It, optionally, can be on the basis of original customer ID, by original customer ID value when establishing new customer ID
Maximum customer ID adds one, to form new customer ID, new customer ID can also be formed by way of random coded.
Step S150: being to determine that the customer is to repeat to shop customer to be no.
As an alternative embodiment, electronic equipment 100 can also be attached with audio frequency apparatus.In determining and institute
Stating the corresponding customer of facial image is after repeating to shop customer, and electronic equipment 100 can also prompt sale people by audio frequency apparatus
The member customer is patron；After determining customer corresponding with the facial image to arrive shop customer for the first time, electronic equipment 100 is also
Can prompt the sales force customer is for the first time to shop customer.
As another optional embodiment, it can also be preserved in database corresponding with each customer ID to shop
Number.It is determining that the customer is to repeat to shop customer still for the first time to after the customer of shop, the customer can also be determined
For customer identification corresponding to add one to shop number.
It under this embodiment, optionally, can be by the repetition after determining the customer is to repeat to shop customer
Add one to shop number to shop customer is corresponding.When determining that the customer is and to save the facial image for the first time to shop customer
In to database as new custom image after, due to the customer be new customer, can for the customer creation newly
After customer ID, directly set one to shop number for the new customer ID is corresponding, that is, complete by add to shop number one operation.
It is worth noting that when determining that the customer is to repeat to shop customer, need to repeat to this shop customer to shop
Number adds for the moment, due to the facial image it is also possible to similar between the custom image of customer ID different from belonging in database
Degree reaches similarity threshold, for example, similarity threshold is 75%, facial image and repeats between the custom image of shop customer A
Similarity reach 76%, the similarity between facial image and the custom image for repeating to shop customer B reaches 80%, at this point,
As a kind of fault-tolerant processing mode, it is believed that carry out similarity-rough set multiple custom images corresponding to repeat to shop care for
In visitor, similarity is highest to repeat to customer corresponding to shop customer and the facial image as the same customer.Therefore, above
Example in, by the corresponding customer of facial image be determined as similarity be 80% custom image corresponding to repeat to shop care for
The corresponding customer of facial image is determined as customer B by visitor, need customer B adding one to shop number at this time.
Optionally, the retention cycle of database can be adjusted according to actual needs, so that businessman preferably analyzes certainly
Oneself management state and service scenario etc. to customer.If the demand of businessman is in available one day repeatedly to the customer in shop
Quantity, then can establish daily face database, if the demand of businessman is in available one month repeatedly to the Gu in shop
Objective quantity, then can establish long-term face database.
Since new public safety has its unique complexity, for example camera shooting picture quality is relatively low and exists a large amount of
Information redundancy, people dress with dress up that changeable and be subject to door curtain/cargo etc. blocks, the above many factors make right
Error rate is higher when portrait picture is identified.
In order to improve the accuracy of identification, as an alternative embodiment, calculating face figure executing step S120
As the assessment of face mass fraction can also be carried out to facial image, to filter out with before the similarity between custom image
The facial image of face mass fraction qualified (reaching quality score thresholds) carries out similarity calculation.Fig. 3 is please referred to, face is carried out
Mass fraction assessment process may include:
Step S111: the characteristic information of the facial image is extracted.
Step S112: fraction assessment is carried out based on the characteristic information, obtains face mass fraction；
Step S113: judge whether the face mass fraction reaches the quality score thresholds pre-saved.
Correspondingly, when judging that the face mass fraction reaches the quality score thresholds, then execute step S120.
I.e. when electronic equipment 100 reaches quality score thresholds in the face mass fraction for judging the facial image captured
When, similarity calculation just is carried out to the facial image and custom image, in the face mass fraction for the facial image that judgement was photographed
Not up to quality score thresholds when, directly the facial image can be abandoned.
As an alternative embodiment, every custom image can correspond to the original face matter pre-saved
Measure score.Under this embodiment for carrying out the assessment of face mass fraction to facial image, optionally, if each repetition
It is corresponding with the custom image to shop customer, then determining that the customer is that some repeats to shop customer, and the customer
Face mass fraction be greater than this when repeating to original face mass fraction corresponding to the customer of shop, with the facial image of the customer
The custom image of the customer is replaced, and is stored in the database.Such as when judging the corresponding customer of current facial image
For the customer A for repeating to shop, and the face mass fraction of current facial image be greater than be stored in database with A pairs of customer
When the original face mass fraction for the custom image answered, electronic equipment 100 is replaced in database with current facial image
Custom image.
Optionally, repeating to shop customer and may being corresponding with multiple custom images when being stored in database simultaneously (such as has
Custom image A, custom image B, custom image C tri- open custom image and are stored in database), and every custom image is all respective
It is corresponding with the original face mass fraction pre-saved.So determine the customer be some repeat to shop customer, and
The face mass fraction of the customer is greater than this and repeats to the corresponding original face quality of the corresponding a certain custom image of shop customer
When score, it is greater than the corresponding original face mass fraction of custom image A, electronic equipment 100 can be replaced with facial image
Fall custom image A.Furthermore, however, it is determined that the face mass fraction of the customer is greater than this and repeats to corresponding multiple customers figure of shop customer
When as corresponding original face mass fraction, such as the corresponding original face matter of simultaneously greater than custom image A and custom image B
Score is measured, at this point, electronic equipment 100 can replace minimum that of mass fraction in multiple custom images with facial image
Custom image, for example, facial image face mass fraction > custom image B original face mass fraction > custom image A
Original face mass fraction, electronic equipment 100 replace custom image B with facial image.
It will be introduced below to by preparatory trained model to calculate the face mass fraction of facial image.
Wherein, trained model may include that face blocks model in advance, for avoiding dress and dressing due to people
The changeable or identification higher problem of error rate caused by be subject to door curtain/cargo etc. blocks, in addition, trained in advance
Model can also include: at least one of human face posture model, face fuzziness model, age models.
Model training process:
Electronic equipment 100 can obtain a large amount of face sample when being trained for some model wait training in advance
Notebook data, wherein it include artificially having marked in each face sample data, study required for model to be trained with this
The corresponding label of feature.Then a large amount of face sample data is input in convolutional neural networks and is trained, to obtain
Corresponding model.
Wherein, when needing to train face attitude mode, label included by face sample data is various artificial labels
Human face posture, when needing to train face fuzziness model, label included by face sample data is various artificial labels
Face values of ambiguity, when needing that face is trained to block model, label included by face sample data is various artificial marks
The type of barrier of note, when needing sport career age model, label included by face sample data is and the face sample data
The real age of corresponding face.
It include below that face blocks model, human face posture model, face fuzziness model and is with preparatory trained model
Example is introduced, correspondingly, in advance trained model extraction to the characteristic information of facial image may include: that face blocks
Characteristic information, face fuzziness characteristic information and human face posture characteristic information.
Electronic equipment 100 can be based respectively on face and block characteristic information, face fuzziness characteristic information and face appearance
State characteristic information is calculated face and blocks score, face values of ambiguity and posture score.Under this embodiment, base
Fraction assessment is carried out in the characteristic information, obtains face mass fraction, comprising:
Based on formula face-quality-score=1/ (f1 × pose+f2 × blurness+f3 × occlusion),
The face mass fraction is calculated, wherein face-quality-score be the face mass fraction, pose be based on
The posture score that the human face posture characteristic information obtains, blurness are obtained based on the face fuzziness characteristic information
Face values of ambiguity, occlusion are to block the face that characteristic information obtains based on the face to block score, f1, f2 and
F3 is respectively the normalized weight value pre-set.F1, f2 and f3 can be adjusted according to the actual situation, face matter
It is higher to measure score, it is higher to represent face quality, is more easily identified correct.
Below will for electronic equipment 100 be based respectively on face block characteristic information, face fuzziness characteristic information and
Human face posture characteristic information, is calculated face and blocks score, face values of ambiguity and posture score and be introduced.
When calculating face blocks score occlusion, gone out first by existing face critical point detection model inspection
Then multiple (such as 84) key points of face with face block model inspection again and go out face and be blocked as is shown in fig. 4 a
Region, as shown in fig 4b.Face block model can provide face key area (face key area includes left/right eyebrow,
Left/right eyes, nose, mouth) area fraction that is blocked, the calculation of the area is as illustrated in fig. 4 c.It is blocked point with mouth
For number (mouth_occlusion), the mouth region that occlusion_mouth=critical point detection arrives is overlapped with occlusion area
The mouth region area that region area/critical point detection arrives similarly can be calculated left eyebrow and block score (left_
Eyebrow_occlusion), right eyebrow blocks score (right_eyebrow_occlusion), left eye eyeball blocks score
(left_eye_occlusion), right eye eyeball blocks score (right_eyebrow_occlusion), nose blocks score
(nose_occlusion), score (contour_occlusion) is blocked in facial contour.Therefore, whole score is blocked
Occlusion=k1 × contour_occlusion+k2 × mouth_occlusion+k3 × nose_occlusion+k4 ×
left_eye_occlusion+k5×right_eye_occlusion+k6×left_eyebrow_occlusion+k7×
Right_eyeb row_occlusion, wherein k1-k7 is the weighted value of each score.Under normal circumstances, k1 > k3 > k2 >
K4=k5 > k6=k7.Occlusion score is higher, indicates that face serious shielding degree is higher, represents face mass fraction and get over
It is low, i.e., it is more difficult to differentiate the ownership of the face.
When calculating face values of ambiguity blurness, face fuzziness model can export one to the facial image of input
Face values of ambiguity blurness, size is between 0-1, and numerical value is bigger, and expression face is fuzzyyer, and face mass fraction is lower,
It is more difficult to differentiate the ownership of the face.
As shown in figure 5, human face posture model can export face to the facial image of input when calculating posture score pose
Tri- Eulerian angles angle value of yaw, row, pitch, wherein pitch be around X-axis rotation pitch angle, yaw be around Y-axis revolve
The yaw angle turned, roll are the roll angles around Z axis rotation.Posture score pose=p1 × yaw+p2 × row+p3 of face ×
Pitch, p1, p2, p3 are the weight of each score.Under normal circumstances, p1 > p3 > p2.Pose score is higher, represents face matter
It is lower to measure score, is more difficult to differentiate the ownership of the face.
A kind of a kind of image processing method applied to electronic equipment 100 provided in an embodiment of the present invention, electronic equipment 100
After the facial image for obtaining customer to be identified, first calculate similar between the facial image and the custom image pre-saved
Degree, obtains at least one similarity；Then judge the similarity whether at least one described similarity all not up to pre-saves
Threshold value；When to be, determine that the customer is for the first time to shop customer, and using the facial image as new custom image progress
It saves；Otherwise determine that the customer is to repeat to shop customer.In this way, the customer that can be automatically identified to shop is newly to care for
Visitor still repeats to shop customer, and hotel owner or sales force do not have to the memory further according to itself to identify to customer.
Corresponding to the image processing method that Fig. 2 is provided, Fig. 6 is please referred to, the embodiment of the invention also provides at a kind of image
Device 400 is managed, which may include:
Module 410 is obtained, for obtaining the facial image of customer to be identified；
Computing module 420 is obtained for calculating the similarity between the facial image and the custom image pre-saved
At least one similarity；
Judge execution module 430, the phase for judging whether all not up to pre-save at least one described similarity
Like degree threshold value；And when the judgment result is yes, determine the customer be for the first time to shop customer, and using the facial image as newly
Custom image saved；Also when the judgment result is No, determine that the customer is to repeat to shop customer.
Optionally, described device can also include: extraction module, evaluation module and judgment module.
The extraction module, for extracting the characteristic information of the facial image；
The evaluation module obtains face mass fraction for carrying out fraction assessment based on the characteristic information；
The judgment module, for judging whether the face mass fraction reaches the quality score thresholds pre-saved.
Correspondingly, computing module 420, for when the judgment module is judged as YES, just calculate the facial image with
The similarity between custom image pre-saved.
Optionally, the acquisition module 410, can be also used for the age characteristics for extracting the facial image；
The judgement execution module 430 can be also used for judging the age characteristics characterization customer in default
Age level when, the first threshold pre-saved is determined as the similarity threshold；Otherwise, the second threshold that will be pre-saved
Value is determined as the similarity threshold；Wherein, the first threshold is greater than the second threshold.
Optionally, described device can also include test module and determining module.
The acquisition module 410 can be also used for obtaining the test set including multiple samples pictures；
The test module, for testing accuracy rate of the test set under different threshold values；
The determining module, the threshold value under highest accuracy rate for will obtain is as the similarity threshold.
Optionally, each described to repeat to shop customer and be corresponding with the custom image, every custom image pair
Answer the original face mass fraction pre-saved.
The judgement execution module 430 can be also used for repeating to described judging that the face mass fraction is greater than
When the original face mass fraction of the corresponding custom image of shop customer, the customer is replaced with the facial image and is schemed
Picture, wherein the similarity between the facial image and the custom image reaches the similarity threshold.
Optionally, each described to repeat to shop customer and be corresponding with multiple described custom images, every custom image pair
There should be the original face mass fraction pre-saved.
The judgement execution module 430 can be also used for repeating to described judging that the face mass fraction is greater than
When the corresponding multiple original face mass fractions of shop customer, by the smallest original face in the multiple original face mass fraction
The corresponding custom image of mass fraction is determined as target customers's image；Target customers's image is replaced with the facial image；
Wherein, the similarity between the facial image and target customers's image reaches the similarity threshold.
Optionally, the characteristic information may include that face blocks characteristic information, face fuzziness characteristic information and people
Face posture feature information, the evaluation module, for based on formula face-quality-score=1/ (f1 × pose+f2 ×
Blurness+f3 × occlusion), the face mass fraction is calculated, wherein face-quality-score is institute
Face mass fraction is stated, pose is the posture score obtained based on the human face posture characteristic information, and blurness is based on institute
The face values of ambiguity that face fuzziness characteristic information obtains is stated, occlusion is to block characteristic information based on the face to obtain
To face block score, f1, f2 and f3 are respectively the weighted value pre-set.
Wherein, the computing module 420 is also used to obtain the face appearance according to the human face posture model pre-saved
State characteristic information, the human face posture characteristic information include three Eulerian angles angle value corresponding with the facial image；Based on public affairs
The posture score is calculated in formula pose=p1 × yaw+p2 × row+p3 × pitch, wherein pose is the posture point
Number, pitch are the pitch angles around X-axis rotation, and yaw is the yaw angle around Y-axis rotation, and roll is turning over around Z axis rotation
Roll angle, p1, p2, p3 are the weighted value pre-set；And
It is also used to that it is special to obtain the face fuzziness including blurness according to the face fuzziness model pre-saved
Reference breath, blurness are the face values of ambiguity；And
It is also used to obtain multiple passes corresponding with the facial image according to the face critical point detection model pre-saved
The area of key range；Model is blocked according to the face pre-saved, the face is obtained and blocks characteristic information, the face blocks
Characteristic information includes the area that is blocked corresponding with each key area；For each key area, by the key
The overlapping region of the area in the region area that is blocked corresponding with the key area is divided by the corresponding face that is blocked of the key area
Product, obtain the key area blocks score；Based on formula occlusion=k1 × A1_occlusion+k2 × A2_
Occlusion+ ...+kn × An_occlusion is calculated the face and blocks score, wherein occlusion is the people
Face blocks score, A1_occlusion, A2_occlusion ... An_occlusion is that each key area blocks score,
K1, k2 ... kn is the fractional weight value pre-set.
Optionally, it described repeat to the corresponding custom image of shop customer with same and pre-saves for corresponding one jointly
Arrive shop number.Described device can also include increasing module, care for for repeating to shop described in being confirmed as the customer
The described of visitor adds one to shop number；Alternatively, by the customer be confirmed as described in set one to shop number to shop customer for the first time.
The technical effect of device provided by the present embodiment, realization principle and generation is identical with previous embodiment, for letter
It describes, Installation practice part does not refer to place, can refer to Fig. 2-Fig. 5 corresponding contents in preceding method embodiment.
In addition, the embodiment of the invention also provides a kind of electronic equipment, including processor and it is connected to the processor
Memory, computer program is stored in the memory, when the computer program is executed by the processor, so that institute
It states electronic equipment and executes image processing method provided by any one of first embodiment embodiment.Wherein, electronic equipment
Structural schematic diagram may refer to Fig. 1.
In addition, the embodiment of the invention also provides a kind of computer readable storage medium, in the computer-readable storage medium
Computer program is stored in matter, when the computer program is run on computers, so that the computer executes this hair
Image processing method provided by any one of bright embodiment.
In addition, the embodiment of the invention also provides a kind of computer program, the computer program can store beyond the clouds or
On the storage medium of person local, when the computer program is run on computers, so that the computer executes the present invention
Image processing method provided by any one embodiment.
In conclusion image processing method, device, electronic equipment, computer-readable storage that the embodiment of the present invention proposes
Medium first calculates between the facial image and the custom image pre-saved after the facial image for obtaining customer to be identified
Similarity, obtain at least one similarity；Then judge what whether at least one described similarity all not up to pre-saved
Similarity threshold；When to be, determine that the customer is for the first time to shop customer, and using the facial image as new customer's figure
As being saved；Otherwise determine that the customer is to repeat to shop customer.In this way, the customer to shop can be automatically identified
It is new customer or repeats to shop customer, hotel owner or sales force does not have to the memory further according to itself to know to customer
Not.
In several embodiments provided herein, it should be understood that disclosed device and method can also pass through
Other modes are realized.The apparatus embodiments described above are merely exemplary, for example, flow chart and block diagram in attached drawing
Show the device of multiple embodiments according to the present invention, the architectural framework in the cards of method and computer program product,
Function and operation.In this regard, each box in flowchart or block diagram can represent the one of a module, section or code
Part, a part of the module, section or code, which includes that one or more is for implementing the specified logical function, to be held
Row instruction.It should also be noted that function marked in the box can also be to be different from some implementations as replacement
The sequence marked in attached drawing occurs.For example, two continuous boxes can actually be basically executed in parallel, they are sometimes
It can execute in the opposite order, this depends on the function involved.It is also noted that every in block diagram and or flow chart
The combination of box in a box and block diagram and or flow chart can use the dedicated base for executing defined function or movement
It realizes, or can realize using a combination of dedicated hardware and computer instructions in the system of hardware.
In addition, each functional module in each embodiment of the present invention can integrate one independent portion of formation together
Point, it is also possible to modules individualism, an independent part can also be integrated to form with two or more modules.
It, can be with if the function is realized and when sold or used as an independent product in the form of software function module
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.
It should be noted that, in this document, relational terms such as first and second and the like are used merely to a reality
Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation
In any actual relationship or order or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended to
Non-exclusive inclusion, so that the process, method, article or equipment including a series of elements is not only wanted including those
Element, but also including other elements that are not explicitly listed, or further include for this process, method, article or equipment
Intrinsic element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that
There is also other identical elements in process, method, article or equipment including the element.
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, any
Those familiar with the art in the technical scope disclosed by the present invention, can easily think of the change or the replacement, and should all contain
Lid is within protection scope of the present invention.Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (12)
1. a kind of image processing method, which is characterized in that the described method includes:
Obtain the facial image of customer to be identified；
The similarity between the facial image and the custom image pre-saved is calculated, at least one similarity is obtained；
Judge the similarity threshold whether at least one described similarity all not up to pre-saves；
When to be, determine that the customer is for the first time to shop customer, and using the facial image as new custom image progress
It saves；
When to be no, determine that the customer is to repeat to shop customer.
2. the method according to claim 1, wherein the method also includes:
Extract the age characteristics of the facial image；
It is when judging that the age characteristics characterization customer is in preset age level, the first threshold pre-saved is true
It is set to the similarity threshold；
Otherwise, the second threshold pre-saved is determined as the similarity threshold；Wherein, the first threshold is greater than described the
Two threshold values.
3. the method according to claim 1, wherein the method also includes:
Obtain the test set including multiple samples pictures；
Test accuracy rate of the test set under different threshold values；
Using the threshold value under obtained highest accuracy rate as the similarity threshold.
4. the method according to claim 1, wherein the method also includes:
Extract the characteristic information of the facial image；
Fraction assessment is carried out based on the characteristic information, obtains face mass fraction；
Judge whether the face mass fraction reaches the quality score thresholds pre-saved；
Correspondingly, calculating the similarity between the facial image and the custom image pre-saved, comprising:
When judging that the face mass fraction reaches the quality score thresholds, calculates the facial image and pre-save
Similarity between custom image.
5. according to the method described in claim 4, it is characterized in that, repeating to shop customer described in each is corresponding with Zhang Suoshu Gu
Objective image, the corresponding original face mass fraction pre-saved of every custom image, the method also includes:
The described original of the corresponding custom image of shop customer is repeated to described judging that the face mass fraction is greater than
When face mass fraction, the custom image is replaced with the facial image, wherein the facial image and the custom image
Between similarity reach the similarity threshold.
6. according to the method described in claim 4, it is characterized in that, repeating to shop customer described in each is corresponding with multiple described Gus
Objective image, every custom image are corresponding with the original face mass fraction pre-saved, the method also includes:
Judge the face mass fraction be greater than with it is described repeat to shop customer corresponding multiple original face mass fractions when,
The corresponding custom image of original face mass fraction the smallest in the multiple original face mass fraction is determined as target to care for
Objective image；
Target customers's image is replaced with the facial image；Wherein, the facial image and target customers's image it
Between similarity reach the similarity threshold.
7. according to the method described in claim 4, it is characterized in that, the characteristic information includes: that face blocks characteristic information, people
Face fuzziness characteristic information and human face posture characteristic information, it is described that fraction assessment is carried out based on the characteristic information, obtain people
Face mass fraction, comprising:
Posture score is obtained based on the human face posture characteristic information；
Face values of ambiguity is obtained based on the face fuzziness characteristic information；
Characteristic information is blocked based on the face obtain face block score；
Based on formula face-quality-score=1/ (f1 × pose+f2 × blurness+f3 × occlusion), calculate
Obtain the face mass fraction, wherein face-quality-score is the face mass fraction, and pose is the appearance
State score, blurness are the face values of ambiguity, and occlusion is that the face blocks score, and f1, f2 and f3 divide
The weighted value that Wei do not pre-set.
8. the method according to the description of claim 7 is characterized in that
It is described that posture score is obtained based on the human face posture characteristic information, comprising:
According to the human face posture model pre-saved, the human face posture characteristic information, the human face posture characteristic information are obtained
Including three Eulerian angles angle value corresponding with the facial image；
Based on formula pose=p1 × yaw+p2 × row+p3 × pitch, the posture score is calculated, wherein pose is
The posture score, pitch are the pitch angles around X-axis rotation, and yaw is the yaw angle around Y-axis rotation, and roll is around Z
The roll angle of axis rotation, p1, p2, p3 are the weighted value pre-set；And
It is described that face values of ambiguity is obtained based on the face fuzziness characteristic information, comprising:
According to the face fuzziness model pre-saved, the face fuzziness characteristic information including blurness is obtained,
Blurness is the face values of ambiguity；And
It is described block characteristic information based on the face and obtain face block score, comprising:
According to the face critical point detection model pre-saved, the face of multiple key areas corresponding with the facial image is obtained
Product；
Model is blocked according to the face pre-saved, the face is obtained and blocks characteristic information, the face blocks characteristic information
Including the area that is blocked corresponding with each key area；
For each key area, by the coincidence of the area of the key area area that is blocked corresponding with the key area
Divided by the corresponding area that is blocked of the key area, obtain the key area blocks score in region；
Based on formula occlusion=k1 × A1_occlusion+k2 × A2_occlusion+ ...+kn × An_occlusion,
The face is calculated and blocks score, wherein occlusion is that the face blocks score, A1_occlusion, A2_
Occlusion ... An_occlusion is that each key area blocks score, k1, k2 ... kn is the score pre-set
Weighted value.
9. method according to claim 1 to 8, which is characterized in that described repeat to shop customer couple with same
The custom image answered correspond to jointly one pre-save to shop number, the method also includes:
By the customer be confirmed as described in repeat to and add one to shop number described in the customer of shop；Alternatively,
By the customer be confirmed as described in set one to shop number to shop customer for the first time.
10. a kind of image processing apparatus, which is characterized in that described device includes:
Module is obtained, for obtaining the facial image of customer to be identified；
Computing module obtains at least one for calculating the similarity between the facial image and the custom image pre-saved
A similarity；
Judge execution module, the similarity threshold for judging whether all not up to pre-save at least one described similarity
Value；And when the judgment result is yes, determine that the customer is for the first time to shop customer, and using the facial image as new customer
Image is saved；Also when the judgment result is No, determine that the customer is to repeat to shop customer.
11. a kind of electronic equipment, which is characterized in that described to deposit including processor and the memory being connected to the processor
Computer program is stored in reservoir, when the computer program is executed by the processor, so that the electronic equipment executes
Method described in any one of claim 1-9.
12. a kind of computer readable storage medium, which is characterized in that be stored with computer in the computer readable storage medium
Program, when the computer program is run on computers, so that the computer is executed as any one in claim 1-9
Method described in.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811328072.3A CN109409962A (en) | 2018-11-08 | 2018-11-08 | Image processing method, device, electronic equipment, computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811328072.3A CN109409962A (en) | 2018-11-08 | 2018-11-08 | Image processing method, device, electronic equipment, computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109409962A true CN109409962A (en) | 2019-03-01 |
Family
ID=65472336
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811328072.3A Pending CN109409962A (en) | 2018-11-08 | 2018-11-08 | Image processing method, device, electronic equipment, computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109409962A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111723678A (en) * | 2020-05-27 | 2020-09-29 | 上海瀛之杰汽车信息技术有限公司 | Human face passenger flow identification method, device, equipment and medium suitable for multi-person scene |
| CN111797773A (en) * | 2020-07-07 | 2020-10-20 | 广州广电卓识智能科技有限公司 | Method, device and equipment for detecting occlusion of key parts of human face |
| CN111814569A (en) * | 2020-06-12 | 2020-10-23 | 深圳禾思众成科技有限公司 | Method and system for detecting human face shielding area |
| CN112016469A (en) * | 2020-08-28 | 2020-12-01 | Oppo广东移动通信有限公司 | Image processing method and device, terminal and readable storage medium |
| CN112131915A (en) * | 2019-06-25 | 2020-12-25 | 杭州海康威视数字技术股份有限公司 | Face attendance system, camera and code stream equipment |
| CN113448925A (en) * | 2021-06-25 | 2021-09-28 | 东莞市小精灵教育软件有限公司 | Test question picture optimization method and device, computer equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106096513A (en) * | 2016-06-01 | 2016-11-09 | 深圳信炜科技有限公司 | Fingerprint identification method, fingerprint recognition system and electronic equipment |
| CN107679613A (en) * | 2017-09-30 | 2018-02-09 | 同观科技（深圳）有限公司 | A kind of statistical method of personal information, device, terminal device and storage medium |
| CN107689069A (en) * | 2017-08-24 | 2018-02-13 | 深圳市唯特视科技有限公司 | A kind of image automatic synthesis method blocked based on identification face |
| CN108109044A (en) * | 2017-12-26 | 2018-06-01 | 南京开为网络科技有限公司 | A kind of intelligence retail crm system |
| CN108230293A (en) * | 2017-05-31 | 2018-06-29 | 深圳市商汤科技有限公司 | Determine method and apparatus, electronic equipment and the computer storage media of quality of human face image |
| CN108241836A (en) * | 2016-12-23 | 2018-07-03 | 同方威视技术股份有限公司 | For the method and device of safety check |
-
2018
- 2018-11-08 CN CN201811328072.3A patent/CN109409962A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106096513A (en) * | 2016-06-01 | 2016-11-09 | 深圳信炜科技有限公司 | Fingerprint identification method, fingerprint recognition system and electronic equipment |
| CN108241836A (en) * | 2016-12-23 | 2018-07-03 | 同方威视技术股份有限公司 | For the method and device of safety check |
| CN108230293A (en) * | 2017-05-31 | 2018-06-29 | 深圳市商汤科技有限公司 | Determine method and apparatus, electronic equipment and the computer storage media of quality of human face image |
| CN107689069A (en) * | 2017-08-24 | 2018-02-13 | 深圳市唯特视科技有限公司 | A kind of image automatic synthesis method blocked based on identification face |
| CN107679613A (en) * | 2017-09-30 | 2018-02-09 | 同观科技（深圳）有限公司 | A kind of statistical method of personal information, device, terminal device and storage medium |
| CN108109044A (en) * | 2017-12-26 | 2018-06-01 | 南京开为网络科技有限公司 | A kind of intelligence retail crm system |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112131915A (en) * | 2019-06-25 | 2020-12-25 | 杭州海康威视数字技术股份有限公司 | Face attendance system, camera and code stream equipment |
| CN111723678A (en) * | 2020-05-27 | 2020-09-29 | 上海瀛之杰汽车信息技术有限公司 | Human face passenger flow identification method, device, equipment and medium suitable for multi-person scene |
| CN111814569A (en) * | 2020-06-12 | 2020-10-23 | 深圳禾思众成科技有限公司 | Method and system for detecting human face shielding area |
| CN111797773A (en) * | 2020-07-07 | 2020-10-20 | 广州广电卓识智能科技有限公司 | Method, device and equipment for detecting occlusion of key parts of human face |
| CN112016469A (en) * | 2020-08-28 | 2020-12-01 | Oppo广东移动通信有限公司 | Image processing method and device, terminal and readable storage medium |
| CN113448925A (en) * | 2021-06-25 | 2021-09-28 | 东莞市小精灵教育软件有限公司 | Test question picture optimization method and device, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109409962A (en) | Image processing method, device, electronic equipment, computer readable storage medium | |
| CN108319953B (en) | Occlusion detection method and device, electronic equipment and the storage medium of target object | |
| CN105518709B (en) | The method, system and computer program product of face for identification | |
| CN106897658B (en) | Method and device for identifying human face living body | |
| EP2546782B1 (en) | Liveness detection | |
| CN107423690A (en) | A kind of face identification method and device | |
| CN109086718A (en) | Biopsy method, device, computer equipment and storage medium | |
| CN108875522A (en) | Face cluster methods, devices and systems and storage medium | |
| CN108701216A (en) | A kind of face shape of face recognition methods, device and intelligent terminal | |
| CN108229330A (en) | Face fusion recognition methods and device, electronic equipment and storage medium | |
| CN108416336A (en) | A kind of method and system of intelligence community recognition of face | |
| CN109858371A (en) | The method and device of recognition of face | |
| WO2016084072A1 (en) | Anti-spoofing system and methods useful in conjunction therewith | |
| CN110287889A (en) | A kind of method and device of identification | |
| CN104915673B (en) | A kind of objective classification method and system of view-based access control model bag of words | |
| CN107316029B (en) | A kind of living body verification method and equipment | |
| CN104166841A (en) | Rapid detection identification method for specified pedestrian or vehicle in video monitoring network | |
| CN107194361A (en) | Two-dimentional pose detection method and device | |
| CN109558833A (en) | A kind of face recognition algorithms evaluating method and device | |
| CN108171158A (en) | Biopsy method, device, electronic equipment and storage medium | |
| CN110472460A (en) | Face image processing process and device | |
| CN110826610A (en) | Method and system for intelligently detecting whether dressed clothes of personnel are standard | |
| CN107741996A (en) | Family's map construction method and device based on recognition of face, computing device | |
| CN109993021A (en) | The positive face detecting method of face, device and electronic equipment | |
| CN108875509A (en) | Biopsy method, device and system and storage medium |
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 |
Application publication date: 20190301 |