CN109034048A - Face recognition algorithms models switching method and apparatus - Google Patents
Face recognition algorithms models switching method and apparatus Download PDFInfo
- Publication number
- CN109034048A CN109034048A CN201810805747.2A CN201810805747A CN109034048A CN 109034048 A CN109034048 A CN 109034048A CN 201810805747 A CN201810805747 A CN 201810805747A CN 109034048 A CN109034048 A CN 109034048A
- Authority
- CN
- China
- Prior art keywords
- recognition algorithms
- face recognition
- face
- model
- threshold value
- 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
- G06V40/12—Fingerprints or palmprints
- G06V40/1365—Matching; Classification
-
- 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
Abstract
The present invention provides a kind of face recognition algorithms models switching method and apparatus, are related to protection and monitor field.The face recognition algorithms models switching method and apparatus calculates similarity by the way that the facial image is compared with each of pre-stored face image database face image template according to presetting face recognition algorithms model;If multiple similarities are respectively less than presetting threshold value, i.e. current face recognition algorithms model can not the facial image accurately to current environment scene when identifying, automatically switch presetting face recognition algorithms model, until the one of similarity being calculated is greater than presetting threshold value, so that switching after face recognition algorithms model can the facial image accurately and efficiently to current environment scene identify, to realize the facial image that the moment adaptively extremely can accurately and efficiently to current environment scene by current face recognition algorithms Model Matching.
Description
Technical field
The present invention relates to protection and monitor fields, in particular to a kind of face recognition algorithms models switching method and dress
It sets.
Background technique
Monitoring system is mainly made of headend equipment and rear end equipment this two large divisions, headend equipment usually by video camera,
Manually or electrically the components such as camera lens, holder, shield, monitor, alarm detector and multi-functional decoder form, their each departments
Its duty, and (transmission is contacted by the way that wired, wireless or optical fiber transmission medium are corresponding with the foundation of the various equipment of central control system
Video/audio signal and control, alarm signal).Face recognition and largely to the valuable application of people occur, it is fast and accurate
Object detection system market is also increasingly flourishing.
In the prior art, facial image is captured by video camera, and by the facial image captured and be pre-stored
Database in facial image be compared, to identify whether current facial image carries out blacklist, and in current face
When image is blacklist, progress and alarm.If the moment saves the facial image captured, to be traced in the future.However
The face recognition algorithms model using fixation in the prior art is frequently subjected to environment scene variation (the high field of light intensity
Scape switches to the low scene of light intensity, and indoor scene switches to outdoor scene) and cause the identification to face not quickly smart enough
Really.
Summary of the invention
In view of this, the embodiment of the present invention is designed to provide a kind of face recognition algorithms models switching method and dress
It sets, to improve above-mentioned problem.
In a first aspect, the embodiment of the invention provides a kind of face recognition algorithms models switching method, the recognition of face
Algorithm model switching method includes:
Receive the facial image that image collecting device is sent;
It will be in the facial image and pre-stored face image database according to presetting face recognition algorithms model
Each facial image template be compared, and calculate similarity, wherein multiple face figures in the face image database
As template includes the facial image template of the same face corresponding with the facial image;
Judge whether the multiple similarities being calculated are respectively less than presetting threshold value;
If multiple similarities are respectively less than presetting threshold value, switch presetting face recognition algorithms model, Zhi Daoji
The one of similarity obtained is greater than presetting threshold value.
Second aspect, the embodiment of the invention also provides a kind of face recognition algorithms models switching device, the face is known
Other algorithm model switching device includes:
Information receiving unit, for receiving the facial image of image collecting device transmission;
Computing unit, for according to presetting face recognition algorithms model by the facial image and pre-stored face
Each of image data base face image template is compared, and calculates similarity;
Whether judging unit, multiple similarities for judging to be calculated are respectively less than presetting threshold value;
If algorithm model switch unit switches presetting people be respectively less than presetting threshold value for multiple similarities
Face recognizer model, until the one of similarity being calculated is greater than presetting threshold value.
Compared with prior art, face recognition algorithms models switching method and apparatus provided by the invention, by according to pre-
The face recognition algorithms model of setting is by each of the facial image and pre-stored face image database face image
Template is compared, and calculates similarity;If multiple similarities are respectively less than presetting threshold value, i.e., current recognition of face is calculated
Method model can not the facial image accurately to current environment scene when identifying, automatically switch presetting face
Recognizer model, until the one of similarity being calculated is greater than presetting threshold value, so that the people after switching
Face recognizer model can the facial image accurately and efficiently to current environment scene identify, thus realize the moment from
Facial image that adaptively extremely can accurately and efficiently to current environment scene by current face recognition algorithms Model Matching.
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 make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.The present invention being usually described and illustrated herein in the accompanying drawings is implemented
The component of example can be arranged and be designed with a variety of different configurations.Therefore, below to the reality of the invention provided in the accompanying drawings
The detailed description for applying example is not intended to limit the range of claimed invention, but is merely representative of selected implementation of the invention
Example.Based on the embodiments of the present invention, obtained by those of ordinary skill in the art without making creative efforts
Every other embodiment, shall fall within the protection scope of the present invention.
Fig. 1 is the interaction schematic diagram of image collecting device provided in an embodiment of the present invention and local terminal;
Fig. 2 is the structural block diagram of local terminal provided in an embodiment of the present invention;
Fig. 3 is the process of the first embodiment of face recognition algorithms models switching method provided in an embodiment of the present invention
Figure;
Fig. 4 is the process of second of embodiment of face recognition algorithms models switching method provided in an embodiment of the present invention
Figure;
Fig. 5 is the process of the third embodiment of face recognition algorithms models switching method provided in an embodiment of the present invention
Figure;
Fig. 6 is the functional block diagram of face recognition algorithms models switching device provided in an embodiment of the present invention.
Icon: 100- image collecting device;The local terminal 200-;300- face recognition algorithms models switching device;101-
Memory;102- storage control;103- processor;104- Peripheral Interface;601- information receiving unit;602- computing unit;
603- judging unit;604- algorithm model switch unit.
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.Usually exist
The component of the embodiment of the present invention described and illustrated in attached drawing can be arranged and be designed with a variety of different configurations herein.Cause
This, is not intended to limit claimed invention to the detailed description of the embodiment of the present invention provided in the accompanying drawings below
Range, but it is merely representative of selected embodiment of the invention.Based on the embodiment of the present invention, those skilled in the art are not doing
Every other embodiment obtained under the premise of creative work out, shall fall within the protection scope of the present invention.
Face recognition algorithms models switching apparatus and method provided by present pre-ferred embodiments can be applied to such as Fig. 1 institute
In the application environment shown.As shown in Figure 1, being communicated to connect between image collecting device 100, local terminal 200, image collecting device
100 carry out data interaction with local terminal 200.In the embodiment of the present invention, image collecting device 100 can use video camera.
As shown in Fig. 2, being the functional block diagram of face recognition algorithms models switching device 300 provided by the invention.
Be equipped with the face recognition algorithms models switching device 300 local terminal 200 include memory 101, storage control 102,
Processor 103 and Peripheral Interface 104.Wherein, memory 101, storage control 102, processor 103, Peripheral Interface 104 are each
Element is directly or indirectly electrically connected between each other, to realize the transmission or interaction of data.For example, these elements are mutual
It can be realized and be electrically connected by one or more communication bus or signal wire.The face recognition algorithms models switching device 300
Described can be stored in the memory 101 or is solidificated in including at least one in the form of software or firmware (firmware)
Software function module in ground terminal.The processor 103 is used to execute the executable module stored in memory 101, such as
The software function module or computer program that the face recognition algorithms models switching device 300 includes.
Wherein, memory 101 may be, but not limited to, random access memory (Random Access Memory,
RAM), read-only memory (Read Only Memory, ROM), programmable read only memory (Programmable Read-Only
Memory, PROM), erasable read-only memory (Erasable Programmable Read-Only Memory, EPROM),
Electricallyerasable ROM (EEROM) (Electric Erasable Programmable Read-Only Memory, EEPROM) etc..
Wherein, memory 101 is for storing program, and the processor 103 executes described program after receiving and executing instruction, aforementioned
Method performed by the local terminal that the stream process that any embodiment of the embodiment of the present invention discloses defines can be applied to processor
In 103, or realized by processor 103.
Processor 103 may be a kind of IC chip, the processing capacity with signal.Above-mentioned processor 103 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),
Ready-made programmable gate array (FPGA) either other programmable logic device, discrete gate or transistor logic, discrete hard
Part component.It may be implemented or execute disclosed each method, step and the logic diagram in the embodiment of the present invention.General processor
It can be microprocessor or the processor 103 be also possible to any conventional processor etc..
Various input/output devices are couple processor 103 and memory 101 by the Peripheral Interface 104.Some
In embodiment, Peripheral Interface 104, processor 103 and storage control 102 can be realized in one single chip.Other one
In a little examples, they can be realized by independent chip respectively.
Referring to Fig. 3, the embodiment of the invention provides a kind of face recognition algorithms models switching methods, as the first reality
Mode is applied, the face recognition algorithms models switching method includes:
Step S301: the facial image that image collecting device 100 is sent is received.
Wherein, the environment scene that image collecting device 100 is installed can be indoor or outdoors or area without shade or back sun
Place, or the inlet and outlet biggish position of flow of the people.
Step S302: according to presetting face recognition algorithms model by the facial image and pre-stored facial image
Each of database face image template is compared, and calculates similarity.
For example, it is assumed that having 100 parts of facial image templates in face image database, then one by one by collected facial image
Each facial image template is compared, to calculate 100 similarities, wherein more in the face image database
A facial image template includes the facial image template of the same face corresponding with the facial image.
Step S303: judge whether the multiple similarities being calculated are respectively less than presetting threshold value;If it is, executing
Step S304.
Wherein, presetting threshold value can be 45%, 50%, 55% etc., set with specific reference to actual demand.
Step S304: switching presetting face recognition algorithms model, until the one of similarity being calculated is big
In presetting threshold value.
For example, it is assumed that the face recognition algorithms model being pre-stored in database is followed successively by Elastic Matching in the present embodiment
Algorithm model converts algorithm model, artificial neural network algorithm model, algorithm of support vector machine model and principal component based on KL
Parser model, when the multiple similarities being calculated using Elastic Matching algorithm model are respectively less than presetting threshold value,
Illustrate that Elastic Matching algorithm model and current environment scene mismatch, switches at this time and algorithm model is converted based on KL, work as utilization
When being still respectively less than presetting threshold value based on multiple similarities that KL transformation algorithm model is calculated, artificial neuron is switched to
Network algorithm model, when one of similarity is big in the multiple similarities being calculated using artificial neural network algorithm model
When presetting threshold value, illustrate Elastic Matching algorithm model and current environment scene matching, then artificial neural network algorithm mould
Type identifies the facial image of current environment scene, realizes the adaptive of face recognition algorithms model and current environment scene
It should match, improve the accuracy of facial image identification.
On the basis of the first above-mentioned embodiment, more adapted to more accurately match with current environment scene
Face recognition algorithms model, as second of embodiment, which can also be under
Column step is implemented:
Step S401: the different time points in preheating setting time receive multiple people that image collecting device 100 is sent
Face image.
Step S402: the different time points in preheating setting time calculate according to presetting different recognitions of face respectively
Different facial images is compared method model with each of pre-stored face image database face image template, and
Calculate the corresponding multiple similarities of different face recognition algorithms models, wherein multiple in the face image database
Facial image template includes the facial image template of the same face corresponding with the facial image.
For example, first 5 minutes in 15 minutes, using Elastic Matching algorithm model by collected facial image with prestore
Each of face image database of storage face image template is compared, and it is corresponding to calculate Elastic Matching algorithm model
Multiple similarities;Centre in 15 minutes 5 minutes, using based on KL transformation algorithm model by collected facial image and pre-
Each of face image database of storage face image template is compared, and calculates and convert algorithm model pair based on KL
The multiple similarities answered;Latter 5 minutes in 15 minutes, using artificial neural network algorithm model by collected facial image
It is compared with each of pre-stored face image database face image template, and calculates artificial neural network algorithm
The corresponding multiple similarities of model.
Step S403: judge whether the corresponding multiple similarities of different face recognition algorithms models being calculated are small
In presetting threshold value.
Step S404: if one of phase in the corresponding multiple similarities of at least two face recognition algorithms models
When being greater than presetting threshold value like degree, find out in the corresponding multiple similarities of at least two face recognition algorithms models
Greater than the highest similarity of presetting threshold value, by current face's recognizer models switching to the corresponding face of highest similarity
Recognizer model, and remain unchanged.
In the present embodiment, directly by multiple similarities be respectively less than the corresponding face recognition algorithms model of presetting threshold value into
Row is rejected.For example, when Elastic Matching algorithm model, right respectively based on KL transformation algorithm model, artificial neural network algorithm model
When the one of similarity in multiple similarities answered is all larger than presetting threshold value, for example, presetting threshold value is 50%,
One of similarity in the corresponding multiple similarities of Elastic Matching algorithm model is 60%, converts algorithm model pair based on KL
The one of similarity in multiple similarities answered is in 70%, the corresponding multiple similarities of artificial neural network algorithm model
One of similarity be 80%, then by current face's recognizer models switching to artificial neural network algorithm model, from
And guarantee current face's recognizer model and current environment scene for best match state.It should be noted that the present embodiment
In, face recognition algorithms model, which remains unchanged, only to be referred to when current environment scene does not change, is remained unchanged.
Further, the selected instruction of input can also be received, is greater than in advance according to selected instruction from one of similarity
Face recognition algorithms model is selected at least two face recognition algorithms models of the threshold value of setting, by current face's recognizer
Models switching is remained unchanged to selected face recognition algorithms model.Staff at any time can according to the actual situation, to working as
Preceding face recognition algorithms model is specified.
On the basis of the first above-mentioned embodiment, in order to which the accurate and recognition efficiency of face recognition algorithms model reaches
To balance, as the third embodiment, which can also carry out according to the following steps
Implement:
Step S501: the different time points in preheating setting time receive multiple people that image collecting device 100 is sent
Face image.
Step S502: the different time points in preheating setting time calculate according to presetting different recognitions of face respectively
Different facial images is compared method model with each of pre-stored face image database face image template, and
Calculate the corresponding multiple similarities of different face recognition algorithms models, wherein multiple in the face image database
Facial image template includes the facial image template of the same face corresponding with the facial image.
Step S503: the similarity result of the facial image quantity and output that are obtained according to each face recognition algorithms model
Calculate the identification completion rate of different face recognition algorithms models.
Step S504: judge whether the corresponding multiple similarities of different face recognition algorithms models being calculated are small
In presetting threshold value.
Step S505: if one of phase in the corresponding multiple similarities of at least two face recognition algorithms models
When being greater than presetting threshold value like degree, current face's identification module is switched into face at least two face recognition algorithms models
The highest face recognition algorithms model of the identification completion rate of image, and remain unchanged.
When image collecting device 100 is in the more position of flow of the people, identification will not only be met to facial image identification
Accurately, and there is sufficiently high recognition efficiency, leakage is avoided to identify as far as possible, therefore, in the present embodiment, when there are multiple recognitions of face
When algorithm model meets one of similarity in corresponding multiple similarities greater than presetting threshold value, selection identification
The highest face recognition algorithms model of completion rate, to have higher recognition efficiency in the case where guaranteeing identification accuracy.
Referring to Fig. 6, needing to illustrate the embodiment of the invention provides a kind of face recognition algorithms models switching device
It is face recognition algorithms models switching device provided by the present embodiment, the technical effect of basic principle and generation and above-mentioned
Method for detecting human face provided by embodiment is identical, and to briefly describe, the present embodiment part does not refer to place, can refer to above-mentioned
Corresponding contents in embodiment.The face recognition algorithms models switching device include information receiving unit 601, computing unit 602,
Judging unit 603 and algorithm model switch unit 604.
As the first embodiment:
Information receiving unit 601 is used to receive the facial image of the transmission of image collecting device 100.
Computing unit 602 is used for the facial image and pre-stored people according to presetting face recognition algorithms model
Each of face image database face image template is compared, and calculates similarity, wherein the face image database
In multiple facial image templates include the same face corresponding with the facial image facial image template.
Whether multiple similarities that judging unit 603 is used to judge to be calculated are respectively less than presetting threshold value.
If algorithm model switch unit 604 is respectively less than presetting threshold value for multiple similarities, switch presetting
Face recognition algorithms model, until the one of similarity being calculated is greater than presetting threshold value.
On the basis of the first above-mentioned embodiment, more adapted to more accurately match with current environment scene
Face recognition algorithms model, as second of embodiment, which can also be under
Column mode is implemented:
Information receiving unit 601 is specifically used for the different time points reception image collecting devices in preheating setting time
The 100 multiple facial images sent.
Computing unit 602 be specifically used for respectively the different time points in preheating setting time according to presetting different
Face recognition algorithms model is by each of different facial image and pre-stored face image database face image template
It is compared, and calculates the corresponding multiple similarities of different face recognition algorithms models, wherein the face image data
Multiple facial image templates in library include the facial image template of the same face corresponding with the facial image.
It is corresponding multiple similar that judging unit 603 is specifically used for the different face recognition algorithms models that judgement is calculated
Whether degree is respectively less than presetting threshold value.
If algorithm model switch unit 604 is specifically used for the corresponding multiple phases of at least two face recognition algorithms models
When being greater than presetting threshold value like one of similarity in degree, it is right respectively to find out at least two face recognition algorithms models
The highest similarity greater than presetting threshold value in the multiple similarities answered, by current face's recognizer models switching to most
The corresponding face recognition algorithms model of high similarity, and remain unchanged;
Or receive the selected instruction of input, according to selected instruction from one of similarity be greater than presetting threshold value to
Face recognition algorithms model is selected in few two kinds of face recognition algorithms models, by current face's recognizer models switching to selected
Face recognition algorithms model, and remain unchanged.
On the basis of the first above-mentioned embodiment, in order to which the accurate and recognition efficiency of face recognition algorithms model reaches
To balance, as the third embodiment, which can also carry out according to the following steps
Implement:
Information receiving unit 601 is specifically used for the different time points reception image collecting devices in preheating setting time
The 100 multiple facial images sent.
Computing unit 602 be specifically used for respectively the different time points in preheating setting time according to presetting different
Face recognition algorithms model is by each of different facial image and pre-stored face image database face image template
It is compared, and calculates the corresponding multiple similarities of different face recognition algorithms models, wherein the face image data
Multiple facial image templates in library include the facial image template of the same face corresponding with the facial image.
Wherein, face recognition algorithms model can include but is not limited to Elastic Matching algorithm model, convert algorithm based on KL
Model, artificial neural network algorithm model, algorithm of support vector machine model and Principal Component Analysis Algorithm model.
The phase of facial image quantity and output that computing unit 602 is also used to obtain according to each face recognition algorithms model
The identification completion rate of different face recognition algorithms models is calculated like degree result.
It is corresponding multiple similar that judging unit 603 is specifically used for the different face recognition algorithms models that judgement is calculated
Whether degree is respectively less than presetting threshold value.
If algorithm model switch unit 604 is specifically used for the corresponding multiple phases of at least two face recognition algorithms models
When being greater than presetting threshold value like one of similarity in degree, current face's identification module is switched at least two faces
The highest face recognition algorithms model of the identification completion rate of facial image in recognizer model, and remain unchanged.
In conclusion face recognition algorithms models switching method and apparatus provided by the invention, by according to presetting
Face recognition algorithms model by each of the facial image and pre-stored face image database face image template into
Row compares, and calculates similarity;If multiple similarities are respectively less than presetting threshold value, i.e., current face recognition algorithms model
Can not the facial image accurately to current environment scene when identifying, automatically switch presetting recognition of face and calculate
Method model, until the one of similarity being calculated is greater than presetting threshold value, so that the recognition of face after switching
Algorithm model can the facial image accurately and efficiently to current environment scene identify, to realize the moment adaptively
Facial image that extremely can accurately and efficiently to current environment scene by current face recognition algorithms Model Matching.
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 model, program segment or code
Part, a part of the model, program segment 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 mode in each embodiment of the present invention can integrate one independent portion of formation together
Point, it is also possible to each model individualism, an independent part can also be formed with two or more model integrateds.
It, can be with if the function is realized and when sold or used as an independent product in the form of software function model
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, local terminal or network equipment etc.) execute all or part of step of each embodiment the method for the present invention
Suddenly.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), deposits at random
The various media that can store program code such as access to memory (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 an entity or behaviour
Make with another entity or operate distinguish, without necessarily requiring or implying between these entities or operation there are it is any this
The actual relationship of kind or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended to nonexcludability
Include so that include a series of elements process, method, article or equipment not only include those elements, but also
Including other elements that are not explicitly listed, or further include for this process, method, article or equipment intrinsic want
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including described want
There is also other identical elements in the process, method, article or equipment of element.
The foregoing is only a preferred embodiment of the present invention, is not intended to restrict the invention, for the skill of this field
For art personnel, the invention may be variously modified and varied.All within the spirits and principles of the present invention, made any to repair
Change, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.It should also be noted that similar label and letter exist
Similar terms are indicated in following attached drawing, therefore, once being defined in a certain Xiang Yi attached drawing, are then not required in subsequent attached drawing
It is further defined and explained.
Claims (10)
1. a kind of face recognition algorithms models switching method, which is characterized in that the face recognition algorithms models switching method packet
It includes:
Receive the facial image that image collecting device is sent;
According to presetting face recognition algorithms model by the facial image with it is every in pre-stored face image database
A facial image template is compared, and calculates similarity, wherein multiple facial image moulds in the face image database
Plate includes the facial image template of the same face corresponding with the facial image;
Judge whether the multiple similarities being calculated are respectively less than presetting threshold value;
If multiple similarities are respectively less than presetting threshold value, switch presetting face recognition algorithms model, until calculating
One of similarity out is greater than presetting threshold value.
2. face recognition algorithms models switching method according to claim 1, which is characterized in that
Described the step of receiving the facial image that image collecting device is sent includes: the different time points in preheating setting time
Receive multiple facial images that image collecting device is sent;
It is described will be in the facial image and pre-stored face image database according to presetting face recognition algorithms model
Each facial image template be compared, and the step of calculating similarity includes: different in preheating setting time respectively
Time point is according to presetting different face recognition algorithms models by different facial image and pre-stored facial image number
It is compared according to each of library face image template, and it is corresponding multiple similar to calculate different face recognition algorithms models
Degree;
It is described to judge that the step of whether multiple similarities being calculated are respectively less than presetting threshold value includes: that judgement is calculated
The corresponding multiple similarities of different face recognition algorithms models whether be respectively less than presetting threshold value;
If multiple similarities are respectively less than presetting threshold value, switch presetting face recognition algorithms model, Zhi Daoji
The one of similarity obtained be greater than presetting threshold value the step of include:
If one of similarity in the corresponding multiple similarities of at least two face recognition algorithms models is greater than default
When fixed threshold value, find out in the corresponding multiple similarities of at least two face recognition algorithms models be greater than it is presetting
The highest similarity of threshold value, by current face's recognizer models switching to the corresponding face recognition algorithms mould of highest similarity
Type, and remain unchanged.
3. face recognition algorithms models switching method according to claim 2, which is characterized in that
If one of similarity in the corresponding multiple similarities of at least two face recognition algorithms models is greater than default
When fixed threshold value,
It finds out in the corresponding multiple similarities of at least two face recognition algorithms models and is greater than presetting threshold value
Highest similarity by current face's recognizer models switching to the corresponding face recognition algorithms model of highest similarity, and is protected
It holds constant;
Or the selected instruction of input is received, it is greater than at least the two of presetting threshold value from one of similarity according to selected instruction
Face recognition algorithms model is selected in kind face recognition algorithms model, by current face's recognizer models switching to selected people
Face recognizer model, and remain unchanged.
4. face recognition algorithms models switching method according to claim 1, which is characterized in that
Described the step of receiving the facial image that image collecting device is sent includes: the different time points in preheating setting time
Receive multiple facial images that image collecting device is sent;
It is described will be in the facial image and pre-stored face image database according to presetting face recognition algorithms model
Each facial image template be compared, and the step of calculating similarity includes: different in preheating setting time respectively
Time point is according to presetting different face recognition algorithms models by different facial image and pre-stored facial image number
It is compared according to each of library face image template, and it is corresponding multiple similar to calculate different face recognition algorithms models
Degree;
The face recognition algorithms models switching method further include: the facial image obtained according to each face recognition algorithms model
Quantity and the similarity result of output calculate the identification completion rate of different face recognition algorithms models;
It is described to judge that the step of whether multiple similarities being calculated are respectively less than presetting threshold value includes: that judgement is calculated
The corresponding multiple similarities of different face recognition algorithms models whether be respectively less than presetting threshold value;
If multiple similarities are respectively less than presetting threshold value, switch presetting face recognition algorithms model, Zhi Daoji
The one of similarity obtained be greater than presetting threshold value the step of include:
If one of similarity in the corresponding multiple similarities of at least two face recognition algorithms models is greater than default
When fixed threshold value, the identification that current face's identification module is switched to facial image at least two face recognition algorithms models is complete
At the highest face recognition algorithms model of rate, and remain unchanged.
5. face recognition algorithms models switching method according to claim 1, which is characterized in that the face recognition algorithms
Model includes Elastic Matching algorithm model, based on KL transformation algorithm model, artificial neural network algorithm model, support vector machines calculation
Method model and Principal Component Analysis Algorithm model.
6. a kind of face recognition algorithms models switching device, which is characterized in that the face recognition algorithms models switching device packet
It includes:
Information receiving unit, for receiving the facial image of image collecting device transmission;
Computing unit, for according to presetting face recognition algorithms model by the facial image and pre-stored facial image
Each of database face image template is compared, and calculates similarity, wherein in the face image database
Multiple facial image templates include the facial image template of the same face corresponding with the facial image;
Whether judging unit, multiple similarities for judging to be calculated are respectively less than presetting threshold value;
If algorithm model switch unit switches presetting face and knows be respectively less than presetting threshold value for multiple similarities
Other algorithm model, until the one of similarity being calculated is greater than presetting threshold value.
7. face recognition algorithms models switching device according to claim 6, which is characterized in that
The information receiving unit is specifically used for the different time points reception image collecting devices in preheating setting time and sends
Multiple facial images;
The computing unit is specifically for the different time points in preheating setting time respectively according to presetting different people
Face recognizer model by each of different facial image and pre-stored face image database face image template into
Row compares, and calculates the corresponding multiple similarities of different face recognition algorithms models;
The judging unit is specifically used for the corresponding multiple similarities of different face recognition algorithms models that judgement is calculated
Whether presetting threshold value is respectively less than;
If it is corresponding multiple similar that the algorithm model switch unit is specifically used at least two face recognition algorithms models
When one of similarity in degree is greater than presetting threshold value, finds out at least two face recognition algorithms models and respectively correspond
Multiple similarities in the highest similarity greater than presetting threshold value, by current face's recognizer models switching to highest
The corresponding face recognition algorithms model of similarity, and remain unchanged.
8. face recognition algorithms models switching device according to claim 7, which is characterized in that
If it is corresponding multiple similar that the algorithm model switch unit is specifically used at least two face recognition algorithms models
When one of similarity in degree is greater than presetting threshold value,
It finds out in the corresponding multiple similarities of at least two face recognition algorithms models and is greater than presetting threshold value
Highest similarity by current face's recognizer models switching to the corresponding face recognition algorithms model of highest similarity, and is protected
It holds constant;
Or the selected instruction of input is received, it is greater than at least the two of presetting threshold value from one of similarity according to selected instruction
Face recognition algorithms model is selected in kind face recognition algorithms model, by current face's recognizer models switching to selected people
Face recognizer model, and remain unchanged.
9. face recognition algorithms models switching device according to claim 6, which is characterized in that
The information receiving unit is specifically used for the different time points reception image collecting devices in preheating setting time and sends
Multiple facial images;
The computing unit is specifically for the different time points in preheating setting time respectively according to presetting different people
Face recognizer model by each of different facial image and pre-stored face image database face image template into
Row compares, and calculates the corresponding multiple similarities of different face recognition algorithms models;
The computing unit be also used to according to each face recognition algorithms model obtain facial image quantity and output it is similar
Degree result calculates the identification completion rate of different face recognition algorithms models;
The judging unit is specifically used for the corresponding multiple similarities of different face recognition algorithms models that judgement is calculated
Whether presetting threshold value is respectively less than;
If it is corresponding multiple similar that the algorithm model switch unit is specifically used at least two face recognition algorithms models
When one of similarity in degree is greater than presetting threshold value, current face's identification module is switched at least two faces and is known
The highest face recognition algorithms model of the identification completion rate of facial image in other algorithm model, and remain unchanged.
10. face recognition algorithms models switching device according to claim 6, which is characterized in that the recognition of face is calculated
Method model includes Elastic Matching algorithm model, based on KL transformation algorithm model, artificial neural network algorithm model, support vector machines
Algorithm model and Principal Component Analysis Algorithm model.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810805747.2A CN109034048A (en) | 2018-07-20 | 2018-07-20 | Face recognition algorithms models switching method and apparatus |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810805747.2A CN109034048A (en) | 2018-07-20 | 2018-07-20 | Face recognition algorithms models switching method and apparatus |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109034048A true CN109034048A (en) | 2018-12-18 |
Family
ID=64644007
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810805747.2A Pending CN109034048A (en) | 2018-07-20 | 2018-07-20 | Face recognition algorithms models switching method and apparatus |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109034048A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110070037A (en) * | 2019-04-22 | 2019-07-30 | 深圳力维智联技术有限公司 | Smooth upgrading method, device and the readable storage medium storing program for executing of human face recognition model |
CN110689655A (en) * | 2019-09-17 | 2020-01-14 | 南昌大学 | Dormitory entrance guard management integration system based on face identification |
CN111091080A (en) * | 2019-12-06 | 2020-05-01 | 贵州电网有限责任公司 | Face recognition method and system |
CN111582305A (en) * | 2020-03-26 | 2020-08-25 | 平安科技(深圳)有限公司 | Biological feature recognition method and device, computer equipment and storage medium |
CN111723609A (en) * | 2019-03-20 | 2020-09-29 | 杭州海康威视数字技术股份有限公司 | Model optimization method and device, electronic equipment and storage medium |
CN111797868A (en) * | 2019-04-09 | 2020-10-20 | Oppo广东移动通信有限公司 | Scene recognition model modeling method and device, storage medium and electronic equipment |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101142586A (en) * | 2005-03-18 | 2008-03-12 | 皇家飞利浦电子股份有限公司 | Method of performing face recognition |
CN105117724A (en) * | 2015-07-30 | 2015-12-02 | 北京邮电大学 | License plate positioning method and apparatus |
CN105354307A (en) * | 2015-11-06 | 2016-02-24 | 腾讯科技(深圳)有限公司 | Image content identification method and apparatus |
CN107301402A (en) * | 2017-06-30 | 2017-10-27 | 锐捷网络股份有限公司 | A kind of determination method, device, medium and the equipment of reality scene key frame |
CN108038948A (en) * | 2017-12-26 | 2018-05-15 | 杭州数梦工场科技有限公司 | Verification method and device, the computer-readable recording medium of passenger identity |
CN108052864A (en) * | 2017-11-17 | 2018-05-18 | 平安科技(深圳)有限公司 | Face identification method, application server and computer readable storage medium |
-
2018
- 2018-07-20 CN CN201810805747.2A patent/CN109034048A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101142586A (en) * | 2005-03-18 | 2008-03-12 | 皇家飞利浦电子股份有限公司 | Method of performing face recognition |
CN105117724A (en) * | 2015-07-30 | 2015-12-02 | 北京邮电大学 | License plate positioning method and apparatus |
CN105354307A (en) * | 2015-11-06 | 2016-02-24 | 腾讯科技(深圳)有限公司 | Image content identification method and apparatus |
CN107301402A (en) * | 2017-06-30 | 2017-10-27 | 锐捷网络股份有限公司 | A kind of determination method, device, medium and the equipment of reality scene key frame |
CN108052864A (en) * | 2017-11-17 | 2018-05-18 | 平安科技(深圳)有限公司 | Face identification method, application server and computer readable storage medium |
CN108038948A (en) * | 2017-12-26 | 2018-05-15 | 杭州数梦工场科技有限公司 | Verification method and device, the computer-readable recording medium of passenger identity |
Non-Patent Citations (6)
Title |
---|
FABIEN CARDINAUX等: "Face Verification Using Adapted Generative Models", 《SIXTH IEEE INTERNATIONAL CONFERENCE ON AUTOMATIC FACE AND GESTURE RECOGNITION》, pages 1 - 12 * |
KUNCHEVA等: "Switching Between Selection and Fusion in", 《IEEE TRANSACTIONS ON SYSTEMS, MAN AND CYBERNETICS》, vol. 32, pages 146 - 156 * |
XI CHENG ET AL.: "Deep similarity learning for multimodal medical images", COMPUTER METHODS IN BIOMECHANICS AND BIOMEDICAL ENGINEERING-IMAGING AND VISUALIZATION, vol. 06, no. 03, pages 248 - 252 * |
中国石油天然气总公司劳资局: "分布式网络系统与Multi-Agent系统编程框架", 哈尔滨:哈尔滨工业大学出版社, pages: 215 - 216 * |
乔杉: "基于BP神经网络的人脸识别方法研究", 《中国优秀硕士学位论文全文数据库 (信息科技辑)》, no. 9, pages 138 - 1090 * |
郑轶: "基于多分类器的人脸识别算法的研究与实现", 《中国优秀硕士学位论文全文数据库 (信息科技辑)》, no. 3, pages 138 - 2090 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111723609A (en) * | 2019-03-20 | 2020-09-29 | 杭州海康威视数字技术股份有限公司 | Model optimization method and device, electronic equipment and storage medium |
CN111723609B (en) * | 2019-03-20 | 2024-01-26 | 杭州海康威视数字技术股份有限公司 | Model optimization method, device, electronic equipment and storage medium |
CN111797868A (en) * | 2019-04-09 | 2020-10-20 | Oppo广东移动通信有限公司 | Scene recognition model modeling method and device, storage medium and electronic equipment |
CN110070037A (en) * | 2019-04-22 | 2019-07-30 | 深圳力维智联技术有限公司 | Smooth upgrading method, device and the readable storage medium storing program for executing of human face recognition model |
CN110070037B (en) * | 2019-04-22 | 2022-11-01 | 深圳力维智联技术有限公司 | Smooth upgrading method and device for face recognition model and readable storage medium |
CN110689655A (en) * | 2019-09-17 | 2020-01-14 | 南昌大学 | Dormitory entrance guard management integration system based on face identification |
CN111091080A (en) * | 2019-12-06 | 2020-05-01 | 贵州电网有限责任公司 | Face recognition method and system |
CN111582305A (en) * | 2020-03-26 | 2020-08-25 | 平安科技(深圳)有限公司 | Biological feature recognition method and device, computer equipment and storage medium |
CN111582305B (en) * | 2020-03-26 | 2023-08-18 | 平安科技(深圳)有限公司 | Biological feature recognition method, apparatus, computer device and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109034048A (en) | Face recognition algorithms models switching method and apparatus | |
CN110147717B (en) | Human body action recognition method and device | |
CN112381075B (en) | Method and system for carrying out face recognition under specific scene of machine room | |
CN109166156B (en) | Camera calibration image generation method, mobile terminal and storage medium | |
US20110103695A1 (en) | Image processing apparatus and image processing method | |
CN109116129B (en) | Terminal detection method, detection device, system and storage medium | |
CN110263603B (en) | Face recognition method and device based on central loss and residual error visual simulation network | |
CN110309060B (en) | Detection method and device for updating identification algorithm, storage medium and computer equipment | |
CN110968712B (en) | BIM family model acquisition method, device and system | |
CN114943936B (en) | Target behavior recognition method and device, electronic equipment and storage medium | |
CN115410275A (en) | Office place personnel state detection method and system based on image recognition | |
CN112884147A (en) | Neural network training method, image processing method, device and electronic equipment | |
US10803353B2 (en) | Method and apparatus for acquiring information | |
CN110909612A (en) | Gait recognition method and system based on deep neural network and machine vision | |
CN111368867A (en) | Archive classification method and system and computer readable storage medium | |
Rodríguez et al. | HD-MR: A new algorithm for number recognition in electrical meters | |
CN108875445B (en) | Pedestrian re-identification method and device | |
CN105787968A (en) | Polygon convex/concave point identification method and device | |
CN113780145A (en) | Sperm morphology detection method, sperm morphology detection device, computer equipment and storage medium | |
CN106651803B (en) | Method and device for identifying house type data | |
CN111460419A (en) | Internet of things artificial intelligence face verification method and Internet of things cloud server | |
CN111291749A (en) | Gesture recognition method and device and robot | |
CN108831158A (en) | It disobeys and stops monitoring method, device and electric terminal | |
CN115147752A (en) | Video analysis method and device and computer equipment | |
CN112819859B (en) | Multi-target tracking method and device applied to intelligent security |
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 | ||
CB03 | Change of inventor or designer information | ||
CB03 | Change of inventor or designer information |
Inventor after: Ning Hao Inventor after: Xu Qing Inventor after: Zhang Lei Inventor after: Zheng Bo Inventor before: Ning Hao Inventor before: Xu Qing Inventor before: Zhang Lei Inventor before: Deng Bo |