CN110414444A - Face identification method and device - Google Patents
Face identification method and device Download PDFInfo
- Publication number
- CN110414444A CN110414444A CN201910701882.7A CN201910701882A CN110414444A CN 110414444 A CN110414444 A CN 110414444A CN 201910701882 A CN201910701882 A CN 201910701882A CN 110414444 A CN110414444 A CN 110414444A
- Authority
- CN
- China
- Prior art keywords
- frame data
- characteristic value
- face
- data
- frame
- 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
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/94—Hardware or software architectures specially adapted for image or video understanding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of face identification method and device, face identification method includes: to generate characteristic value corresponding to the frame data according to every frame data in the human face data obtained in advance；Face recognition result is generated according to the characteristic value of the previous frame data of preset eigenvalue threshold, the characteristic value of the frame data and the frame data.Method provided by the invention, which overcomes face snap technology in the prior art, can not combine the problem of capturing real-time and capturing effect.
Description
Technical field
The present invention relates to object detection technique fields, more particularly, to a kind of face identification method and device.
Background technique
Universal network camera (IPC) does not have the AI function such as Face datection, however universal network camera is acquired
Video flowing transfer to Face datection algorithm process by frame, can be achieved with the function of face snap.By universal network camera and people
The way that face detection algorithm combines, the ability of face snap is imparted to a large amount of universal network cameras, face snap can be wide
It is general to be used for security protection, want the fields such as visitor's identification.
The method of existing face snap generally comprises the critical workflows such as Face datection, picture quality detection；It generally comprises fast
Speed capture, interval frame capture, interval the second capture, leave after capture isotype.Since Face datection, picture quality detect these passes
Key step is all computationally intensive operation, and existing grasp shoot method is caused no matter to use which kind of mode, simultaneous while all can not be preferably
It cares for and captures real-time and capture effect, and its Function Coupling, be unfavorable for extending.
Summary of the invention
For the problems of the prior art, the present invention provides one kind can overcome in the prior art face snap technology without
Method combines the face identification method captured real-time and capture effect problem.
In order to solve the above technical problems, the present invention the following technical schemes are provided:
In a first aspect, the present invention provides a kind of face identification method, comprising:
Characteristic value corresponding to the frame data is generated according to every frame data in the human face data obtained in advance；
According to the spy of the previous frame data of preset eigenvalue threshold, the characteristic value of the frame data and the frame data
Value indicative generates face recognition result.
Preferably, face identification method, further includes:
It is connect by establishing rtsp with camera, obtains the human face data.
Preferably, every frame data according in the human face data obtained in advance generate spy corresponding to the frame data
Value indicative, comprising:
According to every frame data vector quantization in the human face data, one-dimensional vector is generated；
According to the one-dimensional vector generate the frame data corresponding to characteristic value.
Preferably, described according to the upper of preset eigenvalue threshold, the characteristic value of the frame data and the frame data
The characteristic value of one frame data generates face recognition result, comprising:
Compare the characteristic value of the frame data and the size of the eigenvalue threshold；
When the characteristic value of the frame data is greater than the eigenvalue threshold, by the characteristic value of the frame data and the frame data
The characteristic values of previous frame data do difference；
Recognition of face is carried out according to difference result.
Second aspect, the present invention provide a kind of face identification device, which includes:
Characteristic value generation unit, for generating the frame data institute according to every frame data in the human face data obtained in advance
Corresponding characteristic value；
As a result generation unit, for according to preset eigenvalue threshold, the frame data characteristic value and the frame number
According to previous frame data characteristic value generate face recognition result.
Preferably, face identification device further include:
Human face data acquiring unit obtains the human face data for connecting by establishing rtsp with camera.
Preferably, characteristic value generation unit includes:
One-dimensional vector generation module, for generating one-dimensional vector according to every frame data vector quantization in the human face data；
Characteristic value generation module, for according to the one-dimensional vector generate the frame data corresponding to characteristic value.
Preferably, as a result generation unit includes:
Comparison module, for the characteristic value of the frame data and the size of the eigenvalue threshold；
Difference block, for when the characteristic value of the frame data be greater than the eigenvalue threshold when, by the spy of the frame data
The characteristic value of value indicative and the previous frame data of the frame data does difference；
Identification module, for carrying out recognition of face according to difference result.
The third aspect, the present invention provides a kind of electronic equipment, including memory, processor and storage are on a memory and can
The step of computer program run on a processor, processor realizes face identification method when executing program.
Fourth aspect, the present invention provide a kind of computer readable storage medium, are stored thereon with computer program, the calculating
The step of face identification method is realized when machine program is executed by processor.
As can be seen from the above description, face identification method provided by the invention and device, according to the human face data obtained in advance
In every frame data generate characteristic value corresponding to the frame data, and by obtained characteristic value and preset eigenvalue threshold
It makes comparisons, a characteristic value collection is generated according to the characteristic value for being greater than eigenvalue threshold, then, by this feature value set and previous frame
The characteristic value collection of data does difference, finally obtains face recognition result.This method overcomes face snap skill in the prior art
Art can not combine the problem of capturing real-time and capturing effect.Specific advantage is as follows:
1) the face snap method for using frame difference, is truly realized real-time grasp shoot while meeting face quality requirement
Effect.It is traditional quickly capture, leave after the method captured require whithin a period of time comparison chart as mass effect, this side
Formula had not only wasted computing resource but also had been unable to reach the effect of real-time grasp shoot, and traditional interval frame is captured, the interval second captures can not
Guarantee that the quality captured or its implementation are excessively complicated, be not suitable for this computation-intensive of face snap and require to guarantee
The scene of real-time.
2) process provides a kind of flexible face snap system structures, so that face snap set expandability is extremely strong.
There are serious Function Couplings for the system authority of traditional grasp shoot method, are unfavorable for the landing and extension of face snap method.The present invention
The system structure of offer from getting frame, saves all decouplings of the several core apparatus of frame, Face datection, face duplicate removal, different performance
Machine can cooperate, not only facilitate horizontal behavior extension, but also facilitate Function Extension, the support of hot plug be provided.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is the present invention
Some embodiments for those of ordinary skill in the art without creative efforts, can also basis
These attached drawings obtain other attached drawings.
Fig. 1 is face identification method flow diagram one in the embodiment of the present invention；
Fig. 2 is face identification method flow diagram two in the embodiment of the present invention；
Fig. 3 is the flow diagram of face identification method step 100 in the embodiment of the present invention；
Fig. 4 is the flow diagram of face identification method step 200 in the embodiment of the present invention；
Fig. 5 is the flow diagram one of face identification method in specific application example of the invention；
Fig. 6 is the flow diagram two of face identification method in specific application example of the invention；
Fig. 7 is the structural schematic diagram one of face identification device in specific application example of the invention；
Fig. 8 is the structural schematic diagram two of face identification device in specific application example of the invention；
Fig. 9 is the structural schematic diagram of characteristic value generation unit in specific application example of the invention；
Figure 10 is the structural schematic diagram of result generation unit in specific application example of the invention；
Figure 11 is the structural schematic diagram of the electronic equipment in the embodiment of the present invention.
Specific embodiment
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, the 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.Based on the embodiments of the present invention, those of ordinary skill in the art
Every other embodiment obtained without creative efforts, shall fall within the protection scope of the present invention.
In view of the related needs that the safety and ease for use that promote current verification code system exists in the prior art, the present invention
Embodiment a kind of specific embodiment of face identification method is provided, referring to Fig. 1 this method10Specifically include following content:
Step 100: generating feature corresponding to the frame data according to every frame data in the human face data obtained in advance
Value.
It is understood that frame is the single width image frame of minimum unit in image animation, it is equivalent on cinefilm
Each lattice camera lens.As soon as a frame is exactly secondary static picture, continuous frame forms animation, such as TV image.Briefly,
It is the frame number of the picture transmitted in 1 seconds, it is understood that for that can refresh several times, usually graphics processor each second
It is indicated with fps (Frames Per Second).Each frame is all static image, in extremely rapid succession shows that frame just forms fortune
Dynamic illusion.Available more smooth, the more true to nature animation of high frame per second.Frame number each second (fps) the more, shown movement
It will be more smooth.
Step 200: according to the previous frame of preset eigenvalue threshold, the characteristic value of the frame data and the frame data
The characteristic value of data generates face recognition result.
Specifically, the characteristic value collection of certain frame and previous frame are done into difference operation, obtained result set is to capture to arrive
Face.It is understood that difference operation is more in line with actual face snap scene, when camera perceive present frame with
Previous frame has the similarities and differences, and difference operation immediately respond this similarities and differences, without just doing after several seconds or number frame
It responds out.
As can be seen from the above description, face identification method provided by the invention, according to every in the human face data obtained in advance
Frame data generate characteristic value corresponding to the frame data, and obtained characteristic value and preset eigenvalue threshold are made ratio
Compared with according to characteristic value one characteristic value collection of generation for being greater than eigenvalue threshold, then, by this feature value set and previous frame data
Characteristic value collection do difference, finally obtain face recognition result.This method overcome in the prior art face snap technology without
Method combines the problem of capturing real-time and capturing effect.
In one embodiment, referring to fig. 2, face identification method further include:
Step 90: being connect by establishing rtsp with camera, obtain the human face data.
RTSP (Real Time Streaming Protocol), real time streaming transport protocol is in ICP/IP protocol system
An application layer protocol.The protocol define one-to-many applications how effectively to transmit multimedia number by IP network
According to.RTSP is architecturally located on RTP and RTCP, it completes data transmission using TCP or UDP.HTTP and RTSP phase
Than HTTP request is issued by client computer, and server makes a response；When using RTSP, client-server can be issued and be asked
It asks, i.e., RTSP can be two-way.RTSP is the multimedia series flow agreement for controlling sound or image, and is allowed multiple simultaneously
Crossfire demand control, the network communication used of when transmission are reached an agreement on not in the range of its definition, and server end can be selected voluntarily
It selects using TCP or UDP and transmits streamed content, do not emphasize time synchronization especially, it can tolerant network delay so comparing.
In one embodiment, referring to Fig. 3, step 100 includes:
Step 101: according to every frame data vector quantization in the human face data, generating one-dimensional vector.
Specifically, the face picture vector in every frame data in human face data is melted into an one-dimensional vector.
Step 102: according to the one-dimensional vector generate the frame data corresponding to characteristic value.
It is understood that the characteristic value that the one-dimensional vector in step 101 is the face indicates.
In one embodiment, referring to fig. 4, step 200 includes:
Step 201: the size of the characteristic value of the frame data and the eigenvalue threshold.
Step 201 is when implementing, when measuring the characteristic value of certain frame data of human face data greater than preset eigenvalue threshold,
Illustrate that the face has met the standard of recognition of face, execute step 202 at once, without traversing defined time or rule
After all faces in fixed number of frames, reselection one is greater than the human face data of eigenvalue threshold.If measuring human face data
Certain frame data characteristic value be less than preset eigenvalue threshold when, illustrate that human face data quality inspection is unqualified, discard the face number
According to.
Step 202: when the characteristic value of the frame data be greater than the eigenvalue threshold when, by the characteristic value of the frame data with
The characteristic value of the previous frame data of the frame data does difference.
It is understood that video camera acquisition video sequence have the characteristics that it is successional.If do not moved in scene
Target, then the variation of successive frame is very faint, if there is moving target, then has between continuous frame and frame and significantly changes.
Since the target in scene is moving, position of the image of target in different images frame is different.To time upper continuous two frame
Or three frame image carry out calculus of differences, the corresponding pixel of different frame subtracts each other, and gray scale absolute value of the difference is judged, when absolute value is more than
When certain threshold value, it can be judged as moving target, to realize the detection function of target.
Step 202 specifically, after acquiring previous frame data (image) and current data, first carries out ash to it when implementing
Degreeization processing.Then front and back, which makes the difference, takes absolute value, and the new gray level image of gained twice is carried out mutually and by differentiated image two
Edge detection is carried out after value.
Step 203: recognition of face is carried out according to difference result.
As can be seen from the above description, face identification method provided by the invention, according to every in the human face data obtained in advance
Frame data generate characteristic value corresponding to the frame data, and obtained characteristic value and preset eigenvalue threshold are made ratio
Compared with according to characteristic value one characteristic value collection of generation for being greater than eigenvalue threshold, then, by this feature value set and previous frame data
Characteristic value collection do difference, finally obtain face recognition result.This method overcome in the prior art face snap technology without
Method combines the problem of capturing real-time and capturing effect.Specific advantage is as follows:
1) the face snap method for using frame difference, is truly realized real-time grasp shoot while meeting face quality requirement
Effect.It is traditional quickly capture, leave after the method captured require whithin a period of time comparison chart as mass effect, this side
Formula had not only wasted computing resource but also had been unable to reach the effect of real-time grasp shoot, and traditional interval frame is captured, the interval second captures can not
Guarantee that the quality captured or its implementation are excessively complicated, be not suitable for this computation-intensive of face snap and require to guarantee
The scene of real-time.
2) process provides a kind of flexible face snap system structures, so that face snap set expandability is extremely strong.
There are serious Function Couplings for the system authority of traditional grasp shoot method, are unfavorable for the landing and extension of face snap method.The present invention
The system structure of offer from getting frame, saves all decouplings of the several core apparatus of frame, Face datection, face duplicate removal, different performance
Machine can cooperate, not only facilitate horizontal behavior extension, but also facilitate Function Extension, the support of hot plug be provided.
To further explain this programme, the present invention provides the specific application example of face identification method, the concrete application
Example specifically includes following content20, referring to Fig. 5 and Fig. 6.
Step S101: universal network camera establishes rtsp connection, and connection is put into connection pool.
Step S102: circulation takes out a connection from connection pool, reads a frame image data by the connection.
Particularly, when rtsp connection accidental interruption, cause to read frame failure one rtsp connection reconnection thread of unlatching immediately
Task, the connection of horse back reconnection failure.If connection is put into connection pool by successful connection；If it fails to connect, then postponing
2n(wherein, n is n-th reconnection, n < 10) is reconnected after second, reconnects successfully, then connection is put into connection pool, reconnection
Connect failure and n=10 then illustrate can not by reconnection restore connect, it is most likely that be because Physical Links Layer failure cause, because
This needs to trigger alarm, to prompt the road video flowing to have connecting fault.
Step S103: it selects a kind of queue to store the frame data pulled, and monitors the state of queue.
Preferential selection asynchronous queue is then switched to synchronous obstruction queue when asynchronous queue's exception；When synchronous obstruction queue
Consumption terminal speed is faster than the manufacturing side, and this it is in stable condition continue for a period of time, then switch to asynchronous queue.
It is understood that since the rate of consumption data is different, demand is different, for different applied fields
Scape, the different cache policy of dynamic initialization.Such as: if calculation processing ability is limited, and each frame of camera is not required
Data are all processed, that is, under the scene for allowing frame losing, it should cache newest frame using synchronous obstruction queue, have little time to handle
Frame directly abandon；If using the image processing equipment of the high performance such as GPU, and requiring to be unable under the scene of frame losing, answer
The reasonable spatial cache of the distribution, to guarantee that frame data can be handled in time.
Preferential selection asynchronous queue caches frame data, to guarantee that frame data can be processed.When asynchronous queue has expired, explanation
The consumption rate in downstream is too slow, captures a data overflow exception message at this time, and selects synchronous obstruction queue data cached, with
Guarantee the timeliness of frame data；When the consumption terminal speed for monitoring synchronous obstruction queue is faster than manufacturing side speed, and this shape
State continue for a period of time, illustrates that the processing speed in downstream is recovered at this time, is then switched to asynchronous queue again.
Step S104: data are read from connection pool as unit of frame, detect all people's face in frame.
Whether the quality testing face has the standard of subsequent recognition of face (in order to which real-time is captured in maximum guarantee, originally
Passability inspection is only done in the quality testing of invention, has abandoned the way of " selecting optimal " for using in existing candid photograph technology.Such as: shape
As "<A1:29>,<B1:19>,<B2:24>,<A2:14>,<B3:53>,<A3:72>... " and human face data stream pass sequentially through matter
Inspection, it is assumed that the standard of quality inspection is to measure face A1 greater than 20 greater than 20, then illustrate that the face has met the mark of recognition of face
Therefore standard exports face A1 to next processing step, without traversing defined time or defined number of frames at once
After interior all faces, one maximum of reselection is greater than 20 face (such as A3)).If quality inspection is unqualified, directly discarded the people
Face.
Step S105: by facial image vector quantization, face characteristic value set is obtained.
Characteristic value cache set Af1All face characteristic values of previous frame are stored, " characteristic value cache set A is initializedf1For sky
Set.Formula are as follows:
Af1=Φ
Wherein A is the set of a storage face, f1It is the number of frame.
Characteristic value cache set Bf2For storing all face characteristic values of present frame, initialization feature value stores set Bf2
All face set being detected for present frame.Formula are as follows:
Bf2=f (frame)
Wherein B is the set of a storage face, f2It is the number of frame, f is Face datection algorithm, and frame is present frame.
Calculate the face set for needing to export, formula are as follows:
T=Bf2-Af1
Wherein f2Frame is f1A later frame, i.e. f2It is later than f at the time of frame occurs1At the time of appearance.Formula are as follows:
Wherein, fps is frame per second.
The characteristic value collection B of the framef2With the A of previous framef1It is difference (i.e. Bf2-Af1) operation, obtained result set is to grab
The face photographed.Difference operation is more in line with actual face snap scene, and perceiving present frame and previous frame when camera has
The similarities and differences, difference operation immediately respond this similarities and differences, without just responding after several seconds or number frame.And
Interval frame or the way for being spaced the second certainly will will cause the loss of certain faces.The way of frame difference method face duplicate removal of the invention was both
It ensure that not losing frame data in turn ensures response speed.
Step S106: the face captured of output, and with characteristic value collection Bf2Update characteristic value collection Af1。
Characteristic value is stored into set Bf2Value be assigned to characteristic value storage set Af1, formula are as follows:
Af1=Bf2
As can be seen from the above description, face identification method provided by the invention, according to every in the human face data obtained in advance
Frame data generate characteristic value corresponding to the frame data, and obtained characteristic value and preset eigenvalue threshold are made ratio
Compared with according to characteristic value one characteristic value collection of generation for being greater than eigenvalue threshold, then, by this feature value set and previous frame data
Characteristic value collection do difference, finally obtain face recognition result.This method overcome in the prior art face snap technology without
Method combines the problem of capturing real-time and capturing effect.Specific advantage is as follows:
1) the face snap method for using frame difference, is truly realized real-time grasp shoot while meeting face quality requirement
Effect.It is traditional quickly capture, leave after the method captured require whithin a period of time comparison chart as mass effect, this side
Formula had not only wasted computing resource but also had been unable to reach the effect of real-time grasp shoot, and traditional interval frame is captured, the interval second captures can not
Guarantee that the quality captured or its implementation are excessively complicated, be not suitable for this computation-intensive of face snap and require to guarantee
The scene of real-time.
2) process provides a kind of flexible face snap system structures, so that face snap set expandability is extremely strong.
There are serious Function Couplings for the system authority of traditional grasp shoot method, are unfavorable for the landing and extension of face snap method.The present invention
The system structure of offer from getting frame, saves all decouplings of the several core apparatus of frame, Face datection, face duplicate removal, different performance
Machine can cooperate, not only facilitate horizontal behavior extension, but also facilitate Function Extension, the support of hot plug be provided.
Based on the same inventive concept, the embodiment of the present application also provides face identification devices, can be used to implement above-mentioned reality
Apply method described in example, such as the following examples.The principle and face identification method solved the problems, such as due to face identification device
It is similar, therefore the implementation of face identification device may refer to face identification method implementation, overlaps will not be repeated.It is following to be made
, the combination of the software and/or hardware of predetermined function may be implemented in term " unit " or " module ".Although following embodiment
Described system preferably realized with software, but the combined realization of hardware or software and hardware be also may be simultaneously
It is contemplated.
The embodiment of the present invention provides a kind of specific embodiment party of face identification device that can be realized face identification method
Formula, referring to Fig. 7, face identification device specifically includes following content:
Characteristic value generation unit 10, for generating the frame data according to every frame data in the human face data obtained in advance
Corresponding characteristic value；
As a result generation unit 20, for according to preset eigenvalue threshold, the frame data characteristic value and the frame
The characteristic value of the previous frame data of data generates face recognition result.
Preferably, referring to Fig. 8, face identification device further include:
Human face data acquiring unit 30 obtains the human face data for connecting by establishing rtsp with camera.
Preferably, referring to Fig. 9, characteristic value generation unit 10 includes:
One-dimensional vector generation module 101, for according to every frame data vector quantization in the human face data, generate it is one-dimensional to
Amount；
Characteristic value generation module 102, for according to the one-dimensional vector generate the frame data corresponding to characteristic value.
Preferably, referring to Figure 10, as a result generation unit 20 includes:
Comparison module 201, for the characteristic value of the frame data and the size of the eigenvalue threshold；
Difference block 202, for when the characteristic value of the frame data be greater than the eigenvalue threshold when, by the frame data
The characteristic value of characteristic value and the previous frame data of the frame data does difference；
Identification module 203, for carrying out recognition of face according to difference result.
As can be seen from the above description, face identification device provided by the invention, according to every in the human face data obtained in advance
Frame data generate characteristic value corresponding to the frame data, and obtained characteristic value and preset eigenvalue threshold are made ratio
Compared with according to characteristic value one characteristic value collection of generation for being greater than eigenvalue threshold, then, by this feature value set and previous frame data
Characteristic value collection do difference, finally obtain face recognition result.This method overcome in the prior art face snap technology without
Method combines the problem of capturing real-time and capturing effect.Specific advantage is as follows:
1) the face snap method for using frame difference, is truly realized real-time grasp shoot while meeting face quality requirement
Effect.It is traditional quickly capture, leave after the method captured require whithin a period of time comparison chart as mass effect, this side
Formula had not only wasted computing resource but also had been unable to reach the effect of real-time grasp shoot, and traditional interval frame is captured, the interval second captures can not
Guarantee that the quality captured or its implementation are excessively complicated, be not suitable for this computation-intensive of face snap and require to guarantee
The scene of real-time.
2) process provides a kind of flexible face snap system structures, so that face snap set expandability is extremely strong.
There are serious Function Couplings for the system authority of traditional grasp shoot method, are unfavorable for the landing and extension of face snap method.The present invention
The system structure of offer from getting frame, saves all decouplings of the several core apparatus of frame, Face datection, face duplicate removal, different performance
Machine can cooperate, not only facilitate horizontal behavior extension, but also facilitate Function Extension, the support of hot plug be provided.
Embodiments herein, which also provides, can be realized one of Overall Steps in the face identification method in above-described embodiment
The specific embodiment of kind electronic equipment, referring to Figure 11, electronic equipment specifically includes following content:
Processor (processor) 1201, memory (memory) 1202, communication interface (Communications
Interface) 1203 and bus 1204；
Wherein, processor 1201, memory 1202, communication interface 1203 complete mutual communication by bus 1204；
Communication interface 1203 passes for realizing the information between the relevant devices such as server-side devices, capture apparatus and ustomer premises access equipment
It is defeated.
Processor 1201 is used to call the computer program in memory 1202, and processor is realized when executing computer program
The Overall Steps in face identification method in above-described embodiment, for example, processor realizes following steps when executing computer program
It is rapid:
Step 100: generating feature corresponding to the frame data according to every frame data in the human face data obtained in advance
Value.
Step 200: according to the previous frame of preset eigenvalue threshold, the characteristic value of the frame data and the frame data
The characteristic value of data generates face recognition result.
As can be seen from the above description, the electronic equipment in the embodiment of the present application, according to every in the human face data obtained in advance
Frame data generate characteristic value corresponding to the frame data, and obtained characteristic value and preset eigenvalue threshold are made ratio
Compared with according to characteristic value one characteristic value collection of generation for being greater than eigenvalue threshold, then, by this feature value set and previous frame data
Characteristic value collection do difference, finally obtain face recognition result.This method overcome in the prior art face snap technology without
Method combines the problem of capturing real-time and capturing effect.Specific advantage is as follows:
1) the face snap method for using frame difference, is truly realized real-time grasp shoot while meeting face quality requirement
Effect.It is traditional quickly capture, leave after the method captured require whithin a period of time comparison chart as mass effect, this side
Formula had not only wasted computing resource but also had been unable to reach the effect of real-time grasp shoot, and traditional interval frame is captured, the interval second captures can not
Guarantee that the quality captured or its implementation are excessively complicated, be not suitable for this computation-intensive of face snap and require to guarantee
The scene of real-time.
2) process provides a kind of flexible face snap system structures, so that face snap set expandability is extremely strong.
There are serious Function Couplings for the system authority of traditional grasp shoot method, are unfavorable for the landing and extension of face snap method.The present invention
The system structure of offer from getting frame, saves all decouplings of the several core apparatus of frame, Face datection, face duplicate removal, different performance
Machine can cooperate, not only facilitate horizontal behavior extension, but also facilitate Function Extension, the support of hot plug be provided.
Embodiments herein, which also provides, can be realized one of Overall Steps in the face identification method in above-described embodiment
Computer readable storage medium is planted, is stored with computer program on computer readable storage medium, the computer program is processed
Device realizes the Overall Steps of the face identification method in above-described embodiment when executing, for example, when processor executes computer program
Realize following step:
Step 100: generating feature corresponding to the frame data according to every frame data in the human face data obtained in advance
Value.
Step 200: according to the previous frame of preset eigenvalue threshold, the characteristic value of the frame data and the frame data
The characteristic value of data generates face recognition result.
As can be seen from the above description, the computer readable storage medium in the embodiment of the present application, according to the face obtained in advance
Every frame data in data generate characteristic value corresponding to the frame data, and by obtained characteristic value and preset characteristic value
Threshold value is made comparisons, according to be greater than eigenvalue threshold characteristic value generate a characteristic value collection, then, by this feature value set with it is upper
The characteristic value collection of one frame data does difference, finally obtains face recognition result.This method overcomes face in the prior art and grabs
Bat technology can not combine the problem of capturing real-time and capturing effect.Specific advantage is as follows:
1) the face snap method for using frame difference, is truly realized real-time grasp shoot while meeting face quality requirement
Effect.It is traditional quickly capture, leave after the method captured require whithin a period of time comparison chart as mass effect, this side
Formula had not only wasted computing resource but also had been unable to reach the effect of real-time grasp shoot, and traditional interval frame is captured, the interval second captures can not
Guarantee that the quality captured or its implementation are excessively complicated, be not suitable for this computation-intensive of face snap and require to guarantee
The scene of real-time.
2) process provides a kind of flexible face snap system structures, so that face snap set expandability is extremely strong.
There are serious Function Couplings for the system authority of traditional grasp shoot method, are unfavorable for the landing and extension of face snap method.The present invention
The system structure of offer from getting frame, saves all decouplings of the several core apparatus of frame, Face datection, face duplicate removal, different performance
Machine can cooperate, not only facilitate horizontal behavior extension, but also facilitate Function Extension, the support of hot plug be provided.
All the embodiments in this specification are described in a progressive manner, same and similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.Especially for hardware+
For program class embodiment, since it is substantially similar to the method embodiment, so being described relatively simple, related place is referring to side
The part of method embodiment illustrates.
It is above-mentioned that this specification specific embodiment is described.Other embodiments are in the scope of the appended claims
It is interior.In some cases, the movement recorded in detail in the claims or step can be come according to the sequence being different from embodiment
It executes and desired result still may be implemented.In addition, process depicted in the drawing not necessarily require show it is specific suitable
Sequence or consecutive order are just able to achieve desired result.In some embodiments, multitasking and parallel processing be also can
With or may be advantageous.
Although this application provides the method operating procedure of such as embodiment or flow chart, based on routine or without creativeness
Labour may include more or less operating procedure.The step of enumerating in embodiment sequence is only that numerous steps execute
One of sequence mode, does not represent and unique executes sequence.It, can be by when device in practice or client production execute
It is executed according to embodiment or method shown in the drawings sequence or parallel executes (such as parallel processor or multiple threads
Environment).
It should be understood by those skilled in the art that, the embodiment of the present invention can provide as method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the present invention
Apply the form of example.Moreover, it wherein includes the computer of computer usable program code that the present invention, which can be used in one or more,
The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) produces
The form of product.
The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Specific embodiment is applied in the present invention, and principle and implementation of the present invention are described, above embodiments
Explanation be merely used to help understand method and its core concept of the invention；At the same time, for those skilled in the art,
According to the thought of the present invention, there will be changes in the specific implementation manner and application range, in conclusion in this specification
Appearance should not be construed as limiting the invention.
Claims (10)
1. a kind of face identification method characterized by comprising
Characteristic value corresponding to the frame data is generated according to every frame data in the human face data obtained in advance；
According to the characteristic value of the previous frame data of preset eigenvalue threshold, the characteristic value of the frame data and the frame data
Generate face recognition result.
2. face identification method according to claim 1, which is characterized in that further include:
It is connect by establishing rtsp with camera, obtains the human face data.
3. face identification method according to claim 1, which is characterized in that in the human face data that the basis obtains in advance
Every frame data generate characteristic value corresponding to the frame data, comprising:
According to every frame data vector quantization in the human face data, one-dimensional vector is generated；
According to the one-dimensional vector generate the frame data corresponding to characteristic value.
4. face identification method according to claim 1, which is characterized in that described according to preset eigenvalue threshold, institute
The characteristic value for stating the characteristic value of frame data and the previous frame data of the frame data generates face recognition result, comprising:
Compare the characteristic value of the frame data and the size of the eigenvalue threshold；
When the characteristic value of the frame data is greater than the eigenvalue threshold, by the upper of the characteristic value of the frame data and the frame data
The characteristic value of one frame data does difference；
Recognition of face is carried out according to difference result.
5. a kind of face identification device characterized by comprising
Characteristic value generation unit, for according to corresponding to every frame data generation frame data in the human face data obtained in advance
Characteristic value；
As a result generation unit, for according to the characteristic value of preset eigenvalue threshold, the frame data and the frame data
The characteristic value of previous frame data generates face recognition result.
6. face identification device according to claim 5, which is characterized in that further include:
Human face data acquiring unit obtains the human face data for connecting by establishing rtsp with camera.
7. face identification device according to claim 5, which is characterized in that characteristic value generation unit includes:
One-dimensional vector generation module, for generating one-dimensional vector according to every frame data vector quantization in the human face data；
Characteristic value generation module, for according to the one-dimensional vector generate the frame data corresponding to characteristic value.
8. face identification device according to claim 5, which is characterized in that result generation unit includes:
Comparison module, for the characteristic value of the frame data and the size of the eigenvalue threshold；
Difference block, for when the characteristic value of the frame data be greater than the eigenvalue threshold when, by the characteristic value of the frame data
Difference is done with the characteristic value of the previous frame data of the frame data；
Identification module, for carrying out recognition of face according to difference result.
9. a kind of electronic equipment including memory, processor and stores the calculating that can be run on a memory and on a processor
Machine program, which is characterized in that the processor realizes the recognition of face of any one of Claims 1-4 when executing described program
The step of method.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program quilt
The step of any one of the Claims 1-4 face identification method is realized when processor executes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910701882.7A CN110414444A (en) | 2019-07-31 | 2019-07-31 | Face identification method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910701882.7A CN110414444A (en) | 2019-07-31 | 2019-07-31 | Face identification method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110414444A true CN110414444A (en) | 2019-11-05 |
Family
ID=68364711
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910701882.7A Pending CN110414444A (en) | 2019-07-31 | 2019-07-31 | Face identification method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110414444A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108090428A (en) * | 2017-12-08 | 2018-05-29 | 广西师范大学 | A kind of face identification method and its system |
| US20180211101A1 (en) * | 2016-03-09 | 2018-07-26 | International Business Machines Corporation | Face detection, representation, and recognition |
| CN108932465A (en) * | 2017-12-28 | 2018-12-04 | 浙江宇视科技有限公司 | Reduce the method, apparatus and electronic equipment of Face datection false detection rate |
| CN109359625A (en) * | 2018-11-16 | 2019-02-19 | 南京甄视智能科技有限公司 | The method and system of customer identification is judged based on head and shoulder detection and face recognition technology |
-
2019
- 2019-07-31 CN CN201910701882.7A patent/CN110414444A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180211101A1 (en) * | 2016-03-09 | 2018-07-26 | International Business Machines Corporation | Face detection, representation, and recognition |
| CN108090428A (en) * | 2017-12-08 | 2018-05-29 | 广西师范大学 | A kind of face identification method and its system |
| CN108932465A (en) * | 2017-12-28 | 2018-12-04 | 浙江宇视科技有限公司 | Reduce the method, apparatus and electronic equipment of Face datection false detection rate |
| CN109359625A (en) * | 2018-11-16 | 2019-02-19 | 南京甄视智能科技有限公司 | The method and system of customer identification is judged based on head and shoulder detection and face recognition technology |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109313484B (en) | Virtual reality interaction system, method and computer storage medium | |
| CN106412626B (en) | A kind of processing method and processing device of live video | |
| CN105791431B (en) | A kind of distribution on line formula monitor video processing method for scheduling task and device | |
| US10693938B2 (en) | Method and system for interactive transmission of panoramic video | |
| TWI355203B (en) | Image pickup apparatus and image distributing meth | |
| US10609426B2 (en) | Method and system for playing back recorded video | |
| CN104702563B (en) | The acquisition methods and device of streaming media playing duration | |
| US20090185784A1 (en) | Video surveillance system and method using ip-based networks | |
| US11282169B2 (en) | Method and apparatus for processing and distributing live virtual reality content | |
| CN113569825B (en) | Video monitoring method and device, electronic equipment and computer readable medium | |
| CN106454388B (en) | A kind of method and apparatus for determining live streaming setting information | |
| WO2014094537A1 (en) | Immersion communication client and server, and method for obtaining content view | |
| US20090238411A1 (en) | Estimating motion of an event captured using a digital video camera | |
| CN106162105B (en) | A kind of camera parameters control method and device in video monitoring system | |
| WO2016192467A1 (en) | Method and device for playing videos | |
| CN110414444A (en) | Face identification method and device | |
| Li et al. | Digital twin and artificial intelligence-empowered panoramic video streaming: reducing transmission latency in the extended reality-assisted vehicular metaverse | |
| Wen et al. | Cloud-computing-based framework for multi-camera topology inference in smart city sensing system | |
| CN111800604A (en) | Method and device for detecting human shape and human face data based on gun and ball linkage | |
| JP7021842B2 (en) | Video distribution system and video distribution method | |
| Lee et al. | A real-time face tracking system based on a single PTZ camera | |
| JP2015506622A (en) | Method for video surveillance, associated system, associated surveillance server and associated surveillance camera | |
| JP3957509B2 (en) | Video distribution server and video distribution system | |
| Hu et al. | LiveVV: Human-Centered Live Volumetric Video Streaming System | |
| WO2018120857A1 (en) | Streaming media technology-based method and apparatus for processing video data |
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: 20191105 |
|
| RJ01 | Rejection of invention patent application after publication |