CN107169433A - A kind of face identification method based on Streaming Media - Google Patents
A kind of face identification method based on Streaming Media Download PDFInfo
- Publication number
- CN107169433A CN107169433A CN201710324671.7A CN201710324671A CN107169433A CN 107169433 A CN107169433 A CN 107169433A CN 201710324671 A CN201710324671 A CN 201710324671A CN 107169433 A CN107169433 A CN 107169433A
- Authority
- CN
- China
- Prior art keywords
- end system
- face
- data
- equipment
- image
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
- G06V40/169—Holistic features and representations, i.e. based on the facial image taken as a whole
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D3/00—Control of position or direction
- G05D3/12—Control of position or direction using feedback
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L65/00—Network arrangements, protocols or services for supporting real-time applications in data packet communication
- H04L65/60—Network streaming of media packets
- H04L65/65—Network streaming protocols, e.g. real-time transport protocol [RTP] or real-time control protocol [RTCP]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L65/00—Network arrangements, protocols or services for supporting real-time applications in data packet communication
- H04L65/60—Network streaming of media packets
- H04L65/75—Media network packet handling
- H04L65/762—Media network packet handling at the source
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Oral & Maxillofacial Surgery (AREA)
- General Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
A kind of face identification method based on Streaming Media includes:The angle of image capture device is adjusted, and angle configuration information is sent to front end system;Front end system will be sent to back-end system after the information processing;Back-end system determines the need for adjustment angle again;After determining the need for adjustment, by image capture device capture images data;View data is carried out Streaming Media processing by front end system, and is sent to back-end system;Back-end system receives view data and handled, and recognizes face therein;Object in face and database that back-end system will identify that is compared, and determines result.Pass through this method, it can ensure that power consumption and cost do not increase not increasing equipment, the all standing formula capture of three dimensions, strengthen data diversity and flexibility, ensure transmission quality, keep the continuity of data acquisition, save the time, recognition speed is fast, accuracy rate is high result is provided, and causes due to less calculating the power consumption of reduction.
Description
Technical field
The present invention relates generally to computer picture field of signal processing, more specifically, it is related to a kind of based on Streaming Media
Face identification method.
Background technology
Recognition of face is the important branch of computer vision information processing, refers to that the facial feature information based on people carries out body
A kind of biological identification technology of part identification.Image or video flowing containing face with video camera or camera collection, and exist automatically
Detect and track face in image, and then a series of correlation techniques of face are carried out to the face detected.Face recognition technology
Suffer from being widely applied in fields such as national security, military security and public safeties, such as customs's entry-exit management, holder
In authentication, robot research, the scene such as sense of reality virtual game.In addition, with the fast development of information technology, to image
The transmission of data and in time playback have requirements at the higher level, on this basis, and stream media technology is applied and given birth to.Stream media technology is one
New network multimedia technology is planted, it is Compression Techniques of Multimedia Data, data stream scheduling strategy and network data transmission
Control technology organically combines, and user is can be carried out viewing while downloading data, significantly shortens
User etc. it is to be delayed, and saved Internet resources.Streaming media image business refers to that terminal is downloaded while playing image
Data.Being characterized in that user has all downloaded without waiting for picture material can just watch, it is only necessary to data buffer storage to be downloaded
To can just play after a certain amount of, then in playing process, new data are cached to terminal successively, continuous with keep playing
Property.
Due to the demand of application field, above-mentioned technology is combined then can solve the problem that it is many in Object identifying field
Problem.Such as, identification needs quick, accurate, flexible.And in the prior art, due to the original of data acquisition, processing and transmission
, all there is considerable deficiency in the acquisition, processing, result of data are determined in cause.Often the acquisition of pending source data is fast
Degree is not fully up to expectations, and a variety of causes due to view data in conversion, transmitting procedure, data can be caused not clean enough,
Cause that follow-up identification process is time-consuming, accuracy, and more power consumptions are caused due to more calculate.In addition,
Image capture device of the prior art, is often fixed on wall or ceiling, past between capture camera lens and fixing equipment
Toward using being fixedly connected, cause due to fixed single shooting angle, if object moves out the area of establishing shot angle covering
Domain, then can not continue tracking, if it is desired to continuing to track, it is necessary to increase the quantity of image capture device, and now need to carry out
Segmentation, splicing and the linking of image, this had both brought equipment increase and the increase of corresponding power consumption, and had also caused the increase of cost.Even if
Image capture device can be moved relative to fixing equipment, but at most be also rotation of its rotary shaft in a plane, it is impossible to be protected
Card carries out all standing formula capture of three dimensions.Lack the achievement of this respect in the prior art.
The content of the invention
An object of the present invention is to provide a kind of face identification method based on Streaming Media.With it, can be
On the premise of not increasing equipment, it is ensured that power consumption and cost do not increase, it is ensured that carry out all standing formula capture of three dimensions, enhancing
The diversity of data and flexibility, it is ensured that transmission quality, can keep the continuity of data acquisition, and when having saved
Between, it is that the efficiency raising of follow-up recognition of face is made that contribution, the result that recognition speed is fast, accuracy rate is high can be provided, and
Cause the power consumption of reduction due to less calculating.
The present invention is to solve the technical scheme taken of above-mentioned technical problem:A kind of recognition of face side based on Streaming Media
Method, comprises the following steps:In step sl, the angle of image capture device is adjusted, and angle configuration information is sent to front end
System;In step s 2, front end system will be sent to back-end system after the information processing;In step s3, back-end system is determined
Whether again adjustment angle is needed;In step s 4, after determining the need for adjustment, by image capture device capture images number
According to;In step s 5, view data is carried out Streaming Media processing by front end system, and is sent to back-end system;In step s 6,
Back-end system receives view data and handled, and recognizes face therein;And in the step s 7, the people that back-end system will identify that
Face is compared with the object in database, and determines result.
According to another aspect of the present invention, in step s 2, front end system will be sent to rear end system after the information processing
System further comprises:In the step s 21, the information that front end system is captured by angle configuration information and in preset time period is carried out
Processing;And in step S22, the information of processing is sent to back-end system by front end system.
According to another aspect of the present invention, in step s3, back-end system determines the need for adjustment angle again and entered
One step includes:Back-end system is according to the information of the processing of reception, it is determined whether need adjustment angle again, if it is determined that image is caught
The capture angle for obtaining equipment C sets suitably, then directly carries out the capture of image in step s 4;If it is determined that image capture is set
Standby C capture angle sets inappropriate, that is, needs to adjust capture angle again, then return to adjustment angle, and in step s 4
The capture of image is carried out after adjustment.
According to another aspect of the present invention, in step sl, the angle of image capture device is adjusted, and angle is set
Information is sent to front end system, and the front end system 1 includes fixing equipment A, adjustment equipment B, image capture device C, the first processing
Equipment D and the first transceiver E, fixing equipment A are coupled with adjustment equipment B, adjustment equipment B and image capture device C and the
One processing equipment D is coupled respectively, and the first processing equipment D and the coupling of the first transceiver;Adjustment equipment B includes being linked in sequence
Connecting elements B0, rotational electric component B1, connecting elements B2, rotational electric component B3, connecting elements B4, rotatable electricity
Dynamic component B5 and connecting elements B6;When rotational electric component receives electric signal by wire link, during connecting elements is
Hollow structure, it is internally provided with electric wiring, and when rotational electric component receives electric signal by Radio Link, connection
Component can be hollow structure or solid construction, now rotational electric component inside be provided with proximity communication module and letter
Number processing module, it communicates with being again provided with the first processing equipment of proximity communication module;Connecting elements B0 connection is set
It is set to:Connecting elements B0 is directly or indirectly connected to any one in fixing equipment A and the first processing equipment or both;Regulation
Connecting elements B0 in equipment B, rotational electric component B1, connecting elements B2, rotational electric component B3, connecting elements B4,
Rotational electric component B5, connecting elements B6 are specifically configured to:Three rotational electric components are cylindrical structure, and it is each wrapped
Multiple sub- components that can mutually move are included, many sub- components can be moved partly or entirely relative to other equipment;Its
Mutually it is in 90 ° between the axis of middle rotational electric component B1, B3, B5 cylinder bottom surface, i.e. every in these axis
One is placed perpendicular to each other so that the rotary shaft of three rotational electric components cover all three-dimensionals, i.e. X, Y and
Z-direction;Structure can be wholly or partially straight, bending or angled;Further, connecting elements can using L,
S, U, V, spiral shape.
According to another aspect of the present invention, in step s 5, view data is carried out Streaming Media processing by front end system, and
And be sent to back-end system and further comprise:In step s 51, the first processing equipment in front end system receives image capture and set
The standby image information with a certain speed trap;In step S52, the first processing equipment in front end system is compiled the data
Code compression, is specifically included:View data is first divided into fragment, sub-piece is divided into again afterwards, then sub-piece is carried out pre-
Survey, it is in the following ways:It is reference frame by the neighbouring frame re-established for having completed coding, performs motion compensation, afterwards
After the completion of, subtract the actual numerical value of current sub-piece with its numerical value, generate differential data, by converting laggard line number value, via
Entropy code, the data of entropy code is merged into the direction of prediction and motion-vector the data flow of compression;It is preceding in step S53
The data flow is carried out follow-up packing processing by the first processing equipment in end system, and by the data input for processing of packing to buffering
In device;And in step S54, the first transceiver passes through the data in buffer wired or wireless according to transmission speed
Link transmission.
According to another aspect of the present invention, in step S54, the first transceiver is according to transmission speed, by buffer
In data by wired or wireless link transmission, wherein radio communication includes Wi-Fi, bluetooth (BT), near-field communication (NFC),
Global positioning system (GPS), and in the cellular communication including LTE, LTE-A, CDMA, WCDMA, UMTS, WiBro, GSM extremely
It is few one;Wire communication is included in USB (USB), HDMI (HDMI), RS-232 and POTS
At least one.
According to another aspect of the present invention, in step s 6, back-end system receives view data and handled, and identification is wherein
Face further comprise:In step S61, the second transceiver wherein in back-end system is received to be sent out by the first transceiver
The image sent, and be entered into identification equipment;In step S62, identification equipment is according to default agreement, by input
Data are decoded, and obtain image sequence;In step S63, identification equipment is pre-processed to image sequence, and it includes gray scale
Processing, illumination compensation, smoothing denoising, image sharpening etc., to reduce interference signal.Wherein in order to reduce capture due to image,
The noise brought during transmission etc., the present invention reduces interference using denoising mode, is that follow-up identification prepares accurate figure
As data;In step S64, the change in image is judged, when it is determined that foreground area is more than the first critical value, it is determined that
Change, and then enter step S65, and when determination foreground area is not less than the first critical value, it is determined that it is unchanged, terminate the section
The operation of sequence, then carry out the operation of next sequence;In step S65, according to the shape description at face position and they
The distance between characteristic contribute to the characteristic of face classification to obtain, this feature data include characteristic component, this feature point
Amount includes Euclidean distance, curvature and the angle between characteristic point;The characteristic value for extracting face data using PCA algorithms forms facial spy
Levy matrix, the determination of face rectangle carried out using neural-network classification method, determine right and left eyes in face rectangle, left and right eyebrow,
Nose, the left and right corners of the mouth, chin, are rotated and are scaled to the gray-scale map of rectangle, and extract the feature at above-mentioned position;Face rectangle
Operation include:Determine face and mark, set the N*N pixels of face, wherein N is positive integer;Face in prominent image is simultaneously
Rectangle is set, rectangle size is determined and shows;In step S66, face information is determined, and by the image array and information of face
It is stored into storage device H.
According to another aspect of the present invention, back-end system 2 includes the second transceiver F, identification equipment G, storage device
H, the 3rd transceiver I;Step S63 further comprises the steps:The geological information of the subgraph in image sequence is extracted, it is all
Such as size;Build gray-scale map;Carry out the conversion of color space;Create partition holding and perform initialization operation;To gray-scale map
Bar chart handled.
According to another aspect of the present invention, in the step s 7, in back-end system will identify that face and database
Object is compared, and determines that result further comprises:By the 3rd transceiver I of back-end system, via chain road direction data
Storehouse 3 sends request, and human face data and the human face data that has stored are compared, and obtains numerical value, if the numerical value exceedes
Second critical value, it is determined that the identity information of identification object, otherwise returns to no corresponding informance.
According to another aspect of the present invention, this method further comprises:In step s 8, back-end system is according to reception
View data, it is determined whether need another adjustment angle, so as to the corresponding object of the face of Tracking Recognition, through determining to need to adjust
After whole, back-end system sends the feedback information of another adjustment angle, front end system by the second transceiver module, forward end system
Repeat to continue executing with since step S1 based on this.
Brief description of the drawings
Embodiments of the invention, wherein phase are shown by way of example rather than by way of limitation in the accompanying drawings
Same reference represents identical element, wherein:
According to an exemplary embodiment of the invention, Fig. 1 illustrates the flow chart of the face identification method based on Streaming Media.
According to an exemplary embodiment of the invention, Fig. 2 illustrates the structure function figure of front end system.
According to an exemplary embodiment of the invention, the structure function figure and detail view of Fig. 3 A and 3B diagrams adjustment equipment.
According to an exemplary embodiment of the invention, Fig. 4 illustrates the structure function figure of back-end system.
According to an exemplary embodiment of the invention, Fig. 5 illustrates the face identification system based on Streaming Media.
Embodiment
In the following description, refer to the attached drawing and several specific embodiments are diagrammatically shown.It will be appreciated that:
It is contemplated that and other embodiment can be made without departing from the scope of the present disclosure or spirit.Therefore, it is described in detail below should not be by
Think in a limiting sense.
According to an exemplary embodiment of the invention, Fig. 1 illustrates the flow chart of the face identification method based on Streaming Media.
In step sl, the angle of image capture device is adjusted, and angle configuration information is sent to front end system;
In step s 2, front end system will be sent to back-end system after the information processing;
In step s3, back-end system determines the need for adjustment angle again;
In step s 4, after determining the need for adjustment, by image capture device capture images data;
In step s 5, view data is carried out Streaming Media processing by front end system, and is sent to back-end system;
In step s 6, back-end system receives view data and handled, and recognizes face therein;
In the step s 7, the object in back-end system will identify that face and database is compared, and determines result.
Specifically, in step sl, the angle of image capture device is adjusted, and angle configuration information is sent to front end system
System, wherein as shown in Figure 2, front end system 1 is set including fixing equipment A, adjustment equipment B, image capture device C, the first processing
Standby D and the first transceiver E, wherein fixing equipment A are coupled with adjustment equipment B, adjustment equipment B and image capture device C and
First processing equipment D is coupled respectively, and the first processing equipment D and the coupling of the first transceiver.Alternatively, front end system 1 can
Not include fixing equipment A.Wherein as shown in fig. 3, wherein adjustment equipment B includes order to adjustment equipment B functional structure chart
The connecting elements B0 of connection, rotational electric component B1, connecting elements B2, rotational electric component B3, connecting elements B4, can revolve
Turn electrical components B5, connecting elements B6.Preferably, in order to strengthen flexibility, more connecting elements can be used and rotatable
Electrical components.Wherein connecting elements plays connection supporting role, and rotational electric component plays a part of to be rotated according to electric signal;And
And when rotational electric component receives electric signal by wire link, connecting elements is hollow structure, and it is internally provided with
Electric wiring, and when rotational electric component receives electric signal by Radio Link, connecting elements can be hollow structure
Or solid construction, now rotational electric component inside be provided with proximity communication module and signal processing module, its with it is same
Sample is provided with the first processing equipment communication of proximity communication module.Wherein connecting elements B0 connection is set to:If front end
When system 1 does not include fixing equipment A, connecting elements B0 is directly or indirectly connected to the first processing equipment;If front end system 1
During including fixing equipment A, connecting elements B0 can be directly or indirectly connected to appointing in fixing equipment A and the first processing equipment
One or both.Connecting elements B0 in adjustment equipment B, rotational electric component B1, connecting elements B2, rotational electric component
B3, connecting elements B4, rotational electric component B5, connecting elements B6 are specifically configured to:Three rotational electric components are cylinder
Body structure, as shown in Figure 3 B, it each includes multiple sub- component BB1, BB2 ... that can mutually move, and is risen in order to simple
See, Fig. 3 B only show two sub- components, and BB1 fixed and BB2 rotates relative to motion, but this area is common
Technical staff is appreciated that:Can use more sub- components, and many sub- components can partly or entirely relative to
Other equipment is moved.Mutually it is in 90 ° wherein between the axis of rotational electric component B1, B3, B5 cylinder bottom surface, i.e.
Each in these axis is placed perpendicular to each other so that the rotary shaft covering of three rotational electric components is all
Three-dimensional, i.e. X, Y and Z-direction, then allow adjustment equipment to be rotated at any angle in space, and then B6 is connected
The image capture device C connect captures the view data in whole space in which can not stay dead angle, so as to improve many of capture data
Sample, flexibility, improve the specific aim of follow-up recognition of face, and can reduce the usage quantity of image capture device, reduce
Total power consumption and cost.Specifically, structure can be straight on the whole, or its can be in whole some parts bending or into
Angle, this depends on the requirement of the application region of equipment;Further, connecting elements can use any shape, such as and
It is not limited to L, S, U, V, spiral shape etc..
By with the operation of streaming media, on the premise of equipment is not increased, it is ensured that power consumption and cost do not increase, it is ensured that
The all standing formula capture of three dimensions is carried out, diversity and the flexibility of data is enhanced.
Specifically, in step s 2, front end system will be sent to back-end system after the information processing and further comprise:
In the step s 21, the information that front end system is captured by angle configuration information and in preset time period is handled;
And
In step S22, the information of processing is sent to back-end system by front end system.
Specifically, in step s3, back-end system determines the need for adjustment angle again and further comprised:
Back-end system is according to the information of the processing of reception, it is determined whether need adjustment angle again, if it is determined that image is caught
The capture angle for obtaining equipment C sets suitably, then directly carries out the capture of image in step s 4;If it is determined that image capture is set
Standby C capture angle sets inappropriate, that is, needs to adjust capture angle again, then return to adjustment information, and in step s 4
The capture of image is carried out after adjustment.
Specifically, in step s 5, view data is carried out Streaming Media processing by front end system, and is sent to back-end system
Further comprise:
In step s 51, the first processing equipment in front end system receives image capture device with a certain speed trap
Image information;
In step S52, the data are carried out coding compression by the first processing equipment in front end system;
Preferably, the coding compression of data includes:View data is first divided into fragment, sub-piece is divided into again afterwards,
Then sub-piece is predicted, it is in the following ways:It is reference frame by the neighbouring frame re-established for having completed coding,
Perform motion compensation, afterwards after the completion of, subtract the actual numerical value of current sub-piece with its numerical value, generate differential data, by conversion
Laggard line number value, via entropy code, the data of entropy code is merged into the direction of prediction and motion-vector the data of compression
Stream.
In step S53, the data flow is carried out follow-up packing processing by the first processing equipment in front end system, and will beat
The data input of processing is wrapped into buffer;And
In step S54, the data in buffer are passed through wired or wireless chain by the first transceiver according to transmission speed
Transmit on road.
Preferably, the first transceiver is passed the data in buffer by wired or wireless link according to transmission speed
Defeated, wherein radio communication is for example logical including Wi-Fi, bluetooth (BT), near-field communication (NFC), global positioning system (GPS) and honeycomb
Letter (such as LTE, LTE-A, CDMA, WCDMA, UMTS, WiBro, GSM) at least one.Wire communication is for example including logical
With at least one in universal serial bus (USB), HDMI (HDMI), RS-232 and POTS.
By with the operation of streaming media, it is ensured that transmission quality, the continuity of data acquisition can be kept, and is saved
Time, is that the efficiency of follow-up recognition of face improves and is made that contribution.
In step s 6, back-end system receives view data and handled, and recognizes face therein.The structure chart of back-end system
As shown in Figure 4.Wherein back-end system 2 includes the second transceiver F, identification equipment G, storage device H, the 3rd transceiver I.Its
Middle step S6 further comprises:
In step S61, the second transceiver wherein in back-end system receives the image sent by the first transceiver,
And it is entered into identification equipment.
In step S62, identification equipment is decoded the data of input according to default agreement, obtains image sequence;
In step S63, identification equipment is pre-processed to image sequence, and it includes gray proces, illumination compensation, smooth
Denoising, image sharpening etc., to reduce interference signal.Wherein in order to reduce what is brought during capture, transmission due to image etc.
Noise, the present invention reduces interference using denoising mode, is that follow-up identification prepares accurate view data;
Preferably, step S63 further comprises the steps:The geological information of the subgraph in image sequence is extracted, it is all
Such as size;Build gray-scale map;Carry out the conversion of color space;Create partition holding and perform initialization operation;To gray-scale map
Bar chart handled.
In step S64, the change in image is judged, when it is determined that foreground area is more than the first critical value, then really
Surely change, and then enter step S65, and when determination foreground area is not less than the first critical value, it is determined that unchanged, terminating should
Duan Xulie operation, then carry out the operation of next sequence;
In step S65, according to the shape description at face position and they the distance between characteristic contributed to
The characteristic of face classification, this feature data include characteristic component, and this feature component includes the Euclidean distance between characteristic point, song
Rate and angle;The characteristic value formation facial characteristics matrix of face data is extracted using PCA algorithms, neural-network classification method is used
The determination of face rectangle is carried out, right and left eyes, left and right eyebrow, nose, the left and right corners of the mouth, the chin in face rectangle are determined, to rectangle
Gray-scale map is rotated and scaled, and extracts the feature at above-mentioned position.Preferably, the operation of face rectangle includes:Determine face
And mark, the N*N pixels of face are set, wherein N is positive integer;Protrude the face in image and rectangle is set, determine that rectangle is big
It is small and show.
In step S66, face information is determined, and the image array and information of face are stored into storage device H.
In the step s 7, the object in back-end system will identify that face and database is compared, and determines result
Further comprise:By the 3rd transceiver I of back-end system, send and ask via chain road direction database 3, and by human face data
It is compared with the human face data that has stored, and obtains numerical value, if the numerical value is more than the second critical value, it is determined that identification object
Identity information, otherwise return to no corresponding informance.
By above method, the result that recognition speed is fast, accuracy rate is high can be provided, and lead due to less calculating
Cause the power consumption of reduction.
Optionally, in addition, the above-mentioned face identification method based on Streaming Media further comprises:
In step s 8, back-end system is according to the view data of reception, it is determined whether need another adjustment angle, so as to
The corresponding object of face of Tracking Recognition, after determining to need to adjust, back-end system is by the second transceiver module, end system forward
The feedback information of another adjustment angle is sent, front end system is based on this and repeats to continue executing with since step S1.
To sum up, in the inventive solutions, by using a kind of face identification method based on Streaming Media.Pass through
This method, can be on the premise of equipment not be increased, it is ensured that power consumption and cost do not increase, it is ensured that is carried out the complete of three dimensions and is covered
Lid formula is captured, and is enhanced diversity and the flexibility of data, it is ensured that transmission quality, can be kept the continuity of data acquisition,
And the time has been saved, has been that the efficiency of follow-up recognition of face improves and is made that contribution, can provide that recognition speed is fast, accuracy rate is high
Result, and cause due to less calculating the power consumption of reduction.
It will be appreciated that:The example and reality of the present invention can be realized in the form of the combination of hardware, software or hardware and software
Apply example.As described above, any main body for performing this method can be stored, in the form of volatility or non-volatile holographic storage, for example
No matter storage device, as ROM, can erase or whether rewritable, or in the form of a memory, such as RAM, storage core
Piece, equipment or integrated circuit or on the readable medium of light or magnetic, such as CD, DVD, disk or tape.It will be appreciated that:
Storage device and storage medium are suitable for storing the example of the machine readable storage of one or more programs, upon being performed,
One or more of programs realize the example of the present invention.Via any medium, such as couple what is be loaded with by wired or wireless
Signal of communication, can electronically transmit the example of the present invention, and example suitably includes identical content.
It should be noted that:Because all standing formula capture that the present invention solves three dimensions is fast, accurate there is provided recognition speed
The true high recognition result of rate has simultaneously been saved the time, is reduced the technical problem of power consumption, is employed skill in field of computer technology
Art personnel instruct technological means to understand according to it after reading this description, and obtain and can not increase equipment
On the premise of, it is ensured that power consumption and cost do not increase, it is ensured that carry out all standing formula capture of three dimensions, enhance many of data
Sample and flexibility, it is ensured that transmission quality, can keep the continuity of data acquisition, and save the time, be follow-up people
The efficiency of face identification, which is improved, is made that contribution, can provide the result that recognition speed is fast, accuracy rate is high, and due to less meter
Calculate and cause the advantageous effects of the power consumption of reduction, so claimed scheme belongs to special in the following claims
Technical scheme in sharp method meaning.In addition, because the technical scheme that appended claims are claimed can be manufactured in the industry
Or use, therefore the program possesses practicality.
It is described above, it is only the preferably embodiment of the present invention, but protection scope of the present invention is not limited to
This, any one skilled in the art the invention discloses technical scope in, the change that can readily occur in or replace
Change, should all be encompassed within protection scope of the present invention.Unless be otherwise expressly recited, otherwise disclosed each feature is only
It is equivalent or similar characteristics a example for general series.Therefore, protection scope of the present invention should be with claims
Protection domain is defined.
Claims (10)
1. a kind of face identification method based on Streaming Media, comprises the following steps:
In step sl, the angle of image capture device is adjusted, and angle configuration information is sent to front end system;
In step s 2, front end system will be sent to back-end system after the information processing;
In step s3, back-end system determines the need for adjustment angle again;
In step s 4, after determining the need for adjustment, by image capture device capture images data;
In step s 5, view data is carried out Streaming Media processing by front end system, and is sent to back-end system;
In step s 6, back-end system receives view data and handled, and recognizes face therein;
In the step s 7, the object in back-end system will identify that face and database is compared, and determines result.
2. the face identification method as claimed in claim 1 based on Streaming Media, wherein in step s 2, front end system believes this
Back-end system is sent to after breath processing to further comprise:
In the step s 21, the information that front end system is captured by angle configuration information and in preset time period is handled;And
In step S22, the information of processing is sent to back-end system by front end system.
3. the face identification method as claimed in claim 2 based on Streaming Media, wherein in step s3, back-end system determination is
Adjustment angle further comprises no needs again:
Back-end system is according to the information of the processing of reception, it is determined whether need adjustment angle again, if it is determined that image capture is set
Standby C capture angle sets suitably, then directly carries out the capture of image in step s 4;If it is determined that image capture device C
Capture angle set inappropriate, that is, need to adjust capture angle again, then return to adjustment angle, and adjust in step s 4
The capture of image is carried out afterwards.
4. the face identification method as claimed in claim 3 based on Streaming Media, wherein in step sl, adjustment image capture is set
Standby angle, and angle configuration information is sent to front end system, the front end system 1 includes fixing equipment A, adjustment equipment B, figure
As capture device C, the first processing equipment D and the first transceiver E, fixing equipment A are coupled with adjustment equipment B, adjustment equipment B and
Image capture device C and the first processing equipment D are coupled respectively, and the first processing equipment D and the coupling of the first transceiver;Adjust
Section equipment B includes the connecting elements B0, rotational electric component B1, connecting elements B2, rotational electric component B3 being linked in sequence,
Connecting elements B4, rotational electric component B5 and connecting elements B6;When rotational electric component receives electricity by wire link
During signal, connecting elements is hollow structure, and it is internally provided with electric wiring, and when rotational electric component passes through radio chains
When road receives electric signal, connecting elements can be hollow structure or solid construction, now rotational electric component inside setting
There are proximity communication module and signal processing module, it leads to being again provided with the first processing equipment of proximity communication module
Letter;Connecting elements B0 connection is set to:Connecting elements B0 is directly or indirectly connected to fixing equipment A and the first processing equipment
In any one or both;Connecting elements B0 in adjustment equipment B, rotational electric component B1, connecting elements B2, rotatable electricity
Dynamic component B3, connecting elements B4, rotational electric component B5, connecting elements B6 are specifically configured to:Three rotational electric components
For cylindrical structure, it each includes multiple sub- components that can mutually move, and many sub- components can be partly or entirely
Relative to other equipment motion;Wherein mutually it is between the axis of rotational electric component B1, B3, B5 cylinder bottom surface
90 °, i.e. each in these axis is placed perpendicular to each other so that the rotary shaft of three rotational electric components is covered
All three-dimensionals of lid, i.e. X, Y and Z-direction;Structure can be wholly or partially straight, bending or angled;Enter
One step, connecting elements can use L, S, U, V, spiral shape.
5. the face identification method based on Streaming Media as claimed in claim 4, wherein in step s 5, front end system is by image
Data carry out Streaming Media processing, and are sent to back-end system and further comprise:
In step s 51, the first processing equipment in front end system receives image capture device with the image of a certain speed trap
Information;
In step S52, the data are carried out coding compression by the first processing equipment in front end system, are specifically included:First will figure
As data are divided into fragment, it are divided into sub-piece again afterwards, then sub-piece is predicted, it is in the following ways:By neighbour
The near frame re-established for having completed coding is reference frame, performs motion compensation, afterwards after the completion of, subtract current son with its numerical value
The actual numerical value of fragment, generates differential data, by converting laggard line number value, via entropy code, by the data of entropy code with
The direction of prediction and motion-vector are merged into the data flow of compression;
In step S53, the data flow is carried out follow-up packing processing by the first processing equipment in front end system, and by packing
The data input of reason is into buffer;And
In step S54, the first transceiver is passed the data in buffer by wired or wireless link according to transmission speed
It is defeated.
6. the face identification method as claimed in claim 5 based on Streaming Media, wherein in step S54, the first transceiver
According to transmission speed, by the data in buffer by wired or wireless link transmission, wherein radio communication includes Wi-Fi, blue
Tooth (BT), near-field communication (NFC), global positioning system (GPS), and including LTE, LTE-A, CDMA, WCDMA, UMTS,
At least one in WiBro, GSM cellular communication;Wire communication includes USB (USB), high-definition multimedia
At least one in interface (HDMI), RS-232 and POTS.
7. the face identification method as claimed in claim 6 based on Streaming Media, wherein in step s 6, back-end system receives figure
As data and handle, recognize that face therein further comprises:
In step S61, the second transceiver wherein in back-end system receives the image sent by the first transceiver, and
It is entered into identification equipment;
In step S62, identification equipment is decoded the data of input according to default agreement, obtains image sequence;
In step S63, identification equipment is pre-processed to image sequence, and it includes gray proces, illumination compensation, smoothly gone
Make an uproar, image sharpening etc., to reduce interference signal;
In step S64, the change in image is judged, when it is determined that foreground area is more than the first critical value, it is determined that have
Change, and then enter step S65, and when determination foreground area is not less than the first critical value, it is determined that it is unchanged, terminate the Duan Xu
The operation of row, then carry out the operation of next sequence;
In step S65, according to the shape description at face position and they the distance between characteristic to obtain contribute to face
The characteristic of classification, this feature data include characteristic component, this feature component include characteristic point between Euclidean distance, curvature and
Angle;The characteristic value formation facial characteristics matrix of face data is extracted using PCA algorithms, is carried out using neural-network classification method
The determination of face rectangle, determines right and left eyes, left and right eyebrow, nose, the left and right corners of the mouth, the chin in face rectangle, to the gray scale of rectangle
Figure is rotated and scaled, and extracts the feature at above-mentioned position;The operation of face rectangle includes:Determine face and mark, set
The N*N pixels of face, wherein N is positive integer;Protrude the face in image and rectangle is set, determine rectangle size and show;
In step S66, face information is determined, and the image array and information of face are stored into storage device H.
8. the face identification method as claimed in claim 7 based on Streaming Media, wherein back-end system 2 include the second transceiver
F, identification equipment G, storage device H, the 3rd transceiver I;
Step S63 further comprises the steps:Extract the geological information of the subgraph in image sequence, size etc.;Structure
Build gray-scale map;Carry out the conversion of color space;Create partition holding and perform initialization operation;The bar chart of gray-scale map is carried out
Processing.
9. the face identification method as claimed in claim 8 based on Streaming Media, wherein in the step s 7, back-end system will be recognized
Object in the face and database that go out is compared, and determines that result further comprises:
By the 3rd transceiver I of back-end system, send and ask via chain road direction database 3, and by human face data with having deposited
The human face data of storage is compared, and obtains numerical value, if the numerical value is more than the second critical value, it is determined that the identity of identification object
Information, otherwise returns to no corresponding informance.
10. the face identification method as claimed in claim 9 based on Streaming Media, wherein this method further comprises:
In step s 8, back-end system is according to the view data of reception, it is determined whether another adjustment angle is needed, to track
The corresponding object of face of identification, after determining to need to adjust, back-end system is sent by the second transceiver module, forward end system
The feedback information of another adjustment angle, front end system is based on this and repeats to continue executing with since step S1.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710324671.7A CN107169433B (en) | 2017-05-10 | 2017-05-10 | Face recognition method based on streaming media |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710324671.7A CN107169433B (en) | 2017-05-10 | 2017-05-10 | Face recognition method based on streaming media |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107169433A true CN107169433A (en) | 2017-09-15 |
CN107169433B CN107169433B (en) | 2020-05-19 |
Family
ID=59813698
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710324671.7A Expired - Fee Related CN107169433B (en) | 2017-05-10 | 2017-05-10 | Face recognition method based on streaming media |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107169433B (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2013090471A1 (en) * | 2011-12-13 | 2013-06-20 | Google, Inc. | Processing media streams during a multi-user video conference |
CN103310200A (en) * | 2013-06-25 | 2013-09-18 | 郑州吉瑞特电子科技有限公司 | Face recognition method |
CN103324919A (en) * | 2013-06-25 | 2013-09-25 | 郑州吉瑞特电子科技有限公司 | Video monitoring system based on face recognition and data processing method thereof |
CN104036236A (en) * | 2014-05-27 | 2014-09-10 | 厦门瑞为信息技术有限公司 | Human face gender recognition method based on multi-parameter index weighting |
CN104135645A (en) * | 2014-07-31 | 2014-11-05 | 天津市亚安科技股份有限公司 | Video surveillance system and method for face tracking and capturing |
-
2017
- 2017-05-10 CN CN201710324671.7A patent/CN107169433B/en not_active Expired - Fee Related
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2013090471A1 (en) * | 2011-12-13 | 2013-06-20 | Google, Inc. | Processing media streams during a multi-user video conference |
CN103310200A (en) * | 2013-06-25 | 2013-09-18 | 郑州吉瑞特电子科技有限公司 | Face recognition method |
CN103324919A (en) * | 2013-06-25 | 2013-09-25 | 郑州吉瑞特电子科技有限公司 | Video monitoring system based on face recognition and data processing method thereof |
CN104036236A (en) * | 2014-05-27 | 2014-09-10 | 厦门瑞为信息技术有限公司 | Human face gender recognition method based on multi-parameter index weighting |
CN104135645A (en) * | 2014-07-31 | 2014-11-05 | 天津市亚安科技股份有限公司 | Video surveillance system and method for face tracking and capturing |
Non-Patent Citations (1)
Title |
---|
张民安 主编: "《场所隐私权研究:场所隐私权理论的产生、发展、确定和具体适用》", 31 August 2016, 广州:中山大学出版社 * |
Also Published As
Publication number | Publication date |
---|---|
CN107169433B (en) | 2020-05-19 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11671712B2 (en) | Apparatus and methods for image encoding using spatially weighted encoding quality parameters | |
CN107808136A (en) | Image processing method, device, readable storage medium storing program for executing and computer equipment | |
CN103098077B (en) | Real-time video frame Pre-processing hardware | |
CN110334635A (en) | Main body method for tracing, device, electronic equipment and computer readable storage medium | |
JP6662745B2 (en) | Imaging device, information processing system, and polarization image processing method | |
JP2020187753A (en) | Training system for training generator neural network | |
CN109788170A (en) | It is a kind of based on infrared with video image processing system and method for visible light | |
US10846551B2 (en) | Video data processing | |
CN108846807A (en) | Light efficiency processing method, device, terminal and computer readable storage medium | |
CN106165386A (en) | For photo upload and the automatic technology of selection | |
US11695890B2 (en) | Image capture device with an automatic image capture capability | |
US20220059132A1 (en) | Event/object-of-interest centric timelapse video generation on camera device with the assistance of neural network input | |
CN108764321B (en) | Image-recognizing method and device, electronic equipment, storage medium | |
CN108717530A (en) | Image processing method, device, computer readable storage medium and electronic equipment | |
US20190251670A1 (en) | Electronic device and method for correcting images using external electronic device | |
CN107145870A (en) | The identifying system of face in a kind of video | |
CN108616700A (en) | Image processing method and device, electronic equipment, computer readable storage medium | |
CN108111768A (en) | Control method, apparatus, electronic equipment and the computer readable storage medium of focusing | |
CN108848306A (en) | Image processing method and device, electronic equipment, computer readable storage medium | |
CN110121031A (en) | Image-pickup method and device, electronic equipment, computer readable storage medium | |
CN107845076A (en) | Image processing method, device, computer-readable recording medium and computer equipment | |
CN107437268A (en) | Photographic method, device, mobile terminal and computer-readable storage medium | |
CN116721141A (en) | Adaptive facial depth map generation | |
CN109325905A (en) | Image processing method, device, computer readable storage medium and electronic equipment | |
CN109121194B (en) | Method and apparatus for state transition of electronic device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200519 Termination date: 20210510 |