CN206224639U - A kind of face recognition door control system with occlusion detection function - Google Patents
A kind of face recognition door control system with occlusion detection function Download PDFInfo
- Publication number
- CN206224639U CN206224639U CN201621220013.0U CN201621220013U CN206224639U CN 206224639 U CN206224639 U CN 206224639U CN 201621220013 U CN201621220013 U CN 201621220013U CN 206224639 U CN206224639 U CN 206224639U
- Authority
- CN
- China
- Prior art keywords
- module
- face
- recognition
- occlusion detection
- operating system
- 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.)
- Expired - Fee Related
Links
Abstract
The utility model discloses a kind of face recognition door control system with occlusion detection function, it is related to computer vision and area of pattern recognition, including:Camera module, operating system control module, occlusion detection module, face recognition module, recognition result display module, fingerprint login module, face acquisition module.Occlusion detection module can be quickly detected from the occlusion area of face using Markov random field, and eliminate the trouble that user extractd the shelters such as glasses before image is gathered, face recognition module is using sparse coding to facial image Classification and Identification, the fingerprint login module person's of needing to use typing fingerprint consistent with its identity after recognition of face success.The utility model passes through occlusion detection module, eliminate the trouble that user extracts shelter, the operating efficiency of the gate control system based on recognition of face is improve, and the speed of detection is improve using Markov random field model, the security of gate control system is ensure that finally by fingerprint login module.
Description
Technical field
The utility model is related to computer vision and area of pattern recognition, and in particular to a kind of with occlusion detection function
Face recognition door control system.
Background technology
With the development of social economy and science and technology, gate control system is more and more applied to every field, particularly handles official business
The turnover to personnel such as building and concerning security matters place is more strict.On the one hand, it is more widely key, electronics magnetic to use at present
Clamping lock and coded lock etc., but these instruments all have the shortcomings that to be easily lost and be replicated, and face has uniqueness and not
Reproducibility, therefore for gate control system, recognition of face is a kind of more promising technology.On the other hand, current people
Face identifying system is highly susceptible to the influence of shelter, therefore is required for identified person before recognition by the scarf and eye of face
The jewelrys such as mirror are extractd, particularly outdoor gate control system, and in outdoor people, summer needs to wear sunglasses protect eyes and winter
By wearing, scarf is warming for meeting, so this brings many inconvenience to user, also have impact on the operating efficiency of gate control system.
Utility model content
The purpose of this utility model is to solve drawbacks described above of the prior art, there is provided one kind has occlusion detection work(
The face recognition door control system of energy, can efficiently solve the safety problem of gate control system and be brought to user because face is blocked
The problem of inconvenience.
The purpose of this utility model can be reached by adopting the following technical scheme that:
A kind of face recognition door control system with occlusion detection function, including:The face recognition door control system includes:
Camera module 6, operating system control module 2, occlusion detection module 3, face recognition module 4, recognition result display module 5,
Fingerprint login module 7;
The camera module 6, is connected with the operating system control module 2, for gathering facial image, the face
Image is have to block or unobstructed user's facial image;
The operating system control module 2, is connected with the occlusion detection module 3, and the camera module 6 is gathered
Facial image passes to the occlusion detection module 3;
The occlusion detection module 3, the facial image to gathering does occlusion detection, the portion being blocked in detection facial image
Point, the part do not blocked in the image is obtained as facial image to be identified, and passes to coupled face knowledge
Other module 4;
The face recognition module 4, is bi-directionally connected with the operating system control module 2, on the one hand receives the operation
The face dictionary used in recognition of face that system control module 2 is passed over, on the other hand, controls to the operating system
Module 2 transmits face recognition result;
The recognition result display module 5, is connected with the operating system control module 2, for by the recognition of face
Module 4 recognizes that the identity for obtaining shows, and controls the coupled fingerprint login module 7 to gather user fingerprints;
The fingerprint login module 7, is bi-directionally connected with the operating system control module 2, the fingerprint for gathering user,
And obtain the identity of the face login result whether consistent with the identity of fingerprint.
Further, the face recognition door control system also includes face acquisition module 1, for gathering the unobstructed of user
Face image, processes and forms display image by classification, and passes to the coupled operating system control module 2 general
The facial image of collection is stored as dictionary.
Further, the occlusion detection module 3 goes out the occlusion area of face using Markov random field quick detection.
Further, the face recognition module 4 carries out Classification and Identification using sparse coding to facial image.
The utility model has the following advantages and effect relative to prior art:
The utility model passes through occlusion detection module so that gate control system can directly gather the face figure of shelter
Picture, does not on the one hand need user's extra band door lock, and the user that on the other hand omits is removing the trouble of shelter outdoors, very
The user of gate control system is facilitated in big degree, and the Markov model that occlusion detection module is used further is improved
The efficiency of gate control system, by increasing fingerprint login module corresponding with user's identity, prevents by obtaining other people figures
The mode of piece is the behavior of gate control system or by way of making other people fingerprint moulds by the behavior of gate control system, it is ensured that door
The security of access control system.
Brief description of the drawings
Fig. 1 is a kind of structural representation of the face recognition door control system with occlusion detection function disclosed in the utility model
Figure;
Fig. 2 is a kind of workflow of the face recognition door control system with occlusion detection function disclosed in the utility model
Figure;
Fig. 3 is the workflow diagram of occlusion detection module in the utility model;
Fig. 4 is the workflow diagram of face recognition module in the utility model.
Specific embodiment
It is new below in conjunction with this practicality to make the purpose, technical scheme and advantage of the utility model embodiment clearer
Accompanying drawing in type embodiment, is clearly and completely described, it is clear that retouched to the technical scheme in the utility model embodiment
The embodiment stated is a part of embodiment of the utility model, rather than whole embodiments.Based on the implementation in the utility model
Example, the every other embodiment that those of ordinary skill in the art are obtained under the premise of creative work is not made is belonged to
The scope of the utility model protection.
Embodiment
As shown in Figure 1, a kind of face recognition door control system with occlusion detection function includes disclosed in the present embodiment
Face acquisition module 1, operating system control module 2, occlusion detection module 3, face recognition module 4, recognition result display module
5, fingerprint login module 7 and camera module 6.
First, to utilize the unobstructed positive face of the collection user of face acquisition module 1 to shine, process and form aobvious by classification
Diagram picture, is stored the facial image of collection as dictionary by operating system control module 2.Camera module 6 is direct
Collection facial image, without extracing the shelters such as glasses and scarf, occlusion detection module 3 is done to the facial image that camera is gathered
Occlusion detection, the part being blocked in detection face, obtains the part do not blocked in the image as facial image to be identified,
Recognition result display module 5 will recognize that the identity for obtaining shows that fingerprint login module 7 is to make by face recognition module 4
After the identity of user is identified, it is desirable to which user is input into fingerprint corresponding with its identity, operating system control module 2
Output end is connected with the input of face recognition module 4, is used to transmit the face word that face recognition module 4 needs in identification
Recognition result is out sent to operating system control module 2, i.e. operating system control module 2 by allusion quotation, result to be identified again later
Input be connected with the output end of face recognition module 4, another input and the fingerprint of operating system control module 2 are stepped on
Record the output end of module 7 in succession, be used to obtain the login result of fingerprint login module 7.
With reference to accompanying drawing 2, to a kind of the specific of face recognition door control system with occlusion detection function of the present utility model
Workflow is described further, and comprises the following steps that:
Step one:The unobstructed positive face for gathering user in advance by the camera of face acquisition module 1 shines, and completes to make
The Data Enter of user, on the one hand by the Face image synthesis display image of typing, on the other hand as the dictionary of recognition of face,
Enter step 2 afterwards.
Step 2:User before entering, first passes through the collection facial image of camera module 6 outdoors, in this step
In rapid, user can directly go to and adopted by camera before camera without in advance extracing oneself glasses and scarf etc.
Collection image, and the image of collection is transferred to occlusion detection module 3 by operating system control module 2, afterwards into step 3.
Step 3:Occlusion detection, as shown in Figure 3, is blocked using markov random file to the test face sample
Region modeling and iteration renewal, make last numerical convergence, obtain the non-occluded area of face.Its specific calculating process is such as
Under:
S301:Assuming that the image to be identified obtained by step 3 is y, it is assumed that a figure of people of typing in step one
Picture, everyone has b a total of n=a × b of sample image of sample image, i.e. training, and the dimension of every image is m=f
×g.Every training sample image is converted to the column vector of m × 1 by f × g, and generates a standard chemical handwriting practicing allusion quotation
S302:Assuming that label vector isEach component identification respective pixel of s whether be occlusion area picture
Element, wherein s [j]=0 represent that j elements are occlusion area pixels, and s [j]=1 represents that j elements are non-occluded area pixels.S's
All elements are initialized as 1;
S303:According to current region labeling, by the way that following formula is by the non-occluded area in standard chemical handwriting practicing allusion quotation and treats
Recognize that the non-occluded area of face sample is extracted:
A*=A [st-1=1,:]y*=y [st-1=1]
In formula, A*Represent the non-occluded area of standard chemical handwriting practicing allusion quotation, y*Represent the unshielede region of face sample to be identified
Domain, st-1=1 expression takes the pixel that current label vector intermediate value is 1, that is, take the non-occluded area for detecting at present;
S304:The non-occluded area of face sample to be identified and the unshielede region of standard chemical handwriting practicing allusion quotation are calculated using following formula
Similarity between domain, and generate apart from dictionary
D=[d1, d2..., dn]
In formula, Ai *Represent the non-occluded area of the sub- dictionary of the i-th class training sample image composition, EuclideanY is sought in expression*And A*Between Euclidean distance, σ is a constant, for the control weight rate of decay;
S505:Local restriction is applied to code coefficient, face sample is detected by solving the object function in following formula
The code coefficient of non-shield portions:
In formula, | | | |2Represent the l of vector2Norm, x presentation code coefficients, D represented apart from dictionary, before λ is a balance
The positive number of two afterwards;
S306:Be reconstructed according to the code coefficient obtained and study dictionary, obtain correspondence reconstructed image and between weight
Structure error:
S307:Markov random field model is set up to label vector s, the target letter in Algorithm for Solving following formula is cut using figure
Number, updates label vector s:
Value as described in following formula:
In formula, s [i] is the label value of ith pixel,It is the reconstructed error value of ith pixel, μ is a constant,
The intensity of the interphase interaction of different pixels is controlled, τ is a threshold value;
S308:Iteration performs S103-S107, until algorithmic statement or reaching maximum convergence number of times
S309:The rational band occlusion area of facial image to be identified has been detected, and exports facial image to be identified
Non-occluded area.Enter step 4 afterwards.
Step 4:Recognition of face, in this step, using solution l2The method of constrained minimization, it is comprised the following steps that:
S401:The input non-shield portions of images to be recognized.
S401:L is carried out to the non-occluded area part of facial image y to be identified by solving the object function in following formula2
Sparse coding:
In formula, | | | |2Represent the l of vector2Norm, x represents sparse coding coefficient, and θ is one and is balancing front and rear two just
Number,Represent l2Sparse coding coefficient.
S402:Using the sparse coding for solving, the weight between such corresponding reconstructed image and original image is obtained by class
Structure error:
In formula,Represent the reconstructed error of the i-th class, Ai *Represent not hiding for the sub- dictionary that the i-th class training sample image is constituted
Gear region,Represent the sparse coding coefficient of the i-th class.
S403:CalculateSparse concentration index SCI and judge its value whether more than setting threshold value, if less than if this treat
Identification face is not belonging to any kind, i.e., the non-internal staff in dictionary, will refuse its entrance, if being more than or equal to threshold value, enters
Enter step S406
Wherein, k represents the classification number of training sample.
S406:According to the reconstructed error of each class, selection has the class of minimal reconstruction error as the classification of y:
In formula,The reconstructed error of the i-th class is represented,Represent the sparse coding coefficient of the i-th class.
S407:The identity of images to be recognized is exported, afterwards into step 5.
Step 5:Show and be input into fingerprint.If that identify is non-internal staff in the recognition result of step 4,
Then show that the non-internal staff of refusal enters, if recognized successfully, show the facial image of its typing in advance, and require that it is input into
Corresponding fingerprint, if fingerprint mistake, refuses its entrance, if fingerprint is correct, goalkeeper opens allows it to enter.
Above-mentioned occlusion detection module is modeled simultaneously using Markov random field to the facial image occlusion area to be identified
Iteration updates, until final detection is completed, substantially increases the speed of detection, and face recognition module utilizes the mesh with local restriction
Scalar functions obtain the local restriction coding of the test non-shield portions of face sample.
The utility model passes through occlusion detection module so that gate control system can directly gather the face figure of shelter
Picture, does not on the one hand need user's extra band door lock, and the user that on the other hand omits is removing the trouble of shelter outdoors, very
The user of gate control system is facilitated in big degree, and the Markov model that occlusion detection module is used further is improved
The efficiency of system, by increasing fingerprint login module corresponding with user's identity, prevents by obtaining other people pictures
Mode is the behavior of gate control system or by way of making other people fingerprint moulds by the behavior of gate control system, it is ensured that gate inhibition is
The security of system.
Above-described embodiment is the utility model preferably implementation method, but implementation method of the present utility model is not by above-mentioned
The limitation of embodiment, it is other it is any without departing from the change made under Spirit Essence of the present utility model and principle, modify, replace
Generation, combination, simplification, should be equivalent substitute mode, be included within protection domain of the present utility model.
Claims (4)
1. a kind of face recognition door control system with occlusion detection function, it is characterised in that including:The recognition of face gate inhibition
System includes:Camera module (6), operating system control module (2), occlusion detection module (3), face recognition module (4), knowledge
Other result display module (5), fingerprint login module (7);
The camera module (6), is connected, with the operating system control module (2) for gathering facial image, the face
Image is have to block or unobstructed user's facial image;
The operating system control module (2), is connected with the occlusion detection module (3), by the camera module (6) collection
Facial image pass to the occlusion detection module (3);
The occlusion detection module (3), the facial image to gathering does occlusion detection, the portion being blocked in detection facial image
Point, the part do not blocked in the image is obtained as facial image to be identified, and passes to coupled face knowledge
Other module (4);
The face recognition module (4), is bi-directionally connected, on the one hand, the recognition of face with the operating system control module (2)
Module (4) receives the face dictionary used in recognition of face that the operating system control module (2) passes over, the opposing party
Face, the face recognition module (4) transmits face recognition result to the operating system control module (2);
The recognition result display module (5), is connected with the operating system control module (2), for by the recognition of face
The identity that module (4) identification is obtained shows, and controls the coupled fingerprint login module (7) collection user to refer to
Line;
The fingerprint login module (7), is bi-directionally connected, the fingerprint for gathering user with the operating system control module (2),
And obtain the identity of the face login result whether consistent with the identity of fingerprint.
2. a kind of face recognition door control system with occlusion detection function according to claim 1, it is characterised in that
The face recognition door control system also includes face acquisition module (1), the unobstructed face image for gathering user, warp
Cross classification to process and form display image, and pass to the people that the coupled operating system control module (2) will gather
Face image is stored as dictionary.
3. a kind of face recognition door control system with occlusion detection function according to claim 1, it is characterised in that
The occlusion detection module (3) goes out the occlusion area of face using Markov random field quick detection.
4. a kind of face recognition door control system with occlusion detection function according to claim 1, it is characterised in that
The face recognition module (4) carries out Classification and Identification using sparse coding to facial image.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201621220013.0U CN206224639U (en) | 2016-11-14 | 2016-11-14 | A kind of face recognition door control system with occlusion detection function |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201621220013.0U CN206224639U (en) | 2016-11-14 | 2016-11-14 | A kind of face recognition door control system with occlusion detection function |
Publications (1)
Publication Number | Publication Date |
---|---|
CN206224639U true CN206224639U (en) | 2017-06-06 |
Family
ID=58792240
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201621220013.0U Expired - Fee Related CN206224639U (en) | 2016-11-14 | 2016-11-14 | A kind of face recognition door control system with occlusion detection function |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN206224639U (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108805179A (en) * | 2018-05-24 | 2018-11-13 | 华南理工大学 | One kind being based on face local restriction encoded calibration recognition methods |
CN110232369A (en) * | 2019-06-20 | 2019-09-13 | 深圳和而泰家居在线网络科技有限公司 | A kind of face identification method and electronic equipment |
CN111488811A (en) * | 2020-03-31 | 2020-08-04 | 长沙千视通智能科技有限公司 | Face recognition method and device, terminal equipment and computer readable medium |
CN111768543A (en) * | 2020-06-29 | 2020-10-13 | 杭州翔毅科技有限公司 | Traffic management method, device, storage medium and device based on face recognition |
CN111862413A (en) * | 2020-07-28 | 2020-10-30 | 公安部第三研究所 | Method and system for realizing epidemic situation resistant non-contact multidimensional identity rapid identification |
CN113269137A (en) * | 2021-06-18 | 2021-08-17 | 常州信息职业技术学院 | Non-fit face recognition method combining PCANet and shielding positioning |
WO2021203718A1 (en) * | 2020-04-10 | 2021-10-14 | 嘉楠明芯(北京)科技有限公司 | Method and system for facial recognition |
-
2016
- 2016-11-14 CN CN201621220013.0U patent/CN206224639U/en not_active Expired - Fee Related
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108805179A (en) * | 2018-05-24 | 2018-11-13 | 华南理工大学 | One kind being based on face local restriction encoded calibration recognition methods |
CN108805179B (en) * | 2018-05-24 | 2022-03-29 | 华南理工大学 | Face local constraint coding based calibration and recognition method |
CN110232369A (en) * | 2019-06-20 | 2019-09-13 | 深圳和而泰家居在线网络科技有限公司 | A kind of face identification method and electronic equipment |
CN110232369B (en) * | 2019-06-20 | 2021-10-01 | 深圳数联天下智能科技有限公司 | Face recognition method and electronic equipment |
CN111488811A (en) * | 2020-03-31 | 2020-08-04 | 长沙千视通智能科技有限公司 | Face recognition method and device, terminal equipment and computer readable medium |
CN111488811B (en) * | 2020-03-31 | 2023-08-22 | 长沙千视通智能科技有限公司 | Face recognition method, device, terminal equipment and computer readable medium |
WO2021203718A1 (en) * | 2020-04-10 | 2021-10-14 | 嘉楠明芯(北京)科技有限公司 | Method and system for facial recognition |
CN111768543A (en) * | 2020-06-29 | 2020-10-13 | 杭州翔毅科技有限公司 | Traffic management method, device, storage medium and device based on face recognition |
CN111862413A (en) * | 2020-07-28 | 2020-10-30 | 公安部第三研究所 | Method and system for realizing epidemic situation resistant non-contact multidimensional identity rapid identification |
CN113269137A (en) * | 2021-06-18 | 2021-08-17 | 常州信息职业技术学院 | Non-fit face recognition method combining PCANet and shielding positioning |
CN113269137B (en) * | 2021-06-18 | 2023-10-31 | 常州信息职业技术学院 | Non-matching face recognition method combining PCANet and shielding positioning |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN206224639U (en) | A kind of face recognition door control system with occlusion detection function | |
CN108537743B (en) | Face image enhancement method based on generation countermeasure network | |
CN101593272B (en) | Human face feature positioning method based on ASM algorithm | |
CN106570464A (en) | Human face recognition method and device for quickly processing human face shading | |
CN105138998B (en) | Pedestrian based on the adaptive sub-space learning algorithm in visual angle recognition methods and system again | |
CN110472519B (en) | Human face in-vivo detection method based on multiple models | |
CN106503687A (en) | The monitor video system for identifying figures of fusion face multi-angle feature and its method | |
CN106909938B (en) | Visual angle independence behavior identification method based on deep learning network | |
CN106919921B (en) | Gait recognition method and system combining subspace learning and tensor neural network | |
CN204331744U (en) | 3 D stereo intelligent face recognition system | |
CN108875907A (en) | A kind of fingerprint identification method and device based on deep learning | |
CN108960171B (en) | Method for converting gesture recognition into identity recognition based on feature transfer learning | |
CN111104852B (en) | Face recognition technology based on heuristic Gaussian cloud transformation | |
CN112200176B (en) | Method and system for detecting quality of face image and computer equipment | |
CN114549850B (en) | Multi-mode image aesthetic quality evaluation method for solving modal missing problem | |
CN113449704B (en) | Face recognition model training method and device, electronic equipment and storage medium | |
CN111178130A (en) | Face recognition method, system and readable storage medium based on deep learning | |
CN111507320A (en) | Detection method, device, equipment and storage medium for kitchen violation behaviors | |
CN205644823U (en) | Social security self -service terminal device | |
Mahesh et al. | Real-Time Eye Blinking for Password Authentication | |
CN103942545A (en) | Method and device for identifying faces based on bidirectional compressed data space dimension reduction | |
Bekzod | Face recognition based automated student attendance system | |
CN114937298A (en) | Micro-expression recognition method based on feature decoupling | |
CN109886251A (en) | A kind of recognition methods again of pedestrian end to end guiding confrontation study based on posture | |
CN109522865A (en) | A kind of characteristic weighing fusion face identification method based on deep neural network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
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: 20170606 Termination date: 20201114 |