CN108595937A - Micro-sensing intelligent identity authentication method based on behavior characteristics - Google Patents
Micro-sensing intelligent identity authentication method based on behavior characteristics Download PDFInfo
- Publication number
- CN108595937A CN108595937A CN201810392697.XA CN201810392697A CN108595937A CN 108595937 A CN108595937 A CN 108595937A CN 201810392697 A CN201810392697 A CN 201810392697A CN 108595937 A CN108595937 A CN 108595937A
- Authority
- CN
- China
- Prior art keywords
- user
- sample
- dynamic gesture
- probability
- training
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
- G06F21/31—User authentication
- G06F21/36—User authentication by graphic or iconic representation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06V40/20—Movements or behaviour, e.g. gesture recognition
- G06V40/28—Recognition of hand or arm movements, e.g. recognition of deaf sign language
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C9/00—Individual registration on entry or exit
- G07C9/00174—Electronically operated locks; Circuits therefor; Nonmechanical keys therefor, e.g. passive or active electrical keys or other data carriers without mechanical keys
- G07C9/00563—Electronically operated locks; Circuits therefor; Nonmechanical keys therefor, e.g. passive or active electrical keys or other data carriers without mechanical keys using personal physical data of the operator, e.g. finger prints, retinal images, voicepatterns
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Security & Cryptography (AREA)
- Computer Hardware Design (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Psychiatry (AREA)
- Social Psychology (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Collating Specific Patterns (AREA)
Abstract
The intelligent identity authentication method based on the dynamic gesture is high in precision and good in user experience. The technical scheme comprises the following steps: 1. the sample training part is used for acquiring a plurality of times of dynamic gesture information of a set user and finally forming a training sample; and calculating a judgment model and a judgment threshold by using the training sample. 2. The identity authentication part is used for acquiring primary dynamic gesture information of a user as a sample to be identified; and calculating the probability of the sample to be identified under the judgment model, if the probability is greater than the judgment threshold, the user is considered as the set user, otherwise, the user is judged not to be the set user. The invention innovatively provides a scheme for identity authentication by using dynamic gestures, and the authentication process is simpler and more convenient on the premise of ensuring higher authentication accuracy.
Description
Technical field
The present invention relates to identity identifying technology fields, specifically a kind of to realize identity using user's dynamic gesture feature
The method of certification.
Background technology
Traditional identification system (such as identity card, IC card etc.) proved based on third party, security performance is more crisp
Weak, proof of identification is easy to be usurped or falsely used by criminal.And the identification system based on biological characteristic, although in safety
Property aspect compared with increasing for the identification system proved based on third party, but it is in the convenience used, work it is steady
There are still many shortcomings in terms of qualitative and hardware and software cost.
The current more mature identity authorization system based on biological characteristic mainly has:Identity authorization system based on fingerprint
[Peralta D,Galar M,Triguero I,et al.A survey on fingerprint minutiae-based
local matching for verification and identification[J].Information Sciences,
2015,315(C):67-87.], identity authorization system [Huang P, Gao G, Qian C, the et al.Fuzzy based on face
Linear Regression Discriminant Projection for Face Recognition[J].IEEE
Access,2017,5(99):4340-4349.], identity authorization system [the Wakiyama K.iris image based on iris
pickup camera and iris authentication system:EP,EP1696382[P].2006.].To be based on referring to
For the identity authorization system of line, the identification process of fingerprint is relatively complicated, needs user right under the premise of ensureing finger cleaning
Quasi- collecting device pressing, inconvenient for use and recognition success rate are general.In addition, some current technologies have been able to relatively easily
Other people fingerprints are falsely used in realization, and the safety of the identity authorization system based on fingerprint receives larger challenge.In addition, being based on
The identity authorization system of face is there is also similar problem, and recognition of face is to the facial expression of user, accessories and illumination condition
There is more harsh requirement, recognition success rate is not high, and the mode of recognition of face is readily available the modes such as photo, video recording
It is cheated, security performance is poor.And the identity authorization system based on iris, need the eye of user to make close to collecting device
With inconvenience, and the hardware cost of iris identification equipment is high, it is difficult to large-scale promotion.
Invention content
The purpose of the present invention is higher to design a kind of precision, and the intelligent body based on dynamic gesture of better user experience
Identity authentication method.
To achieve the above object, the present invention adopts the following technical scheme that:
1, sample training part
The dynamic gesture information several times of acquisition setting user, eventually forms training sample;
Using training sample, discrimination model and decision threshold are calculated.
2, authentication part
A dynamic gesture information for acquiring user, as sample to be identified;Sample to be identified is calculated in discrimination model
The probability of lower appearance, if probability is more than decision threshold, then it is assumed that the user is setting user, conversely, then judging the user
It is not setting user.
Particularly, dynamic gesture information includes:Acceleration and angular speed of the dynamic gesture in tri- directions X, Y, Z;
Particularly, it calculates discrimination model and decision threshold includes the following steps:Calculate the HMM corresponding to all training samples
The average value of (Hidden Markov Model, hidden Markov model) calculates all training samples and is sentencing as discrimination model
The average value of the probability certainly occurred under model is as decision threshold;
Particularly, it further includes following step to form training sample:It is right that each sample institute is calculated using Baum Welch algorithms
The HMM model answered, and its probability that occurs under HMM model.
The present invention also provides another technical solutions:
A kind of micro- sense intelligent identity identification system of Behavior-based control feature, which is characterized in that including identity information acquisition mould
Block and data processing module, which is characterized in that identity information acquisition module and data processing module are realized described in claim 1
Method
Beneficial effects of the present invention:The present invention innovatively proposes the scheme that authentication is carried out using dynamic gesture,
Under the premise of ensureing higher certification accuracy, keep verification process easier.In addition, company of the present invention according to dynamic gesture
Continuous property feature carries out sample training and authentication using improved HMM algorithms, effectively improves the accuracy of certification.
Description of the drawings
Fig. 1 is the Intelligent door control system structure chart realized based on the present invention;
Fig. 2 is the functional schematic for the Intelligent door control system mobile phone terminal application program realized based on the present invention;
Fig. 3 is the flow chart of the sample training and authentication procedures of the present invention;
Specific implementation mode
Below by a micro- sense identification Intelligent door control system realized using the present invention is introduced, this is further illustrated
The substantive content of invention, but present disclosure is without being limited thereto.
Micro- sense identification Intelligent door control system based on dynamic gesture, as shown in Figure 1, the system include intelligent door lock with
Smart mobile phone.Wherein, intelligent door lock adds a door lock data processing module, a wireless communication using existing electromagnetic lock again
Module.Door lock data processing module based on existing microcontroller development board realize, such as can utilize buy in the market based on
The development board of the ARM microcontrollers either development board etc. based on 51 microcontrollers.Wireless communication module can be utilized and be bought in the market
The realizations such as the bluetooth communication arrived either Wi-Fi communication modules.
Mobile phone terminal processing software is housed, which utilizes technical scheme of the present invention independent development, function on smart mobile phone
Schematic diagram is as shown in Figure 2.The mobile phone terminal processing software includes 4 submodules:Subscriber interface module, data in mobile phone processing module,
Data memory module, communication and encrypting module.Wherein, subscriber interface module provides use for realizing good human-computer interaction
Family operates the visualization interface of the mobile phone terminal processing software, and user can be by clicking the different buttons in subscriber interface module to journey
Sequence assigns different instruction, and the then instruction can be sent to data in mobile phone processing module, and instruction is main (to inform hand including register instruction
The data type that machine data processing module receives is registration dynamic gesture information), unlock instruction (inform data in mobile phone processing module
The data type of reception is unlock dynamic gesture information);Data in mobile phone processing module is having received from subscriber interface module
It can start the dynamic gesture information of acquisition corresponding types after instruction according to different instruction, the dynamic gesture information is by smart mobile phone
Acceleration transducer, the angular-rate sensor set, touch screen provide, if corresponding is registration dynamic gesture information, mobile phone
Data processing module will utilize the information architecture HMM model and calculate correlation model parameters, and model parameter is then sent to number
According to memory module, if corresponding is unlock dynamic gesture information, data in mobile phone processing module will be received and be stored from data
Then the previously stored model parameter data of module is identified unlock dynamic gesture information and sends court verdict information
To communication and encrypting module;The court verdict information that communication is sent with encrypting module primary recipient data in mobile phone processing module, and
The key information sent using data memory module is sent to intelligent door lock after court verdict information is encrypted;Data store
Module is mainly used for receiving, stores and send Various types of data, including key information, model parameter.
Generally, whole experiment process can be divided into sample training and authentication procedures, as shown in Figure 3:
During sample training:
Registration dynamic gesture data input is carried out first.A certain setting user holds mobile phone, clicks in subscriber interface module
Registration button, start dynamic gesture information gathering process, acquisition time length with finger pressing screen be start, lift for knot
Beam.During start to finish, user does customized gesture motion, such as draws circle in the air, draws 8-shaped, hand during acquisition
Machine data processing module collect come from acceleration transducer built in mobile phone, gyroscope tri- directions X, Y, Z acceleration
And the pressing position information of angular velocity information and the user's finger on the touchscreen, to obtain the registration dynamic of the user
Gesture information.Registration dynamic gesture information only includes the action message of setting user.But for accurate, the need of recognition result
It sets user and repeats same action repeatedly, obtain multiple registration dynamic gesture information.The registration dynamic gesture obtained every time
Information forms the time series data of one 7 dimension, 7 dimension sequences include three-dimensional acceleration information, three-dimensional angular velocity information and
One-dimensional location information.In addition, the pressing position of user's finger on the touchscreen, because user is accustomed to difference, so position has
Difference, so as to be utilized as identification feature.
Then dynamic gesture quality testing is carried out.After the completion of setting user by multiple dynamic gesture data input, mobile phone
Data processing module corresponds to collected sequence to each dynamic gesture first and checks, it is too short or long to exclude sequence length
Dynamic gesture, and utilize dynamic time warping algorithm [Salvador S, Chan P.Toward accurate dynamic
time warping in linear time and space[J].Intelligent Data Analysis,2007,11
(5):561-580.] certain dynamic gesture is compared with its previous dynamic gesture, if difference is excessive, by the secondary dynamic hand
Gesture is eliminated.If dynamic gesture is qualified after testing, which will just be saved as a training sample.It is sent out by experiment
Existing, can not only required precision have been met but also user will not be caused defeated by choosing 10 training samples (the corresponding dynamic gesture for acquiring 10 times)
Enter it is excessive, the problem of to bring experience sense to decline.
Followed by data in mobile phone processing module carries out quantization and dimension-reduction treatment to all training samples, obtains each training
The corresponding feature vector of sample.If the obtained corresponding feature vector of m-th of training sample is expressed as T is determined by the time span of training sample.It recycles
Baum Welch algorithms calculate the HMM model corresponding to m-th of training sample, and the input value of the algorithm is corresponding feature vector
Om, in addition, being adopted to observation state transition probability matrix A, hidden state transition probability matrix B and initial state probabilities matrix PI
With the mode of equally distributed initialization.After successive ignition operates, the output of Baum Welch algorithms and feature vector OmIt is right
The matrix A answeredm、BmAnd PIm(i.e. observation state transition probability matrix A, hidden state transition probability matrix B and original state are general
The final value of rate matrix PI), these three matrixes have just corresponded to the HMM model under m-th of training sample.Then, forwards algorithms are utilized
Probability of occurrence of each training sample under its corresponding HMM model is calculated, then the probability of occurrence of all training samples is summed
After being averaged, a certain coefficient is multiplied by as decision threshold, wherein the coefficient determines the relationship between false alarm rate and false dismissed rate, i.e.,
Coefficient is bigger, and corresponding false dismissed rate is bigger, and false alarm rate is smaller, and general coefficient value is between 0.8-1.2 in experiment.Meanwhile it will
The corresponding initial state probabilities matrix PI of all training samples, observation state transition probability matrix A and hidden state transition probability
Matrix B is averaged respectively, obtains matrix PI*, A*, B*, as discrimination model parameter, (HMM that these three parameters are constituted is known as sentencing
Certainly model).In addition, discrimination model parameter will be sent to data memory module with decision threshold.
In authentication procedures:
It is unlocked dynamic gesture data input first.The basic phase of process of its mode and registration dynamic gesture data input
Seemingly, in addition to only needing a gesture information here.
Then data in mobile phone processing module will unlock dynamic gesture information, after identical quantization is operated with dimensionality reduction, then
Its probability of occurrence under discrimination model is calculated using forwards algorithms, and court verdict is exported after being compared with decision threshold,
If probability of occurrence is more than thresholding, user's authentication success is held at this point, communication can be sent with encrypting module to intelligent door lock
The instruction of lock.
By taking three kinds of " O ", " 8 ", " W " font dynamic gestures as an example, 10 users, each user is arranged to be surveyed in every wheel
Setting user is served as in examination successively, remaining 9 user serves as non-setting user, and experiment carries out 10 wheels altogether.Experiment test knot
Fruit is as follows:
" O " type gesture unlock rate
" 8 " type gesture unlock rate
" W " type gesture unlock rate
By data above as can be seen that the successful unlock probability of setting user is about 99.6%, non-setting user unlock
Success rate is about 0.03%, therefore the system has higher safety in the case where ensureing preferable user experience.
Claims (3)
1. a kind of micro- sense intelligent identity identification method of Behavior-based control feature, which is characterized in that including following two process parts:
First part, sample training part:
The dynamic gesture information several times of acquisition setting user, forms several training samples;
Using all training samples, discrimination model and decision threshold are calculated;
Wherein, training sample is formed to include the following steps:The HMM moulds corresponding to each sample are calculated using Baum Welch algorithms
Type, and its probability that occurs under HMM model, HMM refers to Hidden Markov Model;
Wherein, it calculates discrimination model and decision threshold includes the following steps:Calculate being averaged for the HMM corresponding to all training samples
Value is used as discrimination model, calculates the average value for the probability that all training samples occur under discrimination model as decision threshold;
Second part, identity, authentication section:
A dynamic gesture information for acquiring user, as sample to be identified;Sample to be identified is calculated under discrimination model to go out
Existing probability, if probability is more than decision threshold, then it is assumed that the user is setting user, conversely, then judging that the user is not
Set user.
2. according to the method described in claim 1, it is characterized in that, dynamic gesture information includes:Dynamic gesture is in X, Y, Z tri-
The acceleration and angular speed in direction.
3. a kind of micro- sense intelligent identity identification system of Behavior-based control feature, including identity information acquisition module and data processing mould
Block, which is characterized in that utilize identity information acquisition module and data processing module, realize method described in claim 1.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810392697.XA CN108595937A (en) | 2018-04-27 | 2018-04-27 | Micro-sensing intelligent identity authentication method based on behavior characteristics |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810392697.XA CN108595937A (en) | 2018-04-27 | 2018-04-27 | Micro-sensing intelligent identity authentication method based on behavior characteristics |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108595937A true CN108595937A (en) | 2018-09-28 |
Family
ID=63610147
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810392697.XA Pending CN108595937A (en) | 2018-04-27 | 2018-04-27 | Micro-sensing intelligent identity authentication method based on behavior characteristics |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108595937A (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103442114A (en) * | 2013-08-16 | 2013-12-11 | 中南大学 | Identity authentication method based on dynamic gesture |
CN103761466A (en) * | 2014-02-14 | 2014-04-30 | 上海云享科技有限公司 | Method and device for identity authentication |
CN104463250A (en) * | 2014-12-12 | 2015-03-25 | 广东工业大学 | Sign language recognition translation method based on Davinci technology |
CN105205475A (en) * | 2015-10-20 | 2015-12-30 | 北京工业大学 | Dynamic gesture recognition method |
CN106598274A (en) * | 2016-11-02 | 2017-04-26 | 雷飏 | Method for operating virtual key based on motion sensor, and corresponding user terminal |
-
2018
- 2018-04-27 CN CN201810392697.XA patent/CN108595937A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103442114A (en) * | 2013-08-16 | 2013-12-11 | 中南大学 | Identity authentication method based on dynamic gesture |
CN103761466A (en) * | 2014-02-14 | 2014-04-30 | 上海云享科技有限公司 | Method and device for identity authentication |
CN104463250A (en) * | 2014-12-12 | 2015-03-25 | 广东工业大学 | Sign language recognition translation method based on Davinci technology |
CN105205475A (en) * | 2015-10-20 | 2015-12-30 | 北京工业大学 | Dynamic gesture recognition method |
CN106598274A (en) * | 2016-11-02 | 2017-04-26 | 雷飏 | Method for operating virtual key based on motion sensor, and corresponding user terminal |
Non-Patent Citations (2)
Title |
---|
徐佳宾: "基于并行HMM的动态手势识别技术研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
马正华 等: "基于多传感器融合的动态手势识别研究分析", 《计算机工程与应用》 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110362290A (en) | A kind of sound control method and relevant apparatus | |
CN105429969B (en) | A kind of user ID authentication method and equipment | |
CN1972186B (en) | A mobile identity authentication system and its authentication method | |
JP3469031B2 (en) | Face image registration apparatus and method | |
CN110163611A (en) | A kind of personal identification method, device and relevant device | |
CN102890776B (en) | The method that expression figure explanation is transferred by facial expression | |
CN106899968A (en) | A kind of active noncontact identity identifying method based on WiFi channel condition informations | |
CN108875907B (en) | Fingerprint identification method and device based on deep learning | |
Qin et al. | A fuzzy authentication system based on neural network learning and extreme value statistics | |
CN108648310A (en) | A kind of face recognition door control system and its application process of double authentication | |
Fan et al. | Emgauth: An emg-based smartphone unlocking system using siamese network | |
CN110287918A (en) | Vivo identification method and Related product | |
El-Bendary et al. | Investigating of nodes and personal authentications utilizing multimodal biometrics for medical application of WBANs security | |
CN109840451A (en) | A kind of wearable ring of intelligence payment and its method of payment based on electrocardio identification | |
KR20180061819A (en) | Multiple biometric identity authentication apparatus and authentication system, multiple biometric identity authentication method therewith | |
Li et al. | CNN-based continuous authentication on smartphones with conditional Wasserstein generative adversarial network | |
CN112492090A (en) | Continuous identity authentication method fusing sliding track and dynamic characteristics on smart phone | |
Li et al. | Adaptive deep feature fusion for continuous authentication with data augmentation | |
CN113742669A (en) | User authentication method based on twin network | |
US20230231846A1 (en) | Biometric feature verification method and apparatus, electronic device, and storage medium | |
CN108595937A (en) | Micro-sensing intelligent identity authentication method based on behavior characteristics | |
CN114840834B (en) | Implicit identity authentication method based on gait characteristics | |
CN110413106A (en) | A kind of augmented reality input method and system based on voice and gesture | |
CN113065468B (en) | Gait authentication method based on user coordinate system and GRU network | |
CN112990261B (en) | Intelligent watch user identification method based on knocking rhythm |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180928 |