CN109302532B - Identity authentication method based on smart phone acceleration sensor - Google Patents
Identity authentication method based on smart phone acceleration sensor Download PDFInfo
- Publication number
- CN109302532B CN109302532B CN201811301498.XA CN201811301498A CN109302532B CN 109302532 B CN109302532 B CN 109302532B CN 201811301498 A CN201811301498 A CN 201811301498A CN 109302532 B CN109302532 B CN 109302532B
- Authority
- CN
- China
- Prior art keywords
- gait
- data
- vector
- acceleration sensor
- axis
- 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.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M1/00—Substation equipment, e.g. for use by subscribers
- H04M1/72—Mobile telephones; Cordless telephones, i.e. devices for establishing wireless links to base stations without route selection
- H04M1/724—User interfaces specially adapted for cordless or mobile telephones
- H04M1/72448—User interfaces specially adapted for cordless or mobile telephones with means for adapting the functionality of the device according to specific conditions
- H04M1/72463—User interfaces specially adapted for cordless or mobile telephones with means for adapting the functionality of the device according to specific conditions to restrict the functionality of the device
-
- 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
- G06F21/32—User authentication using biometric data, e.g. fingerprints, iris scans or voiceprints
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M1/00—Substation equipment, e.g. for use by subscribers
- H04M1/72—Mobile telephones; Cordless telephones, i.e. devices for establishing wireless links to base stations without route selection
- H04M1/724—User interfaces specially adapted for cordless or mobile telephones
- H04M1/72448—User interfaces specially adapted for cordless or mobile telephones with means for adapting the functionality of the device according to specific conditions
- H04M1/72454—User interfaces specially adapted for cordless or mobile telephones with means for adapting the functionality of the device according to specific conditions according to context-related or environment-related conditions
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M2250/00—Details of telephonic subscriber devices
- H04M2250/12—Details of telephonic subscriber devices including a sensor for measuring a physical value, e.g. temperature or motion
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Security & Cryptography (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Human Computer Interaction (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Hardware Design (AREA)
- Environmental & Geological Engineering (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention discloses an identity authentication method based on an acceleration sensor of a smart phone, and belongs to the field of identity authentication. Firstly, data generated by a mobile phone acceleration sensor is collected and preprocessed to obtain an Euclidean distance curve, gait cycle division is carried out according to the periodicity of the Euclidean distance curve which shows a rule along with the gait cycle, a gait vector corresponding to each gait cycle is calculated, and the Euclidean distance between each gait vector and the last gait vector is calculated by comparing each gait vector with the last gait vector. If the Euclidean distance is less than 2.5, recording the two gait vectors as continuous similar gait vectors, and adding one to the number; and when the number reaches 6, taking the last gait vector as a generated gait vector template, comparing the generated gait vector template with the current gait vector template, and performing identity authentication according to a comparison result. And if the number of the successfully authenticated gait vectors reaches the threshold set by the user, the identity authentication of the user is successful. The method has low computational complexity, excellent anti-attack capability and convenient use for users.
Description
Technical Field
The invention belongs to the field of identity authentication, relates to a mobile phone acceleration sensor, and particularly relates to an identity authentication method based on a smart mobile phone acceleration sensor.
Background
With the popularization and wide use of smart phones in life, people pay more attention to the security of the smart phones than ever before. However, the conventional identification schemes do not satisfy various needs such as digital passwords and picture passwords well, and since users often need to use mobile phones frequently and most of the use time is not long, frequent identification authentication is annoying, and meanwhile, the passwords are easy to be peeped or guessed to be cracked or to be attacked by social engineering, and related researches show that more than 40% of users do not use the identification schemes in their smart phones.
Compared with the traditional identity recognition scheme, the authentication method based on the biological characteristics of the user, such as fingerprint authentication, iris authentication has the advantages of high safety, no need of memory and the like. However, identity authentication based on such biological characteristics requires explicit operation of a user to perform authentication every time, which brings inconvenience to the user to some extent, and therefore, identity recognition research based on biological characteristics such as gait, usage habits, and key characteristics has been developed. Various sensors built in the smart phone, such as an acceleration sensor, a direction sensor and the like, provide possibility for the identity authentication scheme. If the gait is used for identity recognition, the identity recognition process can be completed under the condition of not disturbing the user by analyzing the sensor data, and the identity authentication process is shortened.
Gait features are features generated by a human during walking exercise, and the pioneer of scientific gait analysis is the research of Aristode: gait of the animal. As each person has a unique face, each person also has a distinctive gait. From the anatomical point of view, the physical basis of the gait uniqueness is that the gait uniqueness is determined by the difference of the physiological structure of each person, different leg bone lengths, different muscle strengths and the like.
The existing research for identity authentication by using acceleration sensor data is divided into two types according to an authentication method, wherein the first type is to divide continuous acceleration sensor data into walking periods, and use a Machine learning algorithm, such as Support Vector Machine (SVM), K-near-neighbors (KNN) and the like, to perform identity authentication after extracting characteristics. The second category is identity authentication by comparing continuous acceleration sensor data with a trained template using Dynamic Time Warping (DTW).
In the process of using biological characteristics to identify the identity, the obtained original data needs to be subjected to characteristic extraction, the extracted characteristics depict the biological characteristics of the user, and the result of the characteristic extraction directly influences the final result of the identity identification. However, the existing feature selection method is usually based on experience or experimental trial, so that the features cannot be explained in a biological sense, and the redundant features bring computational burden to real-time identity authentication. This makes it difficult for these methods to perform real-time identity authentication on the handset.
Disclosure of Invention
Aiming at the problems, the invention provides an identity authentication method based on an acceleration sensor of a smart phone, which implicitly performs identity authentication without disturbing a user.
The method mainly comprises the following steps:
step one, when a certain user walks with a mobile phone, collecting data generated by an acceleration sensor of the mobile phone in a certain time period t and preprocessing the data to obtain a Euclidean distance curve;
the method specifically comprises the following steps:
firstly, aiming at a certain time period t, a mobile phone acceleration sensor generates a plurality of groups of continuous triaxial data, and each group of data is respectively synthesized into a sub-combined acceleration;
for data a, the calculation formula of the sub-combined acceleration of the data a is as follows:
rAthe subcontracting accelerations for data set a; x is the number ofAA value on the X-axis of the acceleration sensor representing data A group; y isAA value on the Y-axis of the acceleration sensor representing data a group; z is a radical ofAA value on the Z axis of the acceleration sensor representing the data A group;
combining the sub-combined accelerations of several groups of data in the time period t to form a combined acceleration r continuous in timet；
rt＝{r1,r2,....rA,...}；
Then, for the continuous resultant acceleration rtPreprocessing to obtain a Euclidean distance curve;
the method specifically comprises the following steps: from a succession of resultant accelerations rtAnd intercepting all the combined acceleration data of the sub-groups in the 1 st second as a reference, continuously sliding the reference data to the right, and calculating the Euclidean distance between the reference data and the combined acceleration data at the corresponding position to obtain a Euclidean distance curve.
The initial corresponding position is 1, and the Euclidean distance is the distance between 1 and the reference data; in turn, the distance between 2 and datum +1, and so on.
Step two, according to the periodicity of the Euclidean distance curve showing rules along with the gait cycle, the gait cycle is divided, and the gait vector g corresponding to each gait cycle is calculatedv；
Each gait cycle corresponds to a gait vector representing walking behavior, and each gait cycle comprises a plurality of groups of continuous triaxial data;
gait vector gvThe calculation formula is as follows:
gv＝(x1,x2,x3,x4,x5,x6,x7)
x1～x7for all groups of acceleration sensor values x in the current gait cyclet，ytAnd ztA statistical value of (d);
the method specifically comprises the following steps: x is the number of1The upper quartile of the Z axis of the acceleration sensor in all three-axis data sets in the current gait cycle; x is the number of2The skewness of the X axis of the acceleration sensor in all three-axis data sets in the current gait cycle; x is the number of3The kurtosis of the X axis of the acceleration sensor in all three-axis data sets in the current gait cycle; x is the number of4The kurtosis of the Y axis of the acceleration sensor in all three-axis data sets in the current gait cycle; x is the number of5The combined acceleration is formed by combining all three-axis data in the current gait cycle; x is the number of6The upper quartile of the X axis of the acceleration sensor in all the three-axis data sets in the current gait cycle; x is the number of7The average value of the X axis of the acceleration sensor in all the three axis data sets in the current gait cycle is shown.
And step three, sequentially selecting each gait vector in the time period t, and comparing each gait vector with the previous gait vector to calculate the Euclidean distance between the two vectors.
Step four, judging whether the Euclidean distance is less than 2.5, if so, recording the two gait vectors as continuous similar gait vectors, and adding one to the number; otherwise, recording the number of the continuous similar gait vectors as 0.
The initial value of the number of the continuous similar gait vectors is 0.
And step five, when the number of the continuous similar gait vectors reaches 6, taking the last gait vector as a generated gait vector template.
And step six, comparing the newly generated gait vector with the current gait vector template, and performing identity authentication according to the comparison result.
When the initial value of the current gait vector template is that the number of continuous similar gait vectors reaches 6, the last gait vector is obtained.
The method specifically comprises the following steps: calculating the Euclidean distance between the newly generated gait vector and the obtained gait vector template, judging whether the distance is less than 2.5, if so, successfully authenticating the newly generated gait vector, and entering a seventh step; otherwise, the authentication fails.
Step seven, updating the gait vector template by using the newly generated gait vector, and returning to the step six;
updating the gait vector template formula as follows:
tn＝0.8*t0+0.2*v
t0updating the previous gait vector template; t is tnThe updated gait vector template; v is the gait vector used to update the gait vector template.
And step eight, judging whether the number of the gait vectors successfully authenticated reaches a threshold value set by the user, if so, successfully authenticating the identity of the user, and otherwise, failing to authenticate the identity of the user.
The invention has the advantages and positive effects that:
(1) an identity authentication method based on an acceleration sensor of a smart phone has small identity authentication calculated amount; in the process of identity authentication, no complex calculation is introduced, so that the identity authentication can be carried out in real time without obviously increasing the power consumption of the mobile phone.
(2) An identity authentication method based on an acceleration sensor of a smart phone, which is implicit identity authentication; the identity authentication can be completed when the user walks without needing the user to explicitly operate, and the user can use the mobile phone conveniently.
(3) An identity authentication method based on an acceleration sensor of a smart phone is high in safety; on the premise of ensuring the convenience of use of the user, the anti-attack system has excellent anti-attack capability, and even if an attacker is a trained professional, the attacker cannot effectively finish the attack by simulating the user.
Drawings
FIG. 1 is a flow chart of an identity authentication method based on an acceleration sensor of a smart phone according to the present invention;
FIG. 2 is a schematic diagram of a divided gait cycle of the invention;
FIG. 3 is a ranking of 7 features selected by the present invention to generate a gait vector;
FIG. 4 is a graph of the effect of the number of features in the gait vector on accuracy of the invention;
FIG. 5 is a schematic diagram showing the effect of different values of two coefficients on the result of updating the gait vector template according to the invention;
FIG. 6 is a comparison of the updating effect of the gait feature vector template according to the invention;
FIG. 7 is a schematic diagram of the walk detection sensor trigger of the present invention;
Detailed Description
The present invention will be described in further detail below with reference to the accompanying drawings.
In the prior art, the main flow of the identity authentication process based on user behavior is as follows: firstly, a participant is required to carry a mobile phone to complete certain specified actions, such as walking, running and the like, and a plurality of mobile phone sensor data are collected in the process; the collected sensor data is divided into a plurality of segments, a plurality of features are selected, for each segment, feature values of the corresponding features are calculated, and then the feature values are used for describing the characteristics of the participants. When identity authentication is carried out, received new data are also divided into a plurality of segments to calculate characteristic values, and whether the new data come from the same user or not is judged by comparing the difference condition of the new characteristic values and the old characteristic values, so that the identity authentication is completed.
In this process, the key step is how to characterize the user, i.e., which features to choose for calculation. Unfortunately, the existing methods are different in size when selecting features, and the common statistical indexes such as mean values and variances are often selected and often obtained according to experience of researchers or observation in daily life, and in the actual authentication process, the influence of each feature on the final authentication result is unknown.
This selection method has the following disadvantages. 1) Has no biological significance. The lack of a specific biological meaning means that it is not known why the feature is selected, and it is only known which actions distinguish different persons, why the features can distinguish the persons, if the specific biological meaning of the selected feature is known. 2) The computational burden. The amount of calculation is a factor that has to be considered in the identification process. The identity recognition system is implemented on a mobile phone and often requires real-time authentication. And the too complex characteristics will affect the calculation speed due to the local calculation capability of the mobile phone. In addition, some existing methods have redundant features, and unnecessary features cause unnecessary calculations. 3) The safety is poor. For identity authentication, the attack resistance of features must be considered, if some features are easily imitated, an attacker can increase the attack success rate of the attacker by imitating the actions of a user to complete the attack in the identity identification process. Care should be taken in choosing such features.
Based on the consideration, the invention provides an identity authentication method based on an acceleration sensor of a smart phone by analyzing the data of the acceleration sensor generated by a right-hand-held mobile phone when ten people walk, evaluating various characteristics from multiple aspects and selecting effective characteristics to generate a gait vector, and the identity authentication method mainly comprises the steps of generating the gait vector, generating a gait vector template, authenticating the gait vector, updating the gait vector template and the like.
As shown in fig. 1, the method mainly comprises the following steps:
step one, when a certain user walks with a mobile phone, collecting data generated by an acceleration sensor of the mobile phone in a certain time period t and preprocessing the data to obtain a Euclidean distance curve;
the user walks with the mobile phone, the acceleration sensor generates data in real time, then the real-time data is used for authentication, the speed generated by the acceleration sensor data is not fixed, dozens of groups can be generated in 1 second, each group of data comprises three numbers (x, y and z), the three numbers of each group are calculated, and then a sub-combined acceleration can be obtained, wherein the combined acceleration is a string of sub-combined accelerations and is data with continuous time. The method specifically comprises the following steps:
firstly, aiming at a certain time period t, a mobile phone acceleration sensor generates a plurality of groups of continuous triaxial data, and each group of data is respectively synthesized into a sub-combined acceleration;
for data a, the calculation formula of the sub-combined acceleration of the data a is as follows:
rAthe subcontracting accelerations for data set a; x is the number ofAA value on the X-axis of the acceleration sensor representing data A group; y isAA value on the Y-axis of the acceleration sensor representing data a group; z is a radical ofAA value on the Z axis of the acceleration sensor representing the data A group;
combining the sub-combined accelerations of several groups of data in the time period t to form a combined acceleration r continuous in timet；
rt＝{r1,r2,....rA,...}；
Then, for the continuous resultant acceleration rtPreprocessing to obtain a Euclidean distance curve;
the method specifically comprises the following steps: from a succession of resultant accelerations rtAnd intercepting all the combined acceleration data of the sub-groups with the time of 1S as a reference, continuously sliding the reference data to the right, and calculating the Euclidean distance between the reference data and the combined acceleration data of the corresponding position to obtain a Euclidean distance curve.
The initial corresponding position is 1, and the Euclidean distance is the distance between 1 and the reference data; in turn, the distance between 2 and datum +1, and so on. For example, a total of 1000 sub-combined acceleration data sets are obtained in a certain time period t, and 200 data sets are generated in the first second, the 200 data sets are taken as a reference, and then the reference data is compared 799 times with the reference data set of 1-200, 2-201, 3-202 … … 799 and 1000 times, and the process is called sliding.
Step two, the law is presented along with the gait cycle according to the Euclidean distance curvePeriodically dividing gait cycles, and calculating a gait vector g corresponding to each gait cyclev；
The invention adopts the existing cycle division method, and uses the composite value of three axes of the sensor to measure the walking gait of a person when dividing the walking cycle. The cycle extraction diagram is shown in fig. 2, and it can be seen that the composite values of three axes of the acceleration sensor show obvious periodicity when a person walks, and the time of one step is 0.5-0.8S when the person is in a stable walking state. In order to ensure the information in the gait vector template to be complete, a section of data with the length of 1S is selected as the template from the middle of the data. The template is moved forwards and backwards respectively, then the Euclidean distance between the template and the data is calculated, the obtained result is seen to be periodic, in order to carry out period division, the result is subjected to smooth filtering treatment, the filtering formula is shown as the following formula, M is the filtering length, and the value is selected to be 4 in the embodiment.
And calculating the wave crest and the wave trough of the resultant oscillogram to obtain the walking period, thereby calculating the gait vector.
In the gait vector generation stage, each gait cycle generates a corresponding gait vector representing walking behavior, and each gait cycle comprises a plurality of groups of continuous triaxial data;
in the embodiment, multiple characteristics are evaluated by analyzing acceleration sensor data generated by a right-handed mobile phone when ten people walk, and as shown in fig. 3, 7 characteristics are selected to generate gait vectors:
gv＝(x1,x2,x3,x4,x5,x6,x7)
x1～x7for all groups of acceleration sensor values x in the current gait cyclet，ytAnd ztA statistical value of (d); the effect of selecting these 7 features on accuracy is shown in fig. 4.
The method specifically comprises the following steps: x is the number of1The upper quartile of the Z axis of the acceleration sensor in all three-axis data sets in the current gait cycle; x is the number of2The skewness of the X axis of the acceleration sensor in all three-axis data sets in the current gait cycle; x is the number of3The kurtosis of the X axis of the acceleration sensor in all three-axis data sets in the current gait cycle; x is the number of4The kurtosis of the Y axis of the acceleration sensor in all three-axis data sets in the current gait cycle; x is the number of5The combined acceleration is formed by combining all three-axis data in the current gait cycle; x is the number of6The upper quartile of the X axis of the acceleration sensor in all the three-axis data sets in the current gait cycle; x is the number of7The average value of the X axis of the acceleration sensor in all the three axis data sets in the current gait cycle is shown.
And step three, sequentially selecting each gait vector in the time period t, and comparing each gait vector with the previous gait vector to calculate the Euclidean distance between the two vectors.
Step four, judging whether the Euclidean distance is less than 2.5, if so, recording the two gait vectors as continuous similar gait vectors, and adding one to the number; otherwise, recording the number of the continuous similar gait vectors as 0.
The initial value of the number of the continuous similar gait vectors is 0.
And step five, when the number of the continuous similar gait vectors reaches 6, taking the last gait vector as a generated gait vector template.
When identity authentication is performed, a gait feature vector needs to be obtained in advance as a template, and new data is compared with the template data to obtain an authentication result. In calculating the gait feature template, the user is required to walk to create the template. After the data are obtained, the continuous data are subjected to gait cycle division, and the corresponding gait feature vectors are calculated, so that a series of gait feature vectors can be obtained. When the value of the gait feature vector gradually stabilizes, a gait feature template is obtained: by calculating the distance between each gait vector and the last gait vector, if the distance between the continuous 6 gait vectors and the last gait vector is less than a given threshold value, the last gait vector is used as the gait vector template.
And step six, comparing the newly generated gait vector with the current gait vector template, and performing identity authentication according to the comparison result.
When the initial value of the current gait vector template is that the number of continuous similar gait vectors reaches 6, the last gait vector is obtained.
The method specifically comprises the following steps: calculating the Euclidean distance between the newly generated gait vector and the obtained gait vector template, judging whether the distance is less than 2.5, if so, successfully authenticating the newly generated gait vector, and entering a seventh step; otherwise, the authentication fails.
In the process of generating the gait feature template, a Threshold variable Threshold is introduced to measure whether two gait feature vectors are similar enough. When the Euclidean distance of two gait feature vectors is smaller than the threshold value, the two gait feature vectors are considered to possibly come from the same user.
The invention determines the threshold value by considering the change degree of the gait feature vector of the same user. The data is stable in the middle part of the walking acceleration sensor data, so 100 gait feature vectors are extracted from all gait feature vectors of each person to calculate the average distance between every two gait feature vectors. To ensure the success of identification, the threshold value should be slightly larger than the average distance of the gait feature vectors of the users and smaller than the distance of the gait feature vectors between different users. The Threshold is set accordingly to 2.5.
Step seven, updating the gait vector template by using the newly generated gait vector, and returning to the step six;
the gait tends to be stable when the person walks, however, from a long-term perspective, these gait characteristics may change gradually over time. Therefore, in order to better adapt to such changes and thus enable long-term identification, an update mechanism is needed to ensure that the algorithm will adapt over time. For this purpose, an update algorithm is proposed for updating the gait feature vector. The core idea is that if one-time identity authentication action is passed, namely a user of the mobile phone generates a new gait feature vector, the new gait feature vector represents the latest gait condition and trend of the user to a certain extent, so that the original gait feature vector template can be updated by using the latest gait feature vector to enable the gait feature vector template to adapt to the latest state.
Updating the gait vector template formula as follows:
tn＝0.8*t0+0.2*v
t0updating the previous gait vector template; t is tnThe updated gait vector template; v is the gait vector used to update the gait vector template. The reason for selecting 0.8 and 0.2 is shown in FIG. 5.
And step eight, judging whether the number of the gait vectors successfully authenticated reaches a threshold value set by the user, if so, successfully authenticating the identity of the user, and otherwise, failing to authenticate the identity of the user.
In order to show the necessity of updating, by showing the results of two times of identity authentication, as shown in fig. 6, the abscissa represents data, the ordinate represents the distance between the gait feature vector to be tested and the gait feature vector template, the solid line is updated by using an updating algorithm, and the dotted line is not updated by using an updating algorithm. It can be seen from the figure that after the update algorithm is used, due to the update capability of the gait feature vector template, when new data are continuously received, if the update condition is satisfied, the gait feature vector template can be updated autonomously, so that the new data are more satisfied, and the identity can be better identified.
The power consumption condition of the identity authentication of the application test is realized on the android mobile phone. Since it is desirable to perform authentication without disturbing the user, the program needs to be run in the background all the time, but the continuous operation of the monitoring sensor increases the power consumption, and therefore a walking detection sensor is used as a trigger function. As shown in fig. 7, the walking detection sensor is a sensor provided by Android and triggered each time the user walks, and consumes less power compared to a common sensor. When the walking detection sensor is triggered, the data of the acceleration sensor starts to be collected, and if the walking detection sensor is not triggered any more after a period of time, the program returns to the sleep state, continues to be triggered, and the like.
Claims (2)
1. An identity authentication method based on an acceleration sensor of a smart phone is characterized by mainly comprising the following steps:
step one, when a certain user walks with a mobile phone, collecting data generated by an acceleration sensor of the mobile phone in a certain time period t and preprocessing the data to obtain a Euclidean distance curve;
the method specifically comprises the following steps:
firstly, aiming at a certain time period t, a mobile phone acceleration sensor generates a plurality of groups of continuous triaxial data, and each group of data is respectively synthesized into a sub-combined acceleration;
for data a, the calculation formula of the sub-combined acceleration of the data a is as follows:
rAthe subcontracting accelerations for data set a; x is the number ofAA value on the X-axis of the acceleration sensor representing data A group; y isAA value on the Y-axis of the acceleration sensor representing data a group; z is a radical ofAA value on the Z axis of the acceleration sensor representing the data A group;
combining the sub-combined accelerations of several groups of data in the time period t to form a combined acceleration r continuous in timet；
rt＝{r1,r2,....rA,...}；
Then, for the continuous resultant acceleration rtPreprocessing to obtain a Euclidean distance curve;
the method for obtaining the Euclidean distance curve specifically comprises the following steps: from a succession of resultant accelerations rtIntercepting all the combined acceleration data of the sub-groups in the 1 st second as a reference, continuously sliding the reference data to the right, and calculating the Euclidean distance between the reference data and the combined acceleration data at the corresponding position to obtain an Euclidean distance curve;
the initial corresponding position is 1, and the Euclidean distance is the distance between 1 and the reference data; the distances between 2 and the datum data +1 are sequentially obtained, and the like;
step two, according to the periodicity of the Euclidean distance curve showing rules along with the gait cycle, the gait cycle is divided, and the gait vector g corresponding to each gait cycle is calculatedv；
Each gait cycle corresponds to a gait vector representing walking behavior, and each gait cycle comprises a plurality of groups of continuous triaxial data;
gait vector gvThe calculation formula is as follows:
gv＝(x1,x2,x3,x4,x5,x6,x7)
x1～x7for all groups of acceleration sensor values x in the current gait cyclet，ytAnd ztA statistical value of (d);
step three, sequentially selecting each gait vector in the time period t, and comparing each gait vector with the previous gait vector to calculate the Euclidean distance between the two vectors;
step four, judging whether the Euclidean distance is less than 2.5, if so, recording the two gait vectors as continuous similar gait vectors, and adding one to the number; otherwise, recording the number of the continuous similar gait vectors as 0;
step five, when the number of the continuous similar gait vectors reaches 6, taking the last gait vector as a generated gait vector template;
step six, comparing the newly generated gait vector with the current gait vector template, and performing identity authentication according to the comparison result;
the method specifically comprises the following steps: calculating the Euclidean distance between the newly generated gait vector and the obtained gait vector template, and judging whether the distance is less than 2.5, if so, successfully authenticating the newly generated gait vector, otherwise, failing to authenticate;
step seven, updating the gait vector template by using the newly generated gait vector, and returning to the step six;
updating the gait vector template formula as follows:
tn＝0.8*t0+0.2*v
t0updating the previous gait vector template; t is tnThe updated gait vector template; v is a gait vector used for updating the gait vector template;
and step eight, judging whether the number of the gait vectors successfully authenticated reaches a threshold value set by the user, if so, successfully authenticating the identity of the user, and otherwise, failing to authenticate the identity of the user.
2. The identity authentication method based on the smart phone acceleration sensor as claimed in claim 1, wherein the second step is specifically: x is the number of1The upper quartile of the Z axis of the acceleration sensor in all three-axis data sets in the current gait cycle; x is the number of2The skewness of the X axis of the acceleration sensor in all three-axis data sets in the current gait cycle; x is the number of3The kurtosis of the X axis of the acceleration sensor in all three-axis data sets in the current gait cycle; x is the number of4The kurtosis of the Y axis of the acceleration sensor in all three-axis data sets in the current gait cycle; x is the number of5The combined acceleration is formed by combining all three-axis data in the current gait cycle; x is the number of6The upper quartile of the X axis of the acceleration sensor in all the three-axis data sets in the current gait cycle; x is the number of7The average value of the X axis of the acceleration sensor in all the three axis data sets in the current gait cycle is shown.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811301498.XA CN109302532B (en) | 2018-11-02 | 2018-11-02 | Identity authentication method based on smart phone acceleration sensor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811301498.XA CN109302532B (en) | 2018-11-02 | 2018-11-02 | Identity authentication method based on smart phone acceleration sensor |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109302532A CN109302532A (en) | 2019-02-01 |
| CN109302532B true CN109302532B (en) | 2020-02-07 |
Family
ID=65145758
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811301498.XA Active CN109302532B (en) | 2018-11-02 | 2018-11-02 | Identity authentication method based on smart phone acceleration sensor |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109302532B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110968857B (en) * | 2019-12-03 | 2022-04-08 | 南京航空航天大学 | Smart watch identity authentication method based on arm lifting action |
| CN112966248B (en) * | 2021-03-23 | 2022-10-21 | 西安电子科技大学 | Continuous identity authentication method of mobile equipment in uncontrolled walking scene |
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 |
| CN106850955A (en) * | 2016-12-20 | 2017-06-13 | 陕西尚品信息科技有限公司 | A kind of mobile phone identity verification method based on Gait Recognition |
| CN107103297A (en) * | 2017-04-20 | 2017-08-29 | 武汉理工大学 | Gait identification method and system based on mobile phone acceleration sensor |
| JP6193539B2 (en) * | 2012-01-17 | 2017-09-06 | 日本信号株式会社 | Human body communication card holder and RFID tag authentication system |
| CN107273726A (en) * | 2017-06-02 | 2017-10-20 | 中国人民解放军信息工程大学 | Equipment owner's identity real-time identification method and its device based on acceleration cycle variation law |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107831907A (en) * | 2017-12-07 | 2018-03-23 | 深圳先进技术研究院 | Identity identifying method and device based on Gait Recognition |
-
2018
- 2018-11-02 CN CN201811301498.XA patent/CN109302532B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6193539B2 (en) * | 2012-01-17 | 2017-09-06 | 日本信号株式会社 | Human body communication card holder and RFID tag authentication system |
| CN103442114A (en) * | 2013-08-16 | 2013-12-11 | 中南大学 | Identity authentication method based on dynamic gesture |
| CN106850955A (en) * | 2016-12-20 | 2017-06-13 | 陕西尚品信息科技有限公司 | A kind of mobile phone identity verification method based on Gait Recognition |
| CN107103297A (en) * | 2017-04-20 | 2017-08-29 | 武汉理工大学 | Gait identification method and system based on mobile phone acceleration sensor |
| CN107273726A (en) * | 2017-06-02 | 2017-10-20 | 中国人民解放军信息工程大学 | Equipment owner's identity real-time identification method and its device based on acceleration cycle variation law |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109302532A (en) | 2019-02-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Sun et al. | Accelerometer-based speed-adaptive gait authentication method for wearable IoT devices | |
| CN108776788B (en) | Brain wave-based identification method | |
| Muaaz et al. | Smartphone-based gait recognition: From authentication to imitation | |
| Wu et al. | Liveness is not enough: Enhancing fingerprint authentication with behavioral biometrics to defeat puppet attacks | |
| TWI342548B (en) | Biometrics authentication apparatus | |
| Qin et al. | A fuzzy authentication system based on neural network learning and extreme value statistics | |
| WO2018152711A1 (en) | Electrocardiographic authentication-based door control system and authentication method therefor | |
| CN109302532B (en) | Identity authentication method based on smart phone acceleration sensor | |
| CN109934114A (en) | A kind of finger vena template generation and more new algorithm and system | |
| JP4951415B2 (en) | Biometric authentication system | |
| Chavan et al. | Fingerprint authentication using Gabor filter based matching algorithm | |
| KR20180082948A (en) | Method and apparatus for authenticating a user using an electrocardiogram signal | |
| CN114840834B (en) | Implicit identity authentication method based on gait characteristics | |
| Karnan et al. | Bio password—keystroke dynamic approach to secure mobile devices | |
| Shen et al. | Toward hand-dominated activity recognition systems with wristband-interaction behavior analysis | |
| Abuqadumah et al. | Personal authentication application using deep learning neural network | |
| Holien | Gait recognition under non-standard circumstances | |
| CN113241081B (en) | Far-field speaker authentication method and system based on gradient inversion layer | |
| Stragapede et al. | IJCB 2022 mobile behavioral biometrics competition (MobileB2C) | |
| Li et al. | Intelligent walk authentication: implicit authentication when you walk with smartphone | |
| CN110298159B (en) | Smart phone dynamic gesture identity authentication method | |
| Fan et al. | EmgAuth: Unlocking smartphones with EMG signals | |
| Chen et al. | Modeling interactive sensor-behavior with smartphones for implicit and active user authentication | |
| KR100602526B1 (en) | Verification of Iris Patterns Based on Variable Threshold | |
| Jiang et al. | MAUTH: Continuous user authentication based on subtle intrinsic muscular tremors |
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 |