CN107215307A - Driver identity recognition methods and system based on vehicle sensors correction data - Google Patents
Driver identity recognition methods and system based on vehicle sensors correction data Download PDFInfo
- Publication number
- CN107215307A CN107215307A CN201710374491.XA CN201710374491A CN107215307A CN 107215307 A CN107215307 A CN 107215307A CN 201710374491 A CN201710374491 A CN 201710374491A CN 107215307 A CN107215307 A CN 107215307A
- Authority
- CN
- China
- Prior art keywords
- data
- driver
- vehicle
- sensing data
- identity
- 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
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60R—VEHICLES, VEHICLE FITTINGS, OR VEHICLE PARTS, NOT OTHERWISE PROVIDED FOR
- B60R25/00—Fittings or systems for preventing or indicating unauthorised use or theft of vehicles
- B60R25/20—Means to switch the anti-theft system on or off
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/08—Insurance
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Economics (AREA)
- Mechanical Engineering (AREA)
- Development Economics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Technology Law (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Traffic Control Systems (AREA)
Abstract
Driver identity recognition methods and system based on vehicle sensors correction data, this method is mapped to vehicle movement coordinate system by sensor 3-axis acceleration coordinate system by obtaining vehicle sensor data and being corrected, by sensing data and eliminates the influence to aftermentioned identification process that sensor alignment error is brought；The extraction of statistical nature is carried out to the sensing data after correction again, the characteristic vector for obtaining characterizing personal driving habit is input in the good neutral net of a precondition, you can output identification result；Wherein, the training of neutral net is that the identity data that will be previously entered is exported as training set, is trained the statistical nature of sensing data as training set input.This method can in real time, relatively accurately recognize the identity of current human pilot, and identification process will not be impacted to human pilot.
Description
Technical field
The present invention relates to driving behavior analysis field, and in particular to driver's body based on vehicle sensors correction data
Part recognition methods and system.
Background technology
Know currently used for driver identity mainly has three classes otherwise：External equipment identification, living things feature recognition and people
Face is recognized.
External equipment is recognized, mainly recognizes identity by detecting the external equipment of driver's carrying.Outside herein
Equipment such as Contact Type Ic Card, non-contact IC card or USB moveable magnetic discs, but these external equipments be easily lost and
It is forged, it is difficult to reach the purpose for recognizing driver in real time, exactly, security is low.
Living things feature recognition, mainly recognizes driver identity by detecting human body biological characteristics such as fingerprint, iris etc..
This mode relative to external equipment know otherwise, though in the absence of can lose the problem of and security and real-time are also high,
Implement unrealistic, mainly expensive and identification equipment installation operation is difficult, while to driver when being identified
It is required that harsher, it is necessary to be kept for certain time with specific posture and suitable position, it is difficult to accomplish to gather driver's in real time
Biological characteristic；On the other hand, because most of sensors need and human contact and are pasted into electrode, with very strong interference,
Driver behavior may be influenceed, traffic safety hidden danger is caused, this kind of mode is simultaneously impracticable.
Recognition of face, i.e., by scanning face and recognizing driver identity after carrying out a series of image real time transfer, in
This method has been used in state Publication No. CN202130310U patent document.Such a mode same price is high, Er Qiezhu
To be applied to road monitoring equipment and gather image, face identification system is installed in the car certain difficulty；On the other hand, driver
Dress up, the colour of skin, ambient light light and shade and vehicle rock has considerable influence to the image collected, can directly influence image
Result, so as to influence the degree of accuracy of identification；Furthermore, camera device needs alignment driver's face, and the efficiency that operates is low,
Also it is and impracticable.
The content of the invention
In order to overcome the defect present in above-mentioned existing identification method, the present invention proposes a kind of based on vehicle sensors school
The driver identity recognition methods of correction data, can reach more accurate, concealed identification, and identification with relatively low cost
Process does not interfere with driving.
The present invention is as follows for the technical scheme proposed up to above-mentioned purpose：
A kind of driver identity recognition methods based on vehicle sensors correction data, comprises the following steps：
S1, the identity data for receiving driver's input and preservation；
The sensing data when driver that S2, acquisition vehicle sensors are collected drives, the sensing data is at least wrapped
Include 3-axis acceleration data；
S3, when choosing from the sensing data according to acceleration rate threshold set in advance vehicle horizontal stationary and vehicle
3-axis acceleration data when level ground moves along a straight line, for calculating vehicle movement coordinate system and sensor 3-axis acceleration
Spin matrix between coordinate system；
S4, by spin matrix the sensing data obtained in step S2 is corrected, by the sensing data
The vehicle movement coordinate system is mapped to by the sensor 3-axis acceleration coordinate system；
S5, to by step S4 correction after sensing data extract statistical nature, obtain characteristic vector；
S6, the characteristic vector for obtaining step S5 are input in a neural network model as training set and are trained, and treat
The identification model that network parameter is trained when restraining；
S7, the driver unknown to identity perform step S2 to S5 and obtain characteristic vector to be identified；
S8, the characteristic vector to be identified is input in the identification model, exports identification result.
The above method that the present invention is provided, is realized, sensor device price first based on vehicle sensors gathered data
It is cheap, secondly install concealed, by carrying out identity after performing the data processing of above steps to the sensing data collected
Identification, whole identification process is concealed without influenceing driving procedure；On the other hand, it can be recognized in real time, in the absence of existing
Have equipment in technology it is stolen/risk of loss/forgery.
The present invention separately also provides a kind of driver identity identifying system based on vehicle sensors correction data, including vehicle
Sensor, processor, memory and human-computer interaction interface；Be stored with a computer program on the memory；It is described man-machine
Interactive interface is used to input information to operate the system for human pilot；It is real during computer program described in the computing device
The step of existing preceding method.
The above-mentioned driver identity identifying system that the present invention is provided, is realized based on vehicle sensors, by adopting in real time
The sensing data of collection carries out carrying out identification after volume of data processing, and system is easily achieved and installed and overall cost
Low, system does not influence to drive in identification process, and can in real time, relatively accurately realize that driver identity is recognized.
Brief description of the drawings
Fig. 1 is the graph of a relation between sensor 3-axis acceleration coordinate system and vehicle movement coordinate system；
Fig. 2 is data window schematic diagram；
Fig. 3 is the schematic diagram of neural network training process；
Fig. 4 is the schematic diagram of identification procedure.
Embodiment
The invention will be further described below in conjunction with the accompanying drawings and preferred embodiment.
The embodiment of the present invention provides a kind of driver identity identification based on vehicle sensors correction data
Method, this method includes the identification procedure in two big flows, the training process of identification model and actual use.Its
The principle of middle model training process be by the sensing data in driving procedure counted and obtain characterize each one drive habit
Used characteristic vector, then based on machine learning, the identity data inputted with driver and the spy for characterizing driver's driving habit
The output and input of neutral net when levying vector respectively as training, carry out the training of neutral net until network parameter satisfaction is received
That holds back condition and trained can be used for entering the unknown driver of identity (driver for not living through model training process)
The neutral net (can be described as identification model) of row identification.Identification procedure in actual use is that collection is currently driven
The person of sailing drive the sensing data produced by vehicle and when progress is with model training same data processing and obtain characteristic vector,
Characteristic vector is inputted in identification model again, identification result is exported.
The driver identity recognition methods based on vehicle sensors correction data proposed based on above-mentioned principle, mainly
Including model training process as shown in Figure 3 and identification process as shown in Figure 4, for a certain vehicle, with highest weight
The personnel (such as car owner) of limit can launch into model training process, with the prior typing of the people for being provided as vehicle operator certainly
Oneself identity data and the training for carrying out identification model, training completion can close model training function, then vehicle is in and can carried out
The state of identification.
As shown in figure 3, model training process comprises the following steps 1 to 6：
Step 1, the identity data for receiving driver's input and preservation；
Sensing data and caching, the sensor number when driver that step 2, acquisition vehicle sensors are collected drives
According at least including x-axis, y-axis, z-axis 3-axis acceleration data；Data can first be pre-processed before processing described later is carried out with
Remove the mess code being likely to occur in ablation process；
Step 3, when choosing from the sensing data according to acceleration rate threshold set in advance vehicle horizontal stationary and
3-axis acceleration data of the vehicle when level ground moves along a straight line, add for calculating vehicle movement coordinate system with the axle of sensor three
Spin matrix between velocity coordinate system；
Step 4, by spin matrix the sensing data obtained in step 2 is corrected, by the sensor number
The vehicle movement coordinate system is mapped to according to by the sensor 3-axis acceleration coordinate system；
Step 5, to by step 4 correction after sensing data extract statistical nature obtain characteristic vector；
Step 6, the characteristic vector for obtaining step 5 are as input, while coming the identity data as output to one
Neural network model is trained, until parameter obtains the nerve that can be used for carrying out driver identity identification when meeting the condition of convergence
Network (or making identification model).Wherein, the training process of neural network model is minimized using back-propagation algorithm
The cross entropy of model, carrys out training pattern parameter, until meeting the condition of convergence with batch gradient descent method.
For each driver for being allowed to drive a certain vehicle, above-mentioned step can be passed through in the case where opening model training
Rapid 1 to 6 participates in the training of neural network model, to cause backward under identification state, can pass through body of the invention during driving
Part identification process identifies the identity of oneself.
Model training is completed, then the personnel (such as car owner) with highest authority can close model training function, is closed
I.e. automatic afterwards in that can carry out the state of driver identity identification, identification procedure refers to Fig. 4, comprises the following steps 7 and step
Rapid 8：
Obtaining current identity in real time, continuously after step 7, vehicle launch, unknown (for system, identity is unknown
) sensing data that human pilot drives, and carry out the processing procedure of abovementioned steps 2 to 5, obtain corresponding multiple features to
Amount；
Step 8, the characteristic vector for obtaining process step 7 are input in the above-mentioned neutral net trained, then neural
The exportable identification result of network.Specifically, characteristic vector is inputted by the input layer of neutral net, by hidden layer, finally
Output result is obtained by the Softmax functions of output layer, the output of Softmax functions is the vector that length is specific dimension, to
Result of calculation in amount per dimension represents probability of the input data from the driver corresponding to the dimension, takes probability highest
Dimension be used as identification result output.For a certain vehicle, if the n drivers being allowed to, then it can pass through
Foregoing model training process generation output dimension is the neutral net that n+1 is tieed up, i.e. the identification model of the vehicle has n not
With driver identity data, the output dimension of Softmax functions can be set to be tieed up for n+1, preceding n dimensions correspond to existing n not
Same driver.When in actual identification process, input data (characteristic vector formed after sensing data processing) is not belonging to this n
During individual driver, the probable value highest of the (n+1)th dimension in the output of Softmax functions, output result is n+1 (generally in coding
We are to represent the result with [0,0 ..., 0,1], before n be 0, which dimension is 1 and represents which current driver's belong to
Individual dimension, the value of other dimensions is 0, can be converted into corresponding driver when output is to user, such as result is
When [0,0 ..., 0,1], we, which export, gives user " unknown driver ", and this is the select permeability in interactive mode, does not constitute pair
The limitation of the present invention), represent that the input data is not belonging to known n driver, then simultaneity factor can be for unknown at present
The situation that personnel drive vehicle takes certain measure (such as the aftermentioned control vehicle stall referred to).For example：Assuming that by
Correspond to the data of 3 different identity drivers in the vehicle A of above-mentioned model training identification model, compiled respectively
Code is 1000,0100,0010, when vehicle A is by being not belonging to this 3 driver stored driving, the output of identification model
Recognition result be 0001, it is unknown identity to represent current identity.If the identity None- identified of current driver's, system can be directed to
The situation of None- identified performs further processing, such as by sending an instruction to make vehicle stall to forbid continuing moving ahead,
It can also be alarmed simultaneously by a radio receiving transmitting module to car owner.
It is because vehicle sensors are difficult to avoid that in installation process and can deposited that sensing data, which is corrected, in step 4
In error, with reference to Fig. 1, cause vehicle movement coordinate system XYZ and sensor 3-axis acceleration coordinate system X ' Y ' Z ' misaligned, exist
Relation as shown in Figure 1, can be corrected by calculating spin matrix between the two.For each car, spin matrix is only
It need to calculate once, calculate after spin matrix, matrix multiplication operation is made with the sensing data and spin matrix that collect, you can
Sensing data is mapped in vehicle movement coordinate system, the shadow to later data processing that sensor alignment error is brought is eliminated
Ring.
The calculating of the spin matrix can use such a way：
According to acceleration rate threshold set in advance, when the acceleration long period being less than the threshold value, we take the data to be
The acceleration of horizontal stationary；There is obvious add before and after acceleration is less than the threshold value, and the shorter time period in the short period
During velocity variations, the data that we set now are in the acceleration information under horizontal rectilinear motion state as vehicle.Such as Fig. 1 institutes
Show, under vehicle level inactive state, it is respectively a to obtain 3-axis acceleration sensor datax0、ay0、az0.Due under inactive state
Automobile is only acted on by terrestrial gravitation, and gravity acceleration g is met
Note OZ ' and OZ angle is that α, OY ' and OZ angle are that β, OX ' and OZ angle is γ, then has
In view of automobile, horizontal direction acceleration is zero under static state, therefore is had
When automobile moves along a straight line in level ground, automobile only by it is preceding to tractive force and Action of Gravity Field, therefore to acceleration pass
The acceleration information that sensor is collected carries out decomposition vertically and horizontally, is apparent from vertical direction vector and adds for gravity
Speed.Horizontal direction vector points to OY directions, now obtains 3-axis acceleration sensor data difference axt、ayt、azt, have
According to numerical solution spin matrix R in above-mentioned two situations, wherein R expression formula is as follows.
The transformational relation of the two is (ax, ay, az)T=RT·(ax′, ay′, az′)T.Wherein (ax′, ay′, az′)T(ax, ay, az)T
The respectively data on sensor and the data after correction.
In a preferred embodiment, the extraction that the sensing data after correction carries out statistical nature is specifically included following
Step one and two：
Step 1: collect many groups of sensing datas (each group include x, y, z 3-axis acceleration data ACC_X,
ACC_Y, ACC_Z), that sorts according to time order and function and (receive in real time and data storage, therefore without deliberately be ranked up) is more
Group sensing data is divided into multiple continuous data windows, and dividing mode is, for example,：The head and the tail of adjacent two data window are extremely
One group of sensing data is overlapped less, and each data window includes the sensing data of identical group of number.Such as, with reference to Fig. 2, each
Data window includes L=8 groups data, and (L value is merely illustrative herein, and L is integer and L >=2), and adjacent two data window is (for example
Window1 and Window2 in figure) between overlap have 4 groups of data, i.e., rear 4 groups of data of previous window are adjacent latter
Preceding 4 groups of data of window, the rest may be inferred, and currently available all the sensors data are carried out into such division, obtains many
Data window.Why to allow the head and the tail of adjacent window apertures to have data coincidence, be because previous window tail data with it is latter
Certain relation is there may be between the header data of individual window and with certain information, if misaligned, information can be caused to lose.
Step 2: carrying out after above-mentioned window division, it is assumed that currently available 1000 data windows, respectively to each window
Perform following operate：
Every column data is made the difference point, for the example shown in Fig. 2, obtain the new window containing L-1=7 group differential datas
Mouthful, each column to the new window and the data window seeks 9 kinds of statistical natures respectively, and each column obtains 9 statistics, described nine
Kind of statistical nature include average value, minimum value, 25% quantile, median, 75% quantile, maximum, standard deviation, kurtosis and
The degree of bias.In general, these above-mentioned statistical nature geometry get up relatively accurately characterize the driving habit of a people.
So as to which each data window obtains 54 statistics so that each data window obtains one containing 54 features
Characteristic vector so that 1000 current data windows can obtain 1000 such characteristic vectors.Assuming that be currently at into
During row model training, then this 1000 characteristic vectors can be used as training set and be input in neutral net to be trained.
Assuming that be in identification process, it is necessary to obtain how many characteristic vectors just carry out identification can be in advance by programming
Setting.
The method of the present invention, which can form a set of identification system, is used for vehicle, and vehicle anti-theft and personalized customization can be achieved
The car insurance amount of money, what vehicle anti-theft was for example previously mentioned, when finding identity None- identified, system can control vehicle stall simultaneously
Send alarm.The application of car insurance we same vehicle can each be driven with the different driver of time recording by intelligent terminal
Time, according to the driving age of each driver, record, the information such as age is for vehicle personalization customization insured amount in violation of rules and regulations,
For long-time by the driving age it is longer, record driver-operated vehicle less, in the prime of life in violation of rules and regulations relatively low guarantor can be set
The dangerous amount of money, vice versa.The system should include a radio receiving transmitting module, for the intelligent terminal with car owner or primary flight people
Between carry out radio communication, to be manipulated by intelligent terminal to the system and by the identification result of the system
Feed back to the intelligent terminal.Manipulation is carried out to system includes close/open system, driving time statistics, on/off system
Model training function etc..
Identification system based on the inventive method, including vehicle sensors, processor, memory and man-machine interaction
Interface；Be stored with a computer program on the memory；The human-computer interaction interface be used for for human pilot input information with
Operate the system；The foregoing driver identity identification of the present invention can be achieved described in the computing device during computer program
The step of method.
Above content is to combine specific preferred embodiment further description made for the present invention, it is impossible to assert
The specific implementation of the present invention is confined to these explanations.For those skilled in the art, do not taking off
On the premise of from present inventive concept, some equivalent substitutes or obvious modification can also be made, and performance or purposes are identical, all should
When being considered as belonging to protection scope of the present invention.
Claims (8)
1. a kind of driver identity recognition methods based on vehicle sensors correction data, comprises the following steps：
S1, the identity data for receiving driver's input and preservation；
The sensing data when driver that S2, acquisition vehicle sensors are collected drives, the sensing data at least includes three
Axle acceleration data；
S3, when choosing from the sensing data according to acceleration rate threshold set in advance vehicle horizontal stationary and vehicle is in water
3-axis acceleration data when plane earth moves along a straight line, for calculating vehicle movement coordinate system and sensor 3-axis acceleration coordinate
Spin matrix between system；
S4, by spin matrix the sensing data obtained in step S2 is corrected, by the sensing data by institute
State sensor 3-axis acceleration coordinate system and be mapped to the vehicle movement coordinate system；
S5, to by step S4 correction after sensing data extract statistical nature, obtain characteristic vector；
S6, the characteristic vector for obtaining step S5 are input in a neural network model as training set and are trained, and treat network
The identification model that parameter is trained when restraining；
S7, the driver unknown to identity perform step S2 to S5 and obtain characteristic vector to be identified；
S8, the characteristic vector to be identified is input in the identification model, exports identification result.
2. driver identity recognition methods as claimed in claim 1, it is characterised in that：Also include the biography got to step S2
Sensor data are pre-processed, to remove the mess code occurred in data writing process.
3. driver identity recognition methods as claimed in claim 1, it is characterised in that：The sensing data that step S2 is obtained
Matrix multiplication operation is carried out with the spin matrix, to perform the correction.
4. driver identity recognition methods as claimed in claim 1, it is characterised in that：Neural network model is entered in step S6
The training of row network parameter is specifically included：The cross entropy of neural network model is minimized using back-propagation algorithm, and is used
Batch gradient descent method carrys out training pattern parameter, until parameter meets the condition of convergence.
5. driver identity recognition methods as claimed in claim 1, it is characterised in that：Step S5 specifically includes following sub-step
Suddenly：
S51, the multigroup sensing data sorted according to time order and function is divided into multiple continuous data windows, and it is two neighboring
The head and the tail of data window at least overlap one group of sensing data, and each data window includes L group sensing datas, each group of biography
Sensor packet contains three column datas, and the acceleration information of x-axis, y-axis, z-axis is represented respectively；Wherein, L is whole more than or equal to 2
Number；
S52, following operate is performed to each data window respectively：Every column data is made the difference point, obtained containing L-1 group differential datas
New window, each column to the new window and the data window seeks 9 kinds of statistical natures respectively, and each column obtains 9 statistics,
9 kinds of statistical natures include average value, minimum value, 25% quantile, median, 75% quantile, maximum, standard deviation,
Kurtosis and the degree of bias；
By step S52, each data window obtains 54 statistics so that each data window obtains one and contained
There is the characteristic vector of 54 features.
6. driver identity recognition methods as claimed in claim 1, it is characterised in that：Driver is received by an interactive interface
The identity data for manually selecting and inputting.
7. a kind of driver identity identifying system based on vehicle sensors correction data, including vehicle sensors, processor, deposit
Reservoir and human-computer interaction interface；Be stored with a computer program on the memory；The human-computer interaction interface is used for for driving
Personnel's input information is sailed to operate the system；Such as claim 1 to 6 is realized described in the computing device during computer program
The step of any one methods described.
8. driver identity identifying system as claimed in claim 7, it is characterised in that：Also include a radio receiving transmitting module, use
Radio communication is carried out between intelligent terminal, so that car owner is manipulated by intelligent terminal to the system and will be described
The identification result of system feeds back to the intelligent terminal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710374491.XA CN107215307A (en) | 2017-05-24 | 2017-05-24 | Driver identity recognition methods and system based on vehicle sensors correction data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710374491.XA CN107215307A (en) | 2017-05-24 | 2017-05-24 | Driver identity recognition methods and system based on vehicle sensors correction data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107215307A true CN107215307A (en) | 2017-09-29 |
Family
ID=59944739
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710374491.XA Pending CN107215307A (en) | 2017-05-24 | 2017-05-24 | Driver identity recognition methods and system based on vehicle sensors correction data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107215307A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108280482A (en) * | 2018-01-30 | 2018-07-13 | 广州小鹏汽车科技有限公司 | Driver's recognition methods based on user behavior, apparatus and system |
| CN109409207A (en) * | 2018-09-10 | 2019-03-01 | 百度在线网络技术（北京）有限公司 | The recognition methods of passenger status, device, equipment and storage medium in unmanned vehicle |
| CN109784136A (en) * | 2017-11-15 | 2019-05-21 | 欧姆龙株式会社 | Driver's decision maker, driver condition assessment device, driver's determination method |
| CN110210980A (en) * | 2018-06-15 | 2019-09-06 | 腾讯科技（深圳）有限公司 | A kind of driving behavior appraisal procedure, device and storage medium |
| CN110443185A (en) * | 2019-07-31 | 2019-11-12 | 京东城市(北京)数字科技有限公司 | Driver's recognition methods, driver identification device, electronic equipment and storage medium |
| CN111259935A (en) * | 2020-01-09 | 2020-06-09 | 斑马网络技术有限公司 | Vehicle accident recognition method, device, equipment and storage medium |
| CN114280332A (en) * | 2021-12-31 | 2022-04-05 | 成都路行通信息技术有限公司 | Three-axis acceleration sensor correction method |
| WO2023029966A1 (en) * | 2021-08-31 | 2023-03-09 | 华为技术有限公司 | Identity authentication method, and vehicle |
| US20230174075A1 (en) * | 2021-12-03 | 2023-06-08 | Bendix Commercial Vehicle Systems, Llc | System and Method for Driving Style Driver Identity Determination and Control of Vehicle Functions |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090040054A1 (en) * | 2007-04-11 | 2009-02-12 | Nec Laboratories America, Inc. | Real-time driving danger level prediction |
| WO2015053423A1 (en) * | 2013-10-11 | 2015-04-16 | (주)나노포인트 | System and method for analyzing and diagnosing driving habit |
| CN105718710A (en) * | 2014-12-02 | 2016-06-29 | 高德软件有限公司 | Driving behavior analysis method and equipment |
| CN105761329A (en) * | 2016-03-16 | 2016-07-13 | 成都信息工程大学 | Method of identifying driver based on driving habits |
| CN106127126A (en) * | 2016-06-17 | 2016-11-16 | 杭州好好开车科技有限公司 | The recognition methods of radical driving behaviors based on three anxious data |
| CN106128099A (en) * | 2016-07-01 | 2016-11-16 | 斑马信息科技有限公司 | Driver's recognition methods and device |
-
2017
- 2017-05-24 CN CN201710374491.XA patent/CN107215307A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090040054A1 (en) * | 2007-04-11 | 2009-02-12 | Nec Laboratories America, Inc. | Real-time driving danger level prediction |
| WO2015053423A1 (en) * | 2013-10-11 | 2015-04-16 | (주)나노포인트 | System and method for analyzing and diagnosing driving habit |
| CN105718710A (en) * | 2014-12-02 | 2016-06-29 | 高德软件有限公司 | Driving behavior analysis method and equipment |
| CN105761329A (en) * | 2016-03-16 | 2016-07-13 | 成都信息工程大学 | Method of identifying driver based on driving habits |
| CN106127126A (en) * | 2016-06-17 | 2016-11-16 | 杭州好好开车科技有限公司 | The recognition methods of radical driving behaviors based on three anxious data |
| CN106128099A (en) * | 2016-07-01 | 2016-11-16 | 斑马信息科技有限公司 | Driver's recognition methods and device |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109784136A (en) * | 2017-11-15 | 2019-05-21 | 欧姆龙株式会社 | Driver's decision maker, driver condition assessment device, driver's determination method |
| CN108280482A (en) * | 2018-01-30 | 2018-07-13 | 广州小鹏汽车科技有限公司 | Driver's recognition methods based on user behavior, apparatus and system |
| CN108280482B (en) * | 2018-01-30 | 2020-10-16 | 广州小鹏汽车科技有限公司 | Driver identification method, device and system based on user behaviors |
| CN110210980A (en) * | 2018-06-15 | 2019-09-06 | 腾讯科技（深圳）有限公司 | A kind of driving behavior appraisal procedure, device and storage medium |
| CN109409207A (en) * | 2018-09-10 | 2019-03-01 | 百度在线网络技术（北京）有限公司 | The recognition methods of passenger status, device, equipment and storage medium in unmanned vehicle |
| US11315349B2 (en) | 2018-09-10 | 2022-04-26 | Apollo Intelligent Driving Technology (Beijing) Co., Ltd. | Method, apparatus and device for identifying passenger state in unmanned vehicle, and storage medium |
| CN110443185A (en) * | 2019-07-31 | 2019-11-12 | 京东城市(北京)数字科技有限公司 | Driver's recognition methods, driver identification device, electronic equipment and storage medium |
| CN110443185B (en) * | 2019-07-31 | 2020-11-24 | 北京京东智能城市大数据研究院 | Driver identification method, driver identification device, electronic device, and storage medium |
| CN111259935A (en) * | 2020-01-09 | 2020-06-09 | 斑马网络技术有限公司 | Vehicle accident recognition method, device, equipment and storage medium |
| WO2023029966A1 (en) * | 2021-08-31 | 2023-03-09 | 华为技术有限公司 | Identity authentication method, and vehicle |
| US20230174075A1 (en) * | 2021-12-03 | 2023-06-08 | Bendix Commercial Vehicle Systems, Llc | System and Method for Driving Style Driver Identity Determination and Control of Vehicle Functions |
| US11993270B2 (en) * | 2021-12-03 | 2024-05-28 | Bendix Commercial Vehicle Systems, Llc | System and method for driving style driver identity determination and control of vehicle functions |
| CN114280332A (en) * | 2021-12-31 | 2022-04-05 | 成都路行通信息技术有限公司 | Three-axis acceleration sensor correction method |
| CN114280332B (en) * | 2021-12-31 | 2024-04-23 | 成都路行通信息技术有限公司 | Triaxial acceleration sensor correction method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107215307A (en) | Driver identity recognition methods and system based on vehicle sensors correction data | |
| US10176405B1 (en) | Vehicle re-identification techniques using neural networks for image analysis, viewpoint-aware pattern recognition, and generation of multi- view vehicle representations | |
| CN106570477B (en) | Vehicle cab recognition model building method and model recognizing method based on deep learning | |
| CN106846729B (en) | Tumble detection method and system based on convolutional neural network | |
| CN107463907A (en) | Vehicle collision detection method, device, electronic equipment and vehicle | |
| CN110329271A (en) | A kind of multisensor vehicle driving detection system and method based on machine learning | |
| CN108764059A (en) | A kind of Human bodys' response method and system based on neural network | |
| CN106227341A (en) | Unmanned plane gesture interaction method based on degree of depth study and system | |
| Karim et al. | A dynamic spatial-temporal attention network for early anticipation of traffic accidents | |
| CN110163236A (en) | The training method and device of model, storage medium, electronic device | |
| CN105354986A (en) | Driving state monitoring system and method for automobile driver | |
| CN109034376A (en) | A kind of unmanned plane during flying trend prediction method and system based on LSTM | |
| CN107358242A (en) | Target area color identification method, device and monitor terminal | |
| CN107316067A (en) | A kind of aerial hand-written character recognition method based on inertial sensor | |
| CN107886061A (en) | Human bodys' response method and system based on multi-modal depth Boltzmann machine | |
| CN111753674A (en) | Fatigue driving detection and identification method based on deep learning | |
| CN109740477A (en) | Study in Driver Fatigue State Surveillance System and its fatigue detection method | |
| Tanprasert et al. | Combining unsupervised anomaly detection and neural networks for driver identification | |
| CN205230272U (en) | Driver drive state monitoring system | |
| CN106709475A (en) | Obstacle recognition method and device, computer equipment and readable storage medium | |
| CN108629170A (en) | Personal identification method and corresponding device, mobile terminal | |
| CN110377049B (en) | Brain-computer interface-based unmanned aerial vehicle cluster formation reconfiguration control method | |
| CN109255181A (en) | A kind of distribution of obstacles emulation mode, device and terminal based on multi-model | |
| CN111553326A (en) | Hand motion recognition method and device, electronic equipment and storage medium | |
| CN106599848A (en) | Depth visual feature and support vector machine-based terrain texture recognition algorithm |
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 | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Zhang Kai Inventor after: Li Zhengping Inventor after: Chen Bokui Inventor before: Zhang Kai Inventor before: Li Zhengping |
|
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170929 |