CN109345772A - A kind of fatigue driving recognition methods and relevant device - Google Patents
A kind of fatigue driving recognition methods and relevant device Download PDFInfo
- Publication number
- CN109345772A CN109345772A CN201811198603.1A CN201811198603A CN109345772A CN 109345772 A CN109345772 A CN 109345772A CN 201811198603 A CN201811198603 A CN 201811198603A CN 109345772 A CN109345772 A CN 109345772A
- Authority
- CN
- China
- Prior art keywords
- cluster
- clustering cluster
- state
- described eigenvector
- clustering
- 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
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/02—Alarms for ensuring the safety of persons
- G08B21/06—Alarms for ensuring the safety of persons indicating a condition of sleep, e.g. anti-dozing alarms
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
- B60W40/08—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to drivers or passengers
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
- B60W40/10—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to vehicle motion
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
- B60W40/10—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to vehicle motion
- B60W40/105—Speed
Abstract
This application discloses a kind of fatigue driving recognition methods and relevant devices, comprising: the car status information first in acquisition target vehicle driving process;Then according to the car status information, the feature vector of the state feature of the target vehicle is determined;Then described eigenvector is clustered, determines clustering cluster belonging to described eigenvector;The finally clustering cluster according to belonging to described eigenvector determines the driving condition of the driver of the target vehicle.Using the embodiment of the present application, the recognition accuracy of fatigue driving can be improved.
Description
Technical field
This application involves field of intelligent transportation technology and machine learning field more particularly to a kind of fatigue driving recognition methods
And relevant device.
Background technique
According to statistics, 90% traffic accident that China occurs in recent years is caused by the factor of driver, and wherein fatigue is driven
One of the main reason for driver's accident sailed.Fatigue driving refers to driver's situation in a state of fatigue in physiology or psychology
The case where lower driving vehicle, fatigue driving will lead to driver attention do not concentrate, slow in reacting or even operation error, to make
At traffic accident.Therefore real-time monitoring and identification are carried out to the driving condition of driver, fatigue driving sign occurs in driver
When send a warning in time, will be effectively less because of traffic accident caused by fatigue driving.Currently, method for detecting fatigue driving packet
Include the physiological change index (such as brain electricity, the rhythm of the heart, breathing and myoelectricity information) using biosensor detection driver.Although raw
Reason change indicator can accurately reflect fatigue driving state, but during information collection it is generally necessary on driver's body
It is pasted into surface electrode, this can not only cause the discomfort of driver, but also will affect driving operation behavior and driving safety.Furthermore
Further include utilize and its vision technique or other sensors technology detection driver external change feature (such as eyelid is blinked, is beaten
Yawn etc.) fatigue driving recognition methods, but such methods be usually from the angle of machine vision, it is special for the difference of driver
Sign (face, eyes, mouth, head) is merged, and since image procossing data volume itself is very big, is calculated along with complicated information merges
Method causes driving fatigue recognizer complexity height, hardware cost expensive, and most feature extraction is both for minority
What several pictures extracted, cause each feature extraction inaccuracy, the recognition accuracy of fatigue driving low.
Summary of the invention
The embodiment of the present application provides a kind of fatigue driving recognition methods and relevant device.The identification of fatigue driving can be improved
Accuracy rate.
The embodiment of the present application first aspect provides a kind of fatigue driving recognition methods, comprising:
Obtain the car status information in target vehicle driving process;
According to the car status information, the feature vector of the state feature of the target vehicle is determined;
Described eigenvector is clustered, determines clustering cluster belonging to described eigenvector;
The clustering cluster according to belonging to described eigenvector determines the driving condition of the driver of the target vehicle.
Wherein, described according to the car status information, determine the feature vector packet of the state feature of the target vehicle
It includes:
According to the car status information, the state feature of the target vehicle is determined;
According to the state feature, described eigenvector is determined.
Wherein, described according to the state feature, determine that described eigenvector includes:
Determine the resolution of every kind of state feature in a variety of state features;
According to the resolution, at least one generation described eigenvector is chosen from a variety of state features.
Wherein, the car status information includes at least one in travel speed, traveling acceleration and steering angle, institute
It states according to the car status information, determines that the state feature of the target vehicle includes:
According to the car status information, poor velocity standard, acceleration standard deviation, steering angular velocity and steering angle are determined
At least one of in standard deviation;
The velocity standard is poor, described acceleration standard deviation, the steering angular velocity, the steering mean angular deviation, institute
At least one stated in travel speed, the traveling acceleration and the steering angle is used as the state feature.
Wherein, the clustering cluster includes the first clustering cluster and the second clustering cluster;
It is described that described eigenvector is clustered, determine that clustering cluster belonging to described eigenvector includes:
Determine that described eigenvector belongs to the first degree of membership of first clustering cluster and described eigenvector belongs to institute
State the second degree of membership of the second clustering cluster;
When first degree of membership is greater than second degree of membership, determine that described eigenvector belongs to first cluster
Cluster;Or
When first degree of membership is not more than second degree of membership, it is poly- to determine that described eigenvector belongs to described second
Class cluster.
Wherein, the determining described eigenvector belongs to the first degree of membership and the feature of first clustering cluster
The second degree of membership that vector belongs to second clustering cluster includes:
Determine the first space length between described eigenvector and the first cluster centre of first clustering cluster and
Second space distance between described eigenvector and the second cluster centre of second clustering cluster;
According to first space length and the second space distance, determine that first degree of membership and described second is subordinate to
Category degree.
Wherein, first clustering cluster corresponds to normal condition and second clustering cluster corresponds to fatigue state;
It is described according to the clustering cluster, determine that the driving condition of the driver of the target vehicle includes:
When described eigenvector belongs to first clustering cluster, determine that the driving condition is the normal condition;Or
When described eigenvector belongs to second clustering cluster, determine that the driving condition is the fatigue state.
Wherein, described according to the clustering cluster, after the driving condition for determining the driver of the target vehicle, also wrap
It includes:
If the driving condition is fatigue state, warning information is issued, the warning information is for reminding driver to stop
Only driving behavior.
Correspondingly, the embodiment of the present application second aspect provides a kind of fatigue driving recognition device, comprising:
Module is obtained, for obtaining the car status information in target vehicle driving process;
Determining module, for according to the car status information, determine the feature of the state feature of the target vehicle to
Amount;
Cluster module determines clustering cluster belonging to described eigenvector for clustering to described eigenvector;
The determining module is also used to the clustering cluster according to belonging to described eigenvector, determines driving for the target vehicle
The driving condition for the person of sailing.
Wherein, the determining module is also used to:
According to the car status information, the state feature of the target vehicle is determined;
According to the state feature, described eigenvector is determined.
Wherein, the determining module is also used to:
Determine the resolution of every kind of state feature in a variety of state features;
According to the resolution, at least one generation described eigenvector is chosen from a variety of state features.
Wherein, the car status information includes at least one in travel speed, traveling acceleration and steering angle, institute
Determining module is stated to be also used to:
According to the car status information, poor velocity standard, acceleration standard deviation, steering angular velocity and steering angle are determined
At least one of in standard deviation;
The velocity standard is poor, described acceleration standard deviation, the steering angular velocity, the steering mean angular deviation, institute
At least one stated in travel speed, the traveling acceleration and the steering angle is used as the state feature.
Wherein, the clustering cluster includes the first clustering cluster and the second clustering cluster;
The cluster module is also used to:
Determine that described eigenvector belongs to the first degree of membership of first clustering cluster and described eigenvector belongs to institute
State the second degree of membership of the second clustering cluster;
When first degree of membership is greater than second degree of membership, determine that described eigenvector belongs to first cluster
Cluster;Or
When first degree of membership is not more than second degree of membership, it is poly- to determine that described eigenvector belongs to described second
Class cluster.
Wherein, the cluster module is also used to:
Determine the first space length between described eigenvector and the first cluster centre of first clustering cluster and
Second space distance between described eigenvector and the second cluster centre of second clustering cluster;
According to first space length and the second space distance, determine that first degree of membership and described second is subordinate to
Category degree.
Wherein, first clustering cluster corresponds to normal condition and second clustering cluster corresponds to fatigue state;
The cluster module is also used to:
When described eigenvector belongs to first clustering cluster, determine that the driving condition is the normal condition;Or
When described eigenvector belongs to second clustering cluster, determine that the driving condition is the fatigue state.
Wherein, the determining module is also used to:
If the driving condition is fatigue state, warning information is issued, the warning information is for reminding driver to stop
Only driving behavior.
The embodiment of the present application third aspect discloses a kind of electronic equipment, comprising: processor, memory, communication interface and
Bus;
The processor, the memory are connected by the bus with the communication interface and complete mutual lead to
Letter;
The memory stores executable program code;
The processor is run by reading the executable program code stored in the memory can be performed with described
The corresponding program of program code, for executing in a kind of fatigue driving recognition methods disclosed in the embodiment of the present application first aspect
Operation.
Correspondingly, this application provides a kind of storage mediums, wherein the storage medium is for storing application program, institute
Application program is stated for executing a kind of fatigue driving recognition methods disclosed in the embodiment of the present application first aspect at runtime.
Correspondingly, this application provides a kind of application programs, wherein the application program for executing this Shen at runtime
It please a kind of fatigue driving recognition methods disclosed in embodiment first aspect.
Implement the embodiment of the present application, the first car status information in acquisition target vehicle driving process, the vehicle shape
State information includes at least one in travel speed, traveling acceleration and steering angle;Then according to the car status information,
Determine the feature vector of the state feature of the target vehicle;Then described eigenvector is clustered, determines the feature
Clustering cluster belonging to vector;The finally clustering cluster according to belonging to described eigenvector determines the driver's of the target vehicle
Driving condition.The recognition accuracy of fatigue driving can be improved.
Detailed description of the invention
Technical solution in ord to more clearly illustrate embodiments of the present application, below will be to required use in embodiment description
Attached drawing be briefly described, it should be apparent that, the accompanying drawings in the following description is some embodiments of the present application, for this field
For those of ordinary skill, without creative efforts, it is also possible to obtain other drawings based on these drawings.
Fig. 1 is a kind of flow diagram of fatigue driving recognition methods provided by the embodiments of the present application;
Fig. 2 is the flow diagram of another fatigue driving recognition methods provided by the embodiments of the present application;
Fig. 3 is a kind of structural schematic diagram of fatigue driving recognition device provided by the embodiments of the present application;
Fig. 4 is the structural schematic diagram of a kind of electronic equipment provided by the embodiments of the present application.
Specific embodiment
Below in conjunction with the attached drawing in the embodiment of the present application, technical solutions in the embodiments of the present application carries out clear, complete
Site preparation description, it is clear that described embodiment is some embodiments of the present application, instead of all the embodiments.Based on this Shen
Please in embodiment, every other implementation obtained by those of ordinary skill in the art without making creative efforts
Example, shall fall in the protection scope of this application.
Referring to FIG. 1, Fig. 1 is a kind of flow diagram of fatigue driving recognition methods provided by the embodiments of the present application.Such as
Shown in figure, the method in the embodiment of the present application includes:
S101 obtains the car status information in target vehicle driving process, wherein the car status information includes row
Sail at least one in speed, traveling acceleration and steering angle.
In the specific implementation, can target vehicle travel during according to predeterminated frequency, (such as vehicle often moves forward 5 meters
(m) 1 time or 1 minute/time are acquired) it acquires the travel speed of target vehicle, travel at least one in acceleration and steering angle.
After collecting data, collected steering angle can also be converted to the value between [- 2 π, 2 π], for example, 30 ° are converted
At π/12, wherein "-" expression is counterclockwise steering angle, "+" (omit do not write under normal circumstances) expression are clockwise.
S102 determines the feature vector of the state feature of the target vehicle according to the car status information.
In the specific implementation, the state feature of the target vehicle can be determined first according to the car status information.Its
In, it can be according to collected travel speed calculating speed standard deviation (vstd) and according to collect traveling acceleration calculation add
Poor (a of velocity standardstd) and steering angular velocity (ω) is determined according to steering angle and turns to mean angular deviation (θstd), steering angle speed
Degree can be the amplitude of variation and the ratio of time used in the generation amplitude of variation of steering angle.Wherein, vstdIt can reflect driving
Member is to the control ability of speed, and speed is more steady, and standard deviation is smaller, and acceleration a is also reflection driver to vehicle longitudinal control
One of index of ability, acceleration low frequency signal, which represents, to be accelerated gently, and it is frequent that high-frequency signal represents acceleration;Tired brief acceleration frequency
Composing from low frequency becomes the control ability decline that high frequency illustrates driver to speed, accelerates to neglect and neglect slowly fastly, speed fluctuations are fast.
It therefore, can will traveling acceleration, velocity standard be poor, travel acceleration, acceleration standard deviation, steering angle, steering
Angular speed and at least one turned in mean angular deviation are used as state feature.Wherein, travel speed and traveling acceleration can
Be any primary collected and steering angular velocity can be according to multi collect to steering angle in arbitrarily adopt twice
What the steering angle collected was calculated.In addition, the accuracy in order to guarantee data, travel speed be can also be average traveling speed
Degree, traveling acceleration can also be that average traveling acceleration and steering angular velocity can also be average steering angular velocity.
Then, according to the state feature, described eigenvector is determined.Such as: since 0 moment, in t1, t2 and t3
Quarter has carried out data acquisition to vehicle, obtains travel speed v1、v2And v3, travel acceleration a1、a2And a3And steering angle θ1、θ2
And θ3.Wherein it is possible to calculate v first1、v2And v3Average value obtain average overall travel speedAnd according toIt obtainsThen a is calculated1、a2And a3Mean value obtain average traveling plus
SpeedAnd according toIt obtainsThen steering angle is calculated
θ1、θ2And θ3Average value obtainAssuming that steering angle before time tl is 0, then it can basisObtain steering angle speed
Spend ω1=(θ1-0)/(t1-0)、ω2=(θ2-θ1)/(t2-t1) and ω3=(θ3-θ2)/(t3-t2), and calculate average steering angle
SpeedFinally, obtaining feature vector
S103 clusters described eigenvector, determines the affiliated clustering cluster of described eigenvector.
In the specific implementation, cluster refers to by measuring the similitude between data, different aggregates of data is sorted data into
In process.Driving condition can be divided into fatigue state and normal condition in practice, and therefore, clustering cluster includes fatigue state pair
Corresponding second clustering cluster of the first clustering cluster and normal condition answered, wherein it is poly- can to determine that feature vector belongs to first first
The first degree of membership and feature vector of class cluster belong to the second degree of membership of the second clustering cluster.Then when the first degree of membership is greater than
When the second degree of membership, determine that feature vector belongs to the first clustering cluster.When the first degree of membership is not more than the second degree of membership, determine special
Sign vector belongs to the second clustering cluster.Then the first space length of feature vector and the first cluster centre can be calculated, and (such as Europe is several
In distance) and second space distance with the second cluster centre, wherein the first cluster centre and the second cluster centre are all
It is vector identical with the dimension of feature vector;Finally according to the first space length and second space distance, determine that first is subordinate to
Degree and second be subordinate to, wherein can first the first space length of computer and second space distance and value, then by second space
Distance is with the quotient of this and value as the first degree of membership and using the first space length and the quotient that should and be worth as the second degree of membership.
Such as: feature vector is X=(2,3), the first cluster centre is (4,9) m1=, the second cluster centre be m2=(0,
6) space length of X and m1, then can be calculated firstAnd calculate X's and m2
Space lengthThen X is calculated with respect to the clustering cluster where the first cluster centre
Degree of membership isAnd X is with respect to the clustering cluster where the second cluster centre
Degree of membership be1 < μ 2 of obvious μ, so that it is determined that feature vector, X category
Clustering cluster where the second cluster centre.
In order to determine the first cluster centre and the first clustering cluster and the second cluster centre and the second clustering cluster.It can obtain
The car status information of more vehicles of known driver's driving condition is taken, and determines multiple feature vector training samples, such as X1,
X2,...,Xn.Then clustering algorithm is utilized, such as K- mean value fuzzy clustering algorithm, to X1,X2,...,XnIt is clustered, to obtain
First cluster centre and the first clustering cluster and the second cluster centre and the second clustering cluster.
In order to guarantee the accuracy of cluster centre, the status information of the more vehicles travelled on different sections of highway can be acquired,
And the 5m that often moves forward according to vehicle acquires primary frequency and carries out information collection.For example, 5 kinds as shown in Table 1 can be acquired
The status information of vehicle on section.
1.5 kinds of test segments of table
Section | Basic combining form | Radius | Preceding ratio of slope | Ratio of slope afterwards |
Section 1 | Directly-slow-circle-is slow-straight | 1200m | - 2.96% | 2.42% |
Section 2 | Directly-slow-circle-is slow-straight | 1000m | 0 | 0 |
Section 3 | Directly-slow-circle-is slow-straight | 1300m | - 2.21% | - 2.21% |
Section 4 | Directly-slow-circle-is slow-straight | 2000m | 1.24% | - 3.57% |
Section 5 | Directly-slow-circle-is slow-straight | It is just infinite | 0 | 0 |
Wherein, in K- mean value fuzzy clustering algorithm, can set first the quantity of clustering cluster as 2 (the first clustering cluster and
Second clustering cluster), and by the first cluster centre M of the first clustering cluster1It is initialized as X1, will be in the second cluster of the second clustering cluster
Heart M2It is initialized as X2.Followed by X3,X4,...,XnTwo cluster centres are updated until convergence, to obtain in two clusters
The heart, wherein updating cluster centre, specific step is as follows:
(1) is from Xk(k=3,4 ..., n) starts with (1) formula and calculates separately each feature vector to two cluster centres
The degree of membership of the clustering cluster at place;
Wherein, | | | |2Indicate square of Euclidean distance.
(2) if μ1(Xk)≥μ2(Xk), then M is updated according to (2) formula1.Conversely, then updating M according to (2) formula2;
Wherein, j is equal to 1 or 2.
(3) if update before M1With the M after update1Difference be less than preset threshold (such as 10-5) or update before
M2With the M after update2Difference be less than preset threshold (such as 10-5), then k=k+1 is enabled, and jump to step (1) and followed
Ring;Otherwise end loop, and determine that the first cluster centre is M1, the second cluster centre is M2.For example, M1=0.082,0.664,
0.954,0.431,9.890,0.451,0.632 }, M2=0.041,0.629,3.072,0.634,1.971,0.76,
0.303}。
After determining cluster centre, training sample X can also be determined according to (1) formula1,X2,...,XnIt is poly- relative to first
The degree of membership of class cluster and degree of membership relative to the second clustering cluster, determine the clustering cluster belonging to them, to obtain the first cluster
Cluster and the second clustering cluster.
S104, the clustering cluster according to belonging to described eigenvector determine the driving condition of the driver of the target vehicle.
In the specific implementation, the first clustering cluster is corresponding with fatigue state, the second clustering cluster is corresponding with normal condition.Cause
This determines that driving condition is normal condition when feature vector belongs to the first clustering cluster;When feature vector belongs to the second clustering cluster
When, determine that driving condition is fatigue state.
Optionally, however, it is determined that the driving condition of driver is fatigue state, then issues warning information to remind driver
Stop driving behavior.For example it can beep or play voice reminder information.
In the embodiment of the present application, the car status information in target vehicle driving process, the vehicle shape are obtained first
State information includes at least one in travel speed, traveling acceleration and steering angle;Then according to the car status information,
Determine the feature vector of the state feature of the target vehicle;Then described eigenvector is clustered, determines the feature
The affiliated clustering cluster of vector;The finally clustering cluster according to belonging to described eigenvector determines driving for the driver of the target vehicle
Sail state.The recognition accuracy of fatigue driving can be improved.
Referring to FIG. 2, Fig. 2 is the flow diagram of another fatigue driving recognition methods provided by the embodiments of the present application.
As shown, the method in the embodiment of the present application includes:
S201 obtains the car status information in target vehicle driving process, wherein car status information includes traveling speed
Degree, traveling acceleration and steering angle look at least one of.This step is identical as the step S101 in a upper embodiment, this step
It repeats no more.
S202 determines the various states feature of target vehicle according to car status information.
In the specific implementation, can determine that velocity standard is poor, determines acceleration scale according to traveling acceleration according to travel speed
Standard is poor and determines steering angular velocity according to steering angle and turns to mean angular deviation.Wherein, travel speed, velocity standard are poor, travel
Any one in acceleration, acceleration standard deviation, steering angle and steering angular velocity can be used as state feature.Wherein, row
It sails speed and travels acceleration and can be any primary collected and steering angular velocity and can be and arrived according to multi collect
What any steering angle collected twice in steering angle was calculated.Meanwhile the accuracy in order to guarantee data, travel speed
It can also be and average overall travel speed, traveling acceleration can also be that average traveling acceleration and steering angular velocity can be with
It is average steering angular velocity.
S203, according to the resolution of various states feature, chosen from various states feature it is at least one generate feature to
Amount.
In the specific implementation, can determine various states spy first to further increase the recognition accuracy of fatigue driving
The resolution of every kind of state feature in sign.Wherein, driving condition includes fatigue state and two kinds of normal condition, which can be with
Including the resolution to normal condition and to the resolution of fatigue state.
In order to determine resolution, the car status information of more vehicles of available known driver's driving condition is determined
Various states feature, such as travel speed, velocity standard difference and traveling acceleration.Then it is directed to every kind of state feature, determination is more
A feature vector, such as X1,X2,...,Xn, wherein each feature vector is 1 dimensional vector.For example, being collected for travel speed
Know n sample v of driving condition1=50km/h, v2=69km/h ..., vn=54km/h, it is determined that X1={ 50 }, X2=
{69},...,Xn={ 54 }.Then clustering algorithm is utilized, such as K- mean value fuzzy clustering algorithm, to X1,X2,...,XnGathered
Class obtains corresponding first clustering cluster of normal condition and corresponding second clustering cluster of fatigue state.Finally, statistics is true to drive shape
State is fatigue state and the first cumulative amount of feature vector for being classified into the second clustering cluster, and by the first cumulative amount
Divided by true driving condition be fatigue state feature vector total quantity quotient as corresponding states feature to fatigue state
Resolution.That true driving condition is normal condition and the feature vector that is classified into the first clustering cluster can also be counted
Two cumulative amounts, using the second cumulative amount divided by true driving condition be normal condition feature vector total quantity quotient as
Resolution of the corresponding states feature to normal condition.Further, it is also possible to obtain the whole resolution of every kind of state feature, wherein
Whole resolution can be the quotient of the first cumulative amount and the second cumulative amount divided by the total quantity of feature vector.
For example, as shown in table 2, every kind of state feature is respectively obtained to the resolution of normal condition and fatigue state, and
Whole resolution.
The resolution of the every kind of significant condition of table 2.
v | vstd | a | astd | θ | ω | θstd | |
Total sample number | 388 | 343 | 368 | 382 | 328 | 392 | 408 |
Normal condition | 187 | 215 | 255 | 245 | 120 | 201 | 218 |
Fatigue state | 201 | 128 | 113 | 137 | 208 | 191 | 190 |
Normal condition resolution (%) | 66.97 | 76.79 | 91.07 | 87.5 | 42.86 | 71.79 | 77.86 |
Fatigue state resolution (%) | 95.71 | 60.95 | 53.81 | 65.24 | 99.05 | 90.95 | 90.48 |
Whole resolution (%) | 79.18 | 70.00 | 75.10 | 77.96 | 66.94 | 80.00 | 83.27 |
It, can be from least one shape of the various states Feature Selection of target vehicle after the resolution for determining every kind of state
State feature generates the identification that feature vector carries out fatigue driving.Such as: according to table 2, v and θ divide the resolution of fatigue state
Not Wei 95.71% and 99.05%, be above other state feature, a and astdResolution to normal condition is respectively 91.07%
With 87.5%, it is above other state features, and θstdWhole discrimination highest.Therefore can from v, a of target vehicle,
astd, θ and θstdIn optionally one or more be combined to obtain feature vector.
For another example, v and a can be combined to obtain feature vector { v, a }, can also be by a and θstdObtain feature vector
It { v, a }, can also be by astdIt is combined to obtain feature vector { a with θstd,θ}.Wherein, for every kind of group in this 3 kinds combinations
It closes, the first cluster centre M of the first clustering cluster (normal condition)1With the second cluster centre M of the second clustering cluster (fatigue state)2
As table 3 shows.
The cluster centre that feature combines two-by-two of table 3.
It optionally, can also be from v, v of target vehiclestd、a、astd, θ, ω and θstdIn arbitrarily choose a kind of feature shape
State generates feature vector, to carry out fatigue driving identification, wherein is directed to every kind of significant condition, the first clustering cluster (normal condition)
The first cluster centre M1With the second cluster centre M of the second clustering cluster (fatigue state)2As table 4 shows.
The cluster centre of the single feature of table 4.
S204 clusters feature vector, determines clustering cluster belonging to feature vector.In this step and a upper embodiment
Step S103 it is identical, this step repeats no more.
S205, the clustering cluster according to belonging to feature vector determine the driving condition of the driver of target vehicle.This step with
Step S104 in a upper embodiment is identical, this step repeats no more.
In the embodiment of the present application, the car status information in target vehicle driving process, the vehicle shape are obtained first
State information includes at least one in travel speed, traveling acceleration and steering angle;Then according to the car status information,
Determine the feature vector of the state feature of the target vehicle;Then described eigenvector is clustered, determines the feature
The affiliated clustering cluster of vector;The finally clustering cluster according to belonging to described eigenvector determines driving for the driver of the target vehicle
Sail state.The recognition accuracy of fatigue driving can be improved.
Referring to FIG. 3, Fig. 3 is a kind of structural schematic diagram of fatigue driving recognition device provided by the embodiments of the present application.Such as
Shown in figure, the device in the embodiment of the present application includes:
Module 301 is obtained, for obtaining the car status information in target vehicle driving process.
In the specific implementation, car status information includes at least one in travel speed, traveling acceleration and steering angle.
In order to make acquisition make car status information more can be accurate, can be acquired during target vehicle travels according to predeterminated frequency
At least one of in the travel speed of target vehicle, traveling acceleration and steering angle.
Determining module 302, for determining the feature of the state feature of the target vehicle according to the car status information
Vector.
In the specific implementation, the state feature of the target vehicle can be determined first according to the car status information.Its
In, it can be according to collected travel speed calculating speed standard deviation (vstd) and according to collect traveling acceleration calculation add
Poor (a of velocity standardstd) and steering angular velocity (ω) is determined according to steering angle and turns to mean angular deviation (θstd), steering angle speed
Degree can be the amplitude of variation and the ratio of time used in the generation amplitude of variation of steering angle.Wherein, vstdIt can reflect driving
Member is to the control ability of speed, and speed is more steady, and standard deviation is smaller, and acceleration a is also reflection driver to vehicle longitudinal control
One of index of ability, acceleration low frequency signal, which represents, to be accelerated gently, and it is frequent that high-frequency signal represents acceleration;Tired brief acceleration frequency
Composing from low frequency becomes the control ability decline that high frequency illustrates driver to speed, accelerates to neglect and neglect slowly fastly, speed fluctuations are fast.
Therefore, can will traveling acceleration, travel speed standard deviation, traveling acceleration, acceleration standard deviation, steering angle,
Steering angular velocity and at least one turned in mean angular deviation are used as state feature.Wherein, travel speed and traveling accelerate
Degree can be any primary collected and steering angular velocity can be according to multi collect to steering angle in any two
What secondary collected steering angle was calculated.In addition, the accuracy in order to guarantee data, travel speed can also be average row
Sail speed, traveling acceleration can also be that average traveling acceleration and steering angular velocity can also be average steering angular velocity.
Then, according to the state feature, described eigenvector is determined.
Cluster module 303 determines the affiliated clustering cluster of described eigenvector for clustering to described eigenvector.
In the specific implementation, cluster refers to by measuring the similitude between data source, data source is categorized into different numbers
According to the process in cluster.Driving condition can be divided into fatigue state and normal condition in practice, and therefore, clustering cluster includes tired shape
Corresponding first clustering cluster of state and corresponding second clustering cluster of normal condition, wherein can determine that feature vector belongs to first
The first degree of membership and feature vector of one clustering cluster belong to the second degree of membership of the second clustering cluster, then when the first degree of membership
When greater than the second degree of membership, determine that feature vector belongs to the first clustering cluster.When the first degree of membership is not more than the second degree of membership, really
Determine feature vector and belongs to the second clustering cluster.Wherein, the cluster centre of the first clustering cluster is the first cluster centre, the second clustering cluster
Cluster centre is the second cluster centre, wherein the first cluster centre and the second cluster centre are all the dimension phases with feature vector
Same vector.It is then possible to calculate the first space length (such as Euclidean distance) of feature vector and the first cluster centre, with
And the second space distance with the second cluster centre;Finally according to the first space length and second space distance, the first person in servitude is determined
Category degree and second is subordinate to, wherein can first the first space length of computer and second space distance and value, then by the second sky
Between distance with should and the quotient of value be subordinate to as second as the first degree of membership and using the first space length and the quotient of this and value
Degree.
In order to determine the first cluster centre and the first clustering cluster and the second cluster centre and the second clustering cluster.Obtain mould
Block 301 is also used to obtain the car status information of more vehicles of known driver's driving condition, and determines multiple feature vectors, such as
X1,X2,...,Xn, wherein in order to guarantee the accuracy of cluster centre, the more vehicles travelled on different sections of highway can be acquired
Status information, and the 5m that often moves forward according to vehicle acquires primary frequency and carries out information collection.Cluster module 303 is also used to so
Clustering algorithm is utilized afterwards, such as K- mean value fuzzy clustering algorithm, to X1,X2,...,XnIt is clustered, wherein fuzzy poly- in K- mean value
In class algorithm, the quantity of clustering cluster can be set first as 2 (the first clustering cluster and the second clustering clusters), and by the first clustering cluster
First cluster centre M1It is initialized as X1, by the second cluster centre M of the second clustering cluster2It is initialized as X2.Followed by X3,
X4,...,XnTwo cluster centres are updated until convergence, to obtain two cluster centres.
Optionally, cluster module 303 is also used to the resolution according to various states feature, chooses from various states feature
At least one generates feature vector.Specifically, it in order to further increase the recognition accuracy of fatigue driving, can determine first more
The resolution of every kind of state feature in kind state feature.Wherein, driving condition includes fatigue state and two kinds of normal condition, the knowledge
It Du not may include the resolution to normal condition and the resolution to fatigue state.
In order to determine resolution, the car status information of more vehicles of known driver's driving condition can be collected, is determined
Various states feature, such as travel speed, velocity standard difference and traveling acceleration.Then it is directed to every kind of state feature, determination is more
A feature vector, such as X1,X2,...,Xn, wherein each feature vector is 1 dimensional vector.Then clustering algorithm is utilized, such as K- mean value
Fuzzy clustering algorithm, to X1,X2,...,XnIt is clustered, obtains corresponding first clustering cluster of normal condition and fatigue state is corresponding
The second clustering cluster.Finally, counting the feature vector that true driving condition is fatigue state and is classified into the second clustering cluster
The first cumulative amount, and by the first cumulative amount divided by the total quantity for the feature vector that true driving condition is fatigue state
Quotient is as corresponding states feature to the resolution of fatigue state.True driving condition can also be counted to be normal condition and returned
Second cumulative amount divided by true driving condition is normal to the second cumulative amount of the feature vector in the first clustering cluster by class
The quotient of the total quantity of the feature vector of state is as corresponding states feature to the resolution of normal condition.It, can using the above method
With the resolution of every kind of state feature of determination.Further, it is also possible to obtain the whole resolution of every kind of state feature, wherein whole
Resolution can be the quotient of the first cumulative amount and the second cumulative amount divided by the total quantity of feature vector.
It, can be from least one shape of the various states Feature Selection of target vehicle after the resolution for determining every kind of state
State feature generates the identification that feature vector carries out fatigue driving.
Optionally, cluster module 303 can also be from v, v of target vehiclestd、a、astd, θ, ω and θstdIn arbitrarily choose
A kind of significant condition generation feature vector, to carry out fatigue driving identification.
Determining module 302 is also used to the clustering cluster according to belonging to described eigenvector, determines the driving of the target vehicle
The driving condition of member.
In the specific implementation, the first clustering cluster is corresponding with fatigue state, the second clustering cluster is corresponding with normal condition.Cause
This determines that driving condition is normal condition when feature vector belongs to the first clustering cluster;When feature vector belongs to the second clustering cluster
When, determine that driving condition is fatigue state.
Optionally, however, it is determined that the driving condition of driver is fatigue state, and determining module 302 is also used to issue alarm letter
It ceases to remind driver to stop driving behavior.For example it can beep.
In the embodiment of the present application, the car status information in target vehicle driving process, the vehicle shape are obtained first
State information includes at least one in travel speed, traveling acceleration and steering angle;Then according to the car status information,
Determine the feature vector of the state feature of the target vehicle;Then described eigenvector is clustered, determines the feature
The affiliated clustering cluster of vector;The finally clustering cluster according to belonging to described eigenvector determines driving for the driver of the target vehicle
Sail state.The recognition accuracy of fatigue driving can be improved.
Referring to FIG. 4, Fig. 4 is the structural schematic diagram for a kind of electronic equipment that the embodiment of the present application proposes.As shown, should
Electronic equipment may include: at least one processor 401, such as CPU, at least one communication interface 402, at least one processor
403, at least one bus 404.Wherein, bus 404 is for realizing the connection communication between these components.Wherein, the application is real
The communication interface 402 for applying electronic equipment in example is wired sending port, or wireless device, for example including antenna assembly,
For carrying out the communication of signaling or data with other node devices.Memory 403 can be high speed RAM memory, be also possible to
Non-labile memory (non-volatile memory), for example, at least a magnetic disk storage.Memory 403 is optional
It can also be that at least one is located remotely from the storage device of aforementioned processor 401.Batch processing code is stored in memory 403,
And processor 401 is used to call the program code stored in memory, for performing the following operations:
Obtain the car status information in target vehicle driving process;
According to the car status information, the feature vector of the state feature of the target vehicle is determined;
Described eigenvector is clustered, determines the affiliated clustering cluster of described eigenvector;
The clustering cluster according to belonging to described eigenvector determines the driving condition of the driver of the target vehicle.
Wherein, processor 401 is also used to perform the following operations step:
According to the car status information, the state feature of the target vehicle is determined;
According to the state feature, described eigenvector is determined.
Wherein, processor 401 is also used to perform the following operations step:
Determine the resolution of every kind of state feature in a variety of state features;
According to the resolution, at least one generation described eigenvector is chosen from a variety of state features.
Wherein, the car status information includes at least one in travel speed, traveling acceleration and steering angle, place
Reason device 401 is also used to perform the following operations step:
According to the car status information, poor velocity standard, acceleration standard deviation, steering angular velocity and steering angle are determined
At least one of in standard deviation;
The velocity standard is poor, described acceleration standard deviation, the steering angular velocity, the steering mean angular deviation, institute
At least one stated in travel speed, the traveling acceleration and the steering angle is used as the state feature.
Wherein, the clustering cluster includes the first clustering cluster and the second clustering cluster;
Processor 401 is also used to perform the following operations step:
Determine that described eigenvector belongs to the first degree of membership of first clustering cluster and described eigenvector belongs to institute
State the second degree of membership of the second clustering cluster;
When first degree of membership is greater than second degree of membership, determine that described eigenvector belongs to first cluster
Cluster;Or
When first degree of membership is not more than second degree of membership, it is poly- to determine that described eigenvector belongs to described second
Class cluster.
Wherein, processor 401 is also used to perform the following operations step:
Determine the first space length between described eigenvector and the first cluster centre of first clustering cluster and
Second space distance between described eigenvector and the second cluster centre of second clustering cluster;
According to first space length and the second space distance, determine that first degree of membership and described second is subordinate to
Category degree.
Wherein, first clustering cluster corresponds to normal condition and second clustering cluster corresponds to fatigue state;
Processor 401 is also used to perform the following operations step:
When described eigenvector belongs to first clustering cluster, determine that the driving condition is the normal condition;Or
When described eigenvector belongs to second clustering cluster, determine that the driving condition is the fatigue state.
Wherein, processor 401 is also used to perform the following operations step:
If the driving condition is fatigue state, warning information is issued, the warning information is for reminding driver to stop
Only driving behavior.
It should be noted that the embodiment of the present application also provides a kind of storage medium simultaneously, the storage medium is for storing
Application program, the application program are set for executing electronics in Fig. 1 and a kind of fatigue driving recognition methods shown in Fig. 2 at runtime
The standby operation executed.
It should be noted that the embodiment of the present application also provides a kind of application program simultaneously, the application program is for transporting
The operation that electronic equipment executes in Fig. 1 and a kind of fatigue driving recognition methods shown in Fig. 2 is executed when row.
In the above-described embodiments, can come wholly or partly by software, hardware, firmware or any combination thereof real
It is existing.When implemented in software, it can entirely or partly realize in the form of a computer program product.The computer program
Product includes one or more computer instructions.When loading on computers and executing the computer program instructions, all or
It partly generates according to process or function described in the embodiment of the present application.The computer can be general purpose computer, dedicated meter
Calculation machine, computer network or other programmable devices.The computer instruction can store in computer readable storage medium
In, or from a computer readable storage medium to the transmission of another computer readable storage medium, for example, the computer
Instruction can pass through wired (such as coaxial cable, optical fiber, number from a web-site, computer, server or data center
User's line (DSL)) or wireless (such as infrared, wireless, microwave etc.) mode to another web-site, computer, server or
Data center is transmitted.The computer readable storage medium can be any usable medium that computer can access or
It is comprising data storage devices such as one or more usable mediums integrated server, data centers.The usable medium can be with
It is magnetic medium, (for example, floppy disk, hard disk, tape), optical medium (for example, DVD) or semiconductor medium (such as solid state hard disk
Solid State Disk (SSD)) etc..Above-described specific embodiment, to the purpose of the application, technical solution and beneficial
Effect has been further described.Within the spirit and principles of this application, made any modification, equivalent replacement,
Improve etc., it should be included within the scope of protection of this application.
Claims (10)
1. a kind of fatigue driving recognition methods, which is characterized in that the described method includes:
Obtain the car status information in target vehicle driving process;
According to the car status information, the feature vector of the state feature of the target vehicle is determined;
Described eigenvector is clustered, determines clustering cluster belonging to described eigenvector;
The clustering cluster according to belonging to described eigenvector determines the driving condition of the driver of the target vehicle.
2. the method as described in claim 1, which is characterized in that it is described according to the car status information, determine the target
The feature vector of the state feature of vehicle includes:
According to the car status information, the state feature of the target vehicle is determined;
According to the state feature, described eigenvector is determined.
3. method according to claim 2, which is characterized in that it is described according to the state feature, determine described eigenvector
Include:
Determine the resolution of every kind of state feature in a variety of state features;
According to the resolution, at least one generation described eigenvector is chosen from a variety of state features.
4. method according to claim 2, which is characterized in that the car status information includes travel speed, traveling acceleration
At least one of in degree and steering angle, it is described according to the car status information, determine the state of the target vehicle
Feature includes:
According to the car status information, poor velocity standard, acceleration standard deviation, steering angular velocity and steering angle standard are determined
At least one of in difference;
The velocity standard is poor, described acceleration standard deviation, the steering angular velocity, the steering mean angular deviation, the row
At least one sailed in speed, the traveling acceleration and the steering angle is used as the state feature.
5. the method as described in claim 1, which is characterized in that the clustering cluster includes the first clustering cluster and the second clustering cluster;
It is described that described eigenvector is clustered, determine that clustering cluster belonging to described eigenvector includes:
Determine that described eigenvector belongs to the first degree of membership of first clustering cluster and described eigenvector belongs to described
Second degree of membership of two clustering clusters;
When first degree of membership is greater than second degree of membership, determine that described eigenvector belongs to first clustering cluster;
Or
When first degree of membership is not more than second degree of membership, determine that described eigenvector belongs to second cluster
Cluster.
6. method as claimed in claim 5, which is characterized in that the determining described eigenvector belongs to first clustering cluster
The first degree of membership and described eigenvector belong to the second degree of membership of second clustering cluster and include:
Determine the first space length between described eigenvector and the first cluster centre of first clustering cluster and described
Second space distance between feature vector and the second cluster centre of second clustering cluster;
According to first space length and the second space distance, determine that first degree of membership and described second is subordinate to
Degree.
7. method as claimed in claim 5, which is characterized in that first clustering cluster corresponds to normal condition and described
Two clustering clusters correspond to fatigue state;
It is described according to the clustering cluster, determine that the driving condition of the driver of the target vehicle includes:
When described eigenvector belongs to first clustering cluster, determine that the driving condition is the normal condition;Or
When described eigenvector belongs to second clustering cluster, determine that the driving condition is the fatigue state.
8. the method according to claim 1 to 7, which is characterized in that it is described according to the clustering cluster, determine the mesh
After the driving condition for marking the driver of vehicle, further includes:
If the driving condition is fatigue state, warning information is issued, the warning information is for reminding driver to stop driving
Sail behavior.
9. a kind of fatigue driving recognition device, which is characterized in that described device includes:
Module is obtained, for obtaining the car status information in target vehicle driving process;
Determining module, for determining the feature vector of the state feature of the target vehicle according to the car status information;
Cluster module determines clustering cluster belonging to described eigenvector for clustering to described eigenvector;
The determining module is also used to the clustering cluster according to belonging to described eigenvector, determines the driver of the target vehicle
Driving condition.
10. a kind of electronic equipment characterized by comprising processor, memory, communication interface and bus;
The processor, the memory are connected by the bus with the communication interface and complete mutual communication;
The memory stores executable program code;
The processor is run and the executable program by reading the executable program code stored in the memory
The corresponding program of code, for executing such as the described in any item fatigue driving recognition methods of claim 1-8.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811198603.1A CN109345772A (en) | 2018-10-15 | 2018-10-15 | A kind of fatigue driving recognition methods and relevant device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811198603.1A CN109345772A (en) | 2018-10-15 | 2018-10-15 | A kind of fatigue driving recognition methods and relevant device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109345772A true CN109345772A (en) | 2019-02-15 |
Family
ID=65310337
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811198603.1A Pending CN109345772A (en) | 2018-10-15 | 2018-10-15 | A kind of fatigue driving recognition methods and relevant device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109345772A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110045371A (en) * | 2019-04-28 | 2019-07-23 | 软通智慧科技有限公司 | A kind of identification method, device, equipment and storage medium |
CN112396805A (en) * | 2019-07-31 | 2021-02-23 | Oppo广东移动通信有限公司 | Fatigue driving reminding method, device, terminal and storage medium |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104036619A (en) * | 2013-03-04 | 2014-09-10 | 德尔福电子(苏州)有限公司 | Fatigue detection method and detection apparatus based on turn angle data of steering wheel |
CN104268577A (en) * | 2014-06-27 | 2015-01-07 | 大连理工大学 | Human body behavior identification method based on inertial sensor |
CN104527647A (en) * | 2014-12-15 | 2015-04-22 | 清华大学 | Monitoring and evaluation method of driving behavior risk degrees |
CN104680168A (en) * | 2015-03-10 | 2015-06-03 | 无锡桑尼安科技有限公司 | Extraction equipment of feature vectors for highway landform recognition |
CN106384129A (en) * | 2016-09-13 | 2017-02-08 | 西安科技大学 | Driver driving state recognition method based on closed loop driving model |
CN106740863A (en) * | 2016-11-30 | 2017-05-31 | 武汉长江通信智联技术有限公司 | Driving behavior analysis method |
CN106815312A (en) * | 2016-12-21 | 2017-06-09 | 东软集团股份有限公司 | A kind of driver's evaluation method and device |
DE102016200973A1 (en) * | 2016-01-25 | 2017-07-27 | Robert Bosch Gmbh | A method for detecting a fatigue of a driver of a vehicle and control unit |
CN107316354A (en) * | 2017-07-12 | 2017-11-03 | 哈尔滨工业大学 | A kind of method for detecting fatigue driving based on steering wheel and GNSS data |
CN108229304A (en) * | 2017-11-17 | 2018-06-29 | 清华大学 | A kind of driving behavior recognition methods based on Clustering of systematization |
CN108280415A (en) * | 2018-01-17 | 2018-07-13 | 武汉理工大学 | Driving behavior recognition methods based on intelligent mobile terminal |
-
2018
- 2018-10-15 CN CN201811198603.1A patent/CN109345772A/en active Pending
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104036619A (en) * | 2013-03-04 | 2014-09-10 | 德尔福电子(苏州)有限公司 | Fatigue detection method and detection apparatus based on turn angle data of steering wheel |
CN104268577A (en) * | 2014-06-27 | 2015-01-07 | 大连理工大学 | Human body behavior identification method based on inertial sensor |
CN104527647A (en) * | 2014-12-15 | 2015-04-22 | 清华大学 | Monitoring and evaluation method of driving behavior risk degrees |
CN104680168A (en) * | 2015-03-10 | 2015-06-03 | 无锡桑尼安科技有限公司 | Extraction equipment of feature vectors for highway landform recognition |
DE102016200973A1 (en) * | 2016-01-25 | 2017-07-27 | Robert Bosch Gmbh | A method for detecting a fatigue of a driver of a vehicle and control unit |
CN106384129A (en) * | 2016-09-13 | 2017-02-08 | 西安科技大学 | Driver driving state recognition method based on closed loop driving model |
CN106740863A (en) * | 2016-11-30 | 2017-05-31 | 武汉长江通信智联技术有限公司 | Driving behavior analysis method |
CN106815312A (en) * | 2016-12-21 | 2017-06-09 | 东软集团股份有限公司 | A kind of driver's evaluation method and device |
CN107316354A (en) * | 2017-07-12 | 2017-11-03 | 哈尔滨工业大学 | A kind of method for detecting fatigue driving based on steering wheel and GNSS data |
CN108229304A (en) * | 2017-11-17 | 2018-06-29 | 清华大学 | A kind of driving behavior recognition methods based on Clustering of systematization |
CN108280415A (en) * | 2018-01-17 | 2018-07-13 | 武汉理工大学 | Driving behavior recognition methods based on intelligent mobile terminal |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110045371A (en) * | 2019-04-28 | 2019-07-23 | 软通智慧科技有限公司 | A kind of identification method, device, equipment and storage medium |
CN112396805A (en) * | 2019-07-31 | 2021-02-23 | Oppo广东移动通信有限公司 | Fatigue driving reminding method, device, terminal and storage medium |
CN112396805B (en) * | 2019-07-31 | 2022-10-25 | Oppo广东移动通信有限公司 | Fatigue driving reminding method, device, terminal and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105444770B (en) | Track level map generation and localization method based on smart mobile phone | |
CN108629231B (en) | Obstacle detection method, apparatus, device and storage medium | |
CN107784597B (en) | Travel mode identification method and device, terminal equipment and storage medium | |
CN110501169A (en) | Diagnostic method, device and the electronic equipment of vehicle trouble | |
CN108519094A (en) | Local paths planning method and cloud processing end | |
US10152061B2 (en) | Drive assist apparatus | |
CN109345772A (en) | A kind of fatigue driving recognition methods and relevant device | |
CN103575279B (en) | Based on Data Association and the system of fuzzy message | |
CN108814618A (en) | A kind of recognition methods of motion state, device and terminal device | |
CN107505617A (en) | A kind of radar target method of testing | |
CN110458083A (en) | A kind of lane line vectorization method, device and storage medium | |
CN114970705A (en) | Driving state analysis method, device, equipment and medium based on multi-sensing data | |
CN109918718A (en) | Unmanned plane during flying method for evaluating quality, device, storage medium and electronic equipment | |
CN105953793A (en) | Identification method and apparatus for sharp turn of vehicle | |
CN108986255A (en) | Electronic device, driving behavior methods of marking and storage medium | |
CN106448258A (en) | Parking space state detection method and device | |
WO2024007972A1 (en) | Object association method, computer device, computer readable storage medium, and vehicle | |
CN115546743B (en) | Vehicle road cooperative control method, device, equipment and medium based on adhesion coefficient | |
CN105427625B (en) | Vehicle turning recognition methods and device | |
CN116010543A (en) | Lane information determination method, lane information determination device, electronic equipment and storage medium | |
CN105957182A (en) | Method and apparatus for correcting linear direction used for directing vehicle driving | |
CN114021629B (en) | Vehicle track motion mode extraction method based on mean dynamic time warping | |
CN106537090A (en) | Device and method for determining at least one position of a mobile terminal | |
CN109214466A (en) | A kind of novel clustering algorithm based on density | |
CN114954532A (en) | Lane line determination method, device, equipment and storage medium |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190215 |