CN109543651A - A kind of driver's dangerous driving behavior detection method - Google Patents
A kind of driver's dangerous driving behavior detection method Download PDFInfo
- Publication number
- CN109543651A CN109543651A CN201811487007.5A CN201811487007A CN109543651A CN 109543651 A CN109543651 A CN 109543651A CN 201811487007 A CN201811487007 A CN 201811487007A CN 109543651 A CN109543651 A CN 109543651A
- Authority
- CN
- China
- Prior art keywords
- driver
- steering wheel
- standard
- current
- head
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/59—Context or environment of the image inside of a vehicle, e.g. relating to seat occupancy, driver state or inner lighting conditions
- G06V20/597—Recognising the driver's state or behaviour, e.g. attention or drowsiness
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
- A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
- A61B5/1116—Determining posture transitions
-
- 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
- B60W40/09—Driving style or behaviour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B2503/00—Evaluating a particular growth phase or type of persons or animals
- A61B2503/20—Workers
- A61B2503/22—Motor vehicles operators, e.g. drivers, pilots, captains
Abstract
The invention discloses a kind of driver's dangerous driving behavior detection methods, can obtain the various parameters of steering wheel, realize the steering wheel parameter adaptive detection of various vehicles, input without any manually parameter.After modeling by the steering wheel for vehicle data of depth data acquisition and to steering wheel, may not need any sensor and data-interface can be realized and detect to driving behavior, greatly facilitate the detection of driving behavior；It can quickly start to detect the driving behavior of driver, and testing conditions are not influenced by ambient light, largely improve the feasibility of driver's hazardous act detection of view-based access control model.
Description
Technical field
The present invention relates to behavioral value methods, and in particular to a kind of driver's dangerous driving behavior detection method.
Background technique
At home and abroad compared in the research in forward position, detects and divide mainly for the fatigue state and driving behavior of driver
Analysis, the behavior for analyzing driver by analyzing the state of mind and the behavior of driver are most intuitive effective methods, this is to work as
Preceding main research direction.Research is broadly divided into: driving behavior detection, the operation based on vehicle based on physiological driver's state
The driving behavior of state detects and detects by computer vision technique the behavior of driver.
Currently, more commonly used physiological status has: EEG (EEG signals), EGG (electrocardiosignal), EMG (muscle telecommunications
Number), but to measure these parameters, need to dress special measuring instrument by driver, on the one hand may will affect and drive
The driver behavior for the person of sailing, to influence the accuracy of measurement result, on the other hand this kind of equipment manufacturing cost is relatively high, increase equipment at
This, it is not applicable to use in practice；The driving behavior of operating status based on vehicle detects this method mainly by vehicle
The upper parameter installed various sensors or obtain the vehicle in driving process using CAN bus and OBD interface, according to what is obtained
Parameter judges whether driver violation driving behavior occurs, and this method needs install a large amount of sensor on vehicle,
It influences driver's normal driving and system is excessively complicated, it is possible to vehicle is transformed, cost is relatively high, and due to driving
There are behavioral differences for the driving behavior of member, it is not easy to unified Modeling, so that there are errors for testing result.
Summary of the invention
The purpose of the present invention is to provide a kind of driver's dangerous driving behavior detection methods, exist to solve the prior art
The normally travel that driver is affected when detecting to driver's driving behavior causes accuracy rate of testing result is not high to ask
Topic.
In order to realize above-mentioned task, the invention adopts the following technical scheme:
A kind of driver's dangerous driving behavior detection method, the method include:
When the two sides that step 1, driver's both hands hold steering wheel keep standard posture, driven using the acquisition of Kinect camera
The bone posture of the person of sailing at this time, establishes standard attitude mode；
Described establishes standard attitude mode, comprising:
Step 11 establishes the steering wheel standard enveloping space shown in formula I, the steering wheel standard include space X-axis,
The unit of coordinate in Y-axis and Z axis is unit length:
Wherein, (Xlefthand,Ylefthand,Zlefthand) it is left hand bone three-dimensional coordinate of the driver when keeping standard posture,
(Xrighthand,Yrighthand,Zrighthand) it is right hand bone three-dimensional coordinate of the driver when keeping standard posture, SWxminFor direction
Coordinate minimum value on disk standard enveloping space X axis, SWxmaxFor coordinate maximum value on steering wheel standard enveloping space X axis,
SWyminFor coordinate minimum value in steering wheel standard enveloping space Y-axis, SWymaxTo be sat in steering wheel standard enveloping space Y-axis
Mark maximum value, SWzminFor the upward coordinate minimum value of steering wheel standard enveloping space Z axis, SWzmaxFor steering wheel standard enveloping space Z
Coordinate maximum value in axial direction, LfxpFor projector distance of the length in the steering wheel standard enveloping space on X axis of steering wheel, Hfxp
For the high projector distance in the steering wheel standard enveloping space in Y-axis of steering wheel, WfxpFor steering wheel width in steering wheel
The upward projector distance of Z axis in the standard enveloping space；
Step 12 calculates steering wheel standard rotation frequency f using Formula Vfxp:
Wherein, RfxpFor projector distance of the radius in the steering wheel standard enveloping space on X axis of steering wheel,Unit is unit length, SROIt is racing to detection sensitivity, SRO>0,TfxpIt is needed for steering wheel normal direction of rotation half-turn
The time wanted, unit s；
Step 13 calculates driver's shoulder standard vector V using Formula IVJB:
Wherein, (Xleftshoulder,Yleftshoulder,Zleftshoulder) it is left shoulder joint of the driver when keeping standard posture
Save three-dimensional coordinate, (Xrightshoulder,Yrightshoulder,Zrightshoulder) it is right shoulder joint of the driver when keeping standard posture
Save three-dimensional coordinate；
Step 14 calculates driver's incidence standard vector V using Formula VIINH:
VNH=(Xhead-Xneck,Yhead-Yneck,Zhead-Zneck) Formula VII
Wherein, (Xhead,Yhead,Zhead) it is head three-dimensional coordinate of the driver when keeping standard posture, (Xneck,Yneck,
Zneck) it is neck three-dimensional coordinate of the driver when keeping standard posture；
Step 2 repeats this step in driver's driving procedure, and whether detection driver has in driving procedure
Dangerous driving behavior:
It using the bone posture of Kinect camera acquisition driver, compares, examines with the standard attitude mode
Survey whether driver has dangerous driving behavior, comprising:
Step 21, according to the three-dimensional coordinate of driver's left hand bone and the three-dimensional coordinate of right hand bone whether in the envelope
Judge whether the hand of driver has left steering wheel in space, if the three-dimensional coordinate of left hand bone and the three-dimensional coordinate of right hand bone are equal
In the enveloping space, then the hand of driver does not leave steering wheel；Otherwise the hand of driver has left steering wheel；
Whether step 22 is higher than whether steering wheel standard rotation frequency judges driver according to current steering wheel rotational frequency
There is the dangerous driving behavior of zig zag, if current steering wheel rotational frequency is higher than steering wheel standard rotation frequency, driver has
The dangerous driving behavior of zig zag；Otherwise, the dangerous driving behavior that driver does not take a sudden turn；
Step 23 judges to drive according to the angle of the real-time vector of driver's current time shoulder and the shoulder standard vector
Whether member currently has incorrect sitting posture, if the angle of the real-time vector of current time shoulder and the shoulder standard vector is greater than shoulder
Angle threshold value, then the current abnormal sitting posture of driver is true, and otherwise the current sitting posture of driver is correct；
Step 24 judges according to the angle of the real-time vector of driver's current time incidence and the incidence standard vector
Whether driver head has abnormal shaking, if the angle of the real-time vector of current time incidence and the incidence standard vector
Greater than neck threshold value, then driver's current head has abnormal shaking；Otherwise driver's current head does not occur abnormal shaking.
Further, current T moment steering wheel rotational frequency f in the step 22T fxpIt is obtained using Formula X:
Wherein, tTFor the time at current T moment, tT-1For the time at T-1 moment,For T-1 moment hand bone vector,For left hand bone T-1 moment X axis coordinate,For right hand bone T-1 moment X
Axial coordinate；For current T moment hand bone vector,For left hand bone T moment X axis
Coordinate,For right hand bone T moment X axis coordinate.
Further, the current real-time vector of T moment shoulder of driver and the shoulder standard vector in the step 23
Angle theta1 TIt is obtained using Formula XII:
Wherein, VJSFor the real-time vector of shoulder at current T moment, VJS=(XT rightshoulder-XT leftshoulder,
YT rightshoulder-YT leftshoulder,ZT rightshoulder-ZT leftshoulder), (XT leftshoulder,YT leftshoulder,
ZT leftshoulder) be current T moment left shoulder joint three-dimensional coordinate, (XT rightshoulder,YT rightshoulder,ZT rightshoulder)
For the three-dimensional coordinate of current T moment right shoulder joint.
Further, the current real-time vector of T moment incidence of driver described in the step 24 and the neck
The angle theta of ministerial standard vector2 TIt is obtained using Formula XIV:
Wherein, VNHSFor the real-time vector of incidence at current T moment, VNHS=(XT head-XT neck,YT head-YT neck,ZT head-
ZT neck), (XT head,YT head,ZT head) be current T moment head skeleton three-dimensional coordinate, (XT neck,YT neck,ZT neck) it is current T
The three-dimensional coordinate of moment neck bone.
The present invention has following technical characterstic compared with prior art:
1, driver's dangerous driving behavior detection method provided by the invention drives attitude mode, judgement by establishing standard
Posture in driver's traveling process and standard drive the deviation between attitude mode, judge whether driver danger occurs and drive
The behavior sailed, detection process, to will not influence the normally travel of driver, improve detection without intervening driver
As a result accuracy rate；
2, driver's dangerous driving behavior detection method provided by the invention can obtain the various parameters of steering wheel, realize
The steering wheel parameter adaptive of various vehicles detects, and inputs without any manually parameter.The vehicle obtained by depth data
Steering wheel data and to steering wheel modeling after, may not need any sensor and data-interface can be realized and carry out to driving behavior
Detection, greatly facilitates the detection of driving behavior.
3, driver's dangerous driving behavior detection method provided by the invention can quickly start the driving row to driver
To be detected, and testing conditions are not influenced by ambient light, largely improve driver's danger row of view-based access control model
For the feasibility of detection.
Detailed description of the invention
Fig. 1 is the flow diagram of driver's dangerous driving behavior detection method provided by the invention；
Fig. 2 is the steering wheel enveloping space schematic diagram provided in one embodiment of the present of invention；
Fig. 3 is that the steering wheel provided in one embodiment of the present of invention is projected in X-axis and Z-direction schematic diagram.
Specific embodiment
The standard posture of driver refers to posture when driver's two-handed hand-held steering wheel, when driver keeps standard posture
When, angle, center back line between the distance between driver's both hands, driver's upper arm and postbrachium with perpendicular to horizontal plane side
To angle and incidence center line be held in certain range with perpendicular to the angle in horizontal plane direction, because of different vehicles
Driver's cabin difference, steering wheel angle and position are also different, and the height figure of driver is all different, therefore are driven by definition
Member's standard vehicle driving posture range, when meeting this range, the driver's bone posture of system record at this time is as standard posture.
For example, illustrating to drive when the distance between driver's both hands are maintained between 0.5 times to 1.5 times of shoulder distance
Member's both hands distance complies with standard posture；When angle between driver's left upper arm and left back arm is between 90 degree to 170 degree, drive
Member right upper arm and right postbrachium between angle between 90 degree to 170 degree when, illustrate that driver's both arms posture complies with standard posture；
When driver sits down center back line and perpendicular to horizontal plane direction angle between 0 degree to 20 degree when, illustrate driver sit
Appearance complies with standard posture；When driver sits down incidence center line with perpendicular to horizontal plane direction angle 0 degree to 10 degree it
Between when, illustrate that driver head's posture complies with standard posture.
The steering wheel enveloping space: the space being looped around on the outside of steering wheel is surrounded, which can accommodate completely steering wheel.
Kinect camera: 3D body-sensing camera can capture the athletic posture of human body in three-dimensional space, take the photograph in Kinect
Picture head has bone tracking technique, can capture the three dimensional local information of 20 artis of user's body, that is, obtain the bone of human body
Bone posture.
It is the specific embodiment that inventor provides below, technical solution of the present invention to be described further.
Embodiment one
As shown in Figure 1, the invention discloses a kind of driver's dangerous driving behavior detection method, the method includes:
When step 1, driver's both hands hold the two sides of steering wheel and keep standard posture, acquired using Kinect camera
The bone posture of driver at this time, establishes standard attitude mode；
The invention proposes it is a kind of using Kinect depth camera extract driver posture and analyze driver whether there is or not
The method of dangerous driving behavior.This method is big without installing on vehicle without allowing driver to dress various measurement sensors
Quantity sensor really realizes contactless driving behavior detection, and any interference will not be generated to the normal driving of driver, and
And it can accomplish to detect driving behavior under the low light situations such as daytime and night.
There are three camera lenses for the Kinect somatosensory camera of Microsoft Corporation, and intermediate camera lens is RGB color video camera,
For acquiring color image.The right and left camera lens is then respectively the 3D that RF transmitter and infrared C MOS video camera are constituted
Structure light depth inductor is used to sampling depth data, i.e., distance of the object to camera in scene.
When this method starts, it is necessary first to initialize standard bone posture, standard attitude mode, Kinect camera,
OpenGL three-dimensional library and OpenGL rendering contexts.
In the present embodiment, when initializing to Kinect camera, NiTE library command NiTE: is used:
Initialize () is completed.
Parameter is initialized again, the included angle A including steering wheel plane and groundfxp, normal direction disk turns to half-turn need to
The time T wantedfxp, racing is to detection sensitivity SRO, identification sensitivity is SA。
In the present embodiment, A is initializedfxp=30 °, Tfxp=4s, SRO=1.0, SA=1.0.
It when starting that driver is allowed to carry out pose calibrating, is delayed 3 seconds first, in order to allow driver to there is the sufficient time to make
The vehicle driving posture of standard.After 3 seconds, the bone posture of driver is recorded, and bone posture information is saved into structure
In body BodyPoints, standard bone posture is obtained, comprising:
Step A, usertracking device object UserTracker is created, and is created using order UserTracker::create ()
Build user tracking device；
Step B, user's scene point cloud frame object UserTrackerFrameRef is created, and uses order
UserTracker::readFrame (&UserTrackerFrameRef) read point cloud frame；
Step C, using order UserTracker::startSkeletonTracking () Cong Dianyun frame in track skeleton；
Step D, obtained skeleton is decomposed using order UserTracker::getSkeleton (), obtains driver and works as
Preceding bone posture is denoted as standard bone posture.
In the present embodiment, driver is sitting on driver's seat while after meeting the following conditions, then illustrates that driver is current
Maintain standard posture:
Wherein, driver's left hand bone three-dimensional coordinate (Xlefthand,Ylefthand,Zlefthand), right hand bone three-dimensional coordinate
(Xrighthand,Yrighthand,Zrighthand)；The direction vector V of driver's left forearmlfa=(Xlefthand-Xleftelbow,Ylefthand-
Yleftelbow,Zlefthand-Zleftelbow), the direction vector V of right forearmrfa=(Xrighthand-Xrightelbow,Yrighthand-
Yrightelbow,Zrighthand-Zrightelbow)；The direction vector V of the left back arm of driverlua=(Xleftshoulder-Xleftelbow,
Yleftshoulder-Yleftelbow,Zleftshoulder-Zleftelbow), the direction vector V of right postbrachiumrua=(Xrightshoulder-
Xrightelbow,Yrightshoulder-Yrightelbow,Zrightshoulder-Zrightelbow) wherein, left elbow joint three-dimensional coordinate (Xleftelbow,
Yleftelbow,Zleftelbow), right elbow joint three-dimensional coordinate (Xrightelbow,Yrightelbow,Zrightelbow), left shoulder joint three-dimensional coordinate
(Xleftshoulder,Yleftshoulder,Zleftshoulder), right shoulder joint three-dimensional coordinate (Xrightshoulder,Yrightshoulder,
Zrightshoulder)；
Driver's sitting posture vector Vzz=(Xhc-Xneck,Yhc-Yneck,Zhc-Zneck), incidence vector Vhn=(Xhead-Xneck,
Yhead-Yneck,Zhead-Zneck)；
Wherein, head three-dimensional coordinate (Xhead,Yhead,Zhead), neck three-dimensional coordinate (Xneck,Yneck,Zneck), hip central point
Three-dimensional coordinate is (Xhc,Yhc,Zhc)。
In the present embodiment, the left shoulder coordinate of driver is (Xleftshoulder,Yleftshoulder,Zleftshoulder)=(-
5.0,0.0,0.0), right shoulder coordinate is (Xrighthoulder,Yrightshoulder,Zrightshoulder)=(5.0,0.0,0.0)；Left elbow is sat
It is designated as (Xleftelbow,Yleftelbow,Zleftelbow)=(- 5.0, -1.0, -2.5), right elbow coordinate is (Xrightelbow,Yrightelbow,
Zrightelbow)=(5.0, -1.0, -2.5)；Left hand bone coordinate is (Xlefthand,Ylefthand,Zlefthand)=(- 5.0, -1.0, -
5.0), right hand bone coordinate is (Xrighthand,Yrighthand,Zrighthand)=(5.0, -1.0, -5.0)；Head coordinate is (Xhead,
Yhead,Zhead)=(0.0,2.0,0.0), neck coordinate is (Xneck,Yneck,Zneck)=(0.0,1.0,0.0), hip central point three
Dimension coordinate is (Xhc,Yhc,Zhc)=(0.0, -4.0,0.0).
Because | Xleftshoulder-Xrighthoulder|=10.0, | Xlefthand-Xrighthand|=10.0, then meet range 0.5*
|Xleftshoulder-Xrighthoulder|<|Xlefthand-Xrighthand|<1.5*|Xleftshoulder-Xrighthoulder|, therefore, it is considered that driving
Member's both hands distance complies with standard vehicle driving posture.
Driver's left forearm direction vector Vlfa=(0.0,0.0, -2.5), the direction vector V of available right forearmrfa=
(0.0,0.0, -2.5), the direction vector V of the available left back arm of driverlua=(0.0,1.0,2.5), it is available it is right after
The direction vector V of armrua=(0.0,1.0,2.5), then the angle theta of left upper arm and left back arm can be calculatedla=158.2 °,
The angle theta of right upper arm and right postbrachium can similarly be calculatedra=158.2 °, meet 90 ° < θra=θra< 170 °, therefore can sentence
The both arms posture of disconnected driver complies with standard vehicle driving posture.
Pass through hip central point and the available driver's sitting posture vector V of neck three-dimensional coordinatezz=(0.0, -5.0,0.0), then
Driver's sitting posture vector and vehicle Y-axis angle may be calculated θzz=0, then meet 0 °≤θzz≤ 20 °, it is believed that driver
Sitting posture complies with standard vehicle driving posture.
Pass through the available incidence vector V of the three-dimensional coordinate of neck and headhn=(0.0,1.0,0.0), then incidence
Vector and vehicle Y-axis angle may be calculated θhn=0, then meet 0 °≤θhn≤ 10 ° it is considered that driver head's posture meets
Standard vehicle driving posture.
In conclusion driver's both hands distance, both arms posture, sitting posture, head pose comply with standard posture, system thinks
Driver has made standard posture, and the numerical value of the driver's bone posture of system record at this time establishes standard attitude mode.
In this step, it when driver's both hands hold the two sides of steering wheel and keep standard posture, is imaged using Kinect
The bone posture of head acquisition driver at this time includes: left hand bone three-dimensional coordinate (Xlefthand,Ylefthand,Zlefthand), right hand bone three
Tie up coordinate (Xrighthand,Yrighthand,Zrighthand), left shoulder joint X axis coordinate (Xleftshoulder,Yleftshoulder,
Zleftshoulder), right shoulder joint X axis coordinate (Xrightshoulder,Yrightshoulder,Zrightshoulder), head three-dimensional coordinate
(Xhead,Yhead,Zhead), neck three-dimensional coordinate (Xneck,Yneck,Zneck)。
Wherein, standard attitude mode is established, comprising:
In the present solution, in order to quickly detect whether hand bone coordinate exceeds steering wheel envelope domain range, it is assumed that steering wheel packet
Network space is the hexahedron parallel with solid axes, as shown in Figure 2.
Step 11 establishes the steering wheel standard enveloping space shown in formula I, the steering wheel standard include space X-axis,
The unit of coordinate in Y-axis and Z axis is unit length:
Wherein, (Xlefthand,Ylefthand,Zlefthand) it is left hand bone three-dimensional coordinate of the driver when keeping standard posture,
(Xrighthand,Yrighthand,Zrighthand) it is right hand bone three-dimensional coordinate of the driver when keeping standard posture, SWxminFor direction
Coordinate minimum value on disk standard enveloping space X axis, SWxmaxFor coordinate maximum value on steering wheel standard enveloping space X axis,
SWyminFor coordinate minimum value in steering wheel standard enveloping space Y-axis, SWymaxTo be sat in steering wheel standard enveloping space Y-axis
Mark maximum value, SWzminFor the upward coordinate minimum value of steering wheel standard enveloping space Z axis, SWzmaxFor steering wheel standard enveloping space Z
Coordinate maximum value in axial direction, LfxpFor projector distance of the length in the steering wheel standard enveloping space on X axis of steering wheel, Hfxp
For the high projector distance in the steering wheel standard enveloping space in Y-axis of steering wheel, WfxpFor steering wheel width in steering wheel
The upward projector distance of Z axis in the standard enveloping space；
In this step, the long L of steering wheel is calculated by Formula IIfxpProjector distance in the X-axis, unit are unit length:
Lfxp=| Xrighthand-Xlefthand|×SA
Formula II calculates the wide W of steering wheel by formula IIIfxpIn the upward projector distance of Z axis, unit is unit length:
Wfxp=Lfxp·cos(Afxp) formula III
The high H of steering wheel is calculated by formula IVfxpProjector distance in Y-axis, unit are unit length:
Hfxp=Lfxp·sin(Afxp) formula IV
In the present embodiment, as shown in figure 3, projection of the length of steering wheel in the steering wheel standard enveloping space on X axis away from
From Lfxp=2.0 unit lengths, projector distance W of the width of steering wheel in the steering wheel standard enveloping space in Y-axisfxp=
2.0 unit lengths, the height of steering wheel upward projector distance H of Z axis in the steering wheel standard enveloping spacefxp=1.0 units
Length, therefore the three-dimensional coordinate of the steering wheel standard enveloping space are as follows:
Step 12 calculates steering wheel standard rotation frequency f using Formula Vfxp:
Wherein, RfxpFor projector distance of the radius in the steering wheel standard enveloping space on X axis of steering wheel,SROIt is racing to detection sensitivity, SRO>0,TfxpFor the time that steering wheel normal direction of rotation half-turn needs, unit is
s；
In the present embodiment, projector distance R of the radius of steering wheel in steering wheel standard enveloping space X-axisfxp=2.0
A unit length, racing is to detection sensitivity SWRoSensiIt is 1.0, the time T that steering wheel normal direction of rotation half-turn needsfxp=4s,
Then f can be calculated in steering wheel standard rotation frequencyfxp=0.5.
Step 13 calculates driver's shoulder standard vector V using Formula IVJB:
Wherein, (Xleftshoulder,Yleftshoulder,Zleftshoulder) it is left shoulder joint of the driver when keeping standard posture
Save three-dimensional coordinate, (Xrightshoulder,Yrightshoulder,Zrightshoulder) it is right shoulder joint of the driver when keeping standard posture
Save three-dimensional coordinate；
In the present embodiment, Xrightshoulder=5.0, Xleftshoulder=-5.0, Yrightshoulder=Yleftshoulder=
0.0, Zrightshoulder=Zleftshoulder=0.0, then shoulder standard vector VJB=(10.0,0.0,0.0).
Step 14 calculates driver's incidence standard vector V using Formula VIINH:
VNH=(Xhead-Xneck,Yhead-Yneck,Zhead-Zneck) Formula VII
Wherein, (Xhead,Yhead,Zhead) it is head three-dimensional coordinate of the driver when keeping standard posture, (Xneck,Yneck,
Zneck) it is neck three-dimensional coordinate of the driver when keeping standard posture；
In this step, Xhead=0.0, Xneck=0.0, Yhead=5.0, Yneck=0.0, Zhead=0.0, Zneck=0.0,
So incidence standard vector VNH=(0.0,5.0,0.0).
Step 2 repeats this step in driver's driving procedure, and whether detection driver has in driving procedure
Dangerous driving behavior:
It using the bone posture of Kinect camera acquisition driver, compares, examines with the standard attitude mode
Survey whether driver has dangerous driving behavior, comprising:
Step 21, according to the three-dimensional coordinate of driver's left hand bone and the three-dimensional coordinate of right hand bone whether in the envelope
Judge whether the hand of driver has left steering wheel in space, if the three-dimensional coordinate of left hand bone and the three-dimensional coordinate of right hand bone are equal
In the enveloping space, then the hand of driver does not leave steering wheel；Otherwise the hand of driver has left steering wheel；
In the present embodiment, current T moment left hand bone three-dimensional coordinate is (XT lefthand,YT lefthand,ZT lefthand)=(-
15,0,0), current T moment right hand bone three-dimensional coordinate (XT righthand,YT righthand,ZT righthand)=(9,0,0), then with standard
The enveloping space compares, and driver has left steering wheel in T moment left hand, and the right hand does not leave steering wheel, and the current T+1 moment is left
Hand bone three-dimensional coordinate is (XT+1 lefthand,YT+1 lefthand,ZT+1 lefthand)=(0,5,0), current T+1 moment right hand bone three-dimensional is sat
It is designated as (XT+1 righthand,YT+1 righthand,ZT+1 righthand)=(0, -15,0), then it compares, drives with the standard enveloping space
Member does not have left steering wheel in T+1 moment left hand, and the right hand has left steering wheel.
Whether step 22 is higher than whether steering wheel standard rotation frequency judges driver according to current steering wheel rotational frequency
There is the dangerous driving behavior of zig zag, if current steering wheel rotational frequency is higher than steering wheel standard rotation frequency, driver has
The dangerous driving behavior of zig zag；Otherwise, the dangerous driving behavior that driver does not take a sudden turn；
Optionally, current T moment steering wheel rotational frequency f in the step 22T fxpIt is obtained using Formula X:
Wherein, tTFor the time at current T moment, tT-1For the time at T-1 moment,For T-1 moment hand bone vector,For left hand bone T-1 moment X axis coordinate,For right hand bone T-1 moment X
Axial coordinate；For current T moment hand bone vector,For left hand bone T moment X-axis
To coordinate,For right hand bone T moment X axis coordinate.
Therefore, the step 22 includes:
Step 221 generates T-1 moment hand bone vector using Formula VIIIAnd T moment hand bone vector
Wherein,For left hand bone T moment X axis coordinate,For right hand bone T moment X axis coordinate,
For left hand bone T-1 moment X axis coordinate,For right hand bone T-1 moment X axis coordinate；
In the present embodiment,ThenNext moment right-hand man bone
Coordinate are as follows:Then
Step 222 judges whether driver has turning action at the current T moment using Formula IX:
If Formula IX is set up, illustrates that driver has turning action, execute step 223；Otherwise illustrate that driver does not turn
Movement executes step 23；
In the present embodiment, it is calculated in previous stepRfxp=2.0, racing
To detection sensitivity SWRoSensiIt is 1.0, then Then above formula is set up, therefore can be with
Detect that driver has turning action at current time.
Step 223 judges whether driver has the dangerous driving behavior of zig zag the current T moment using Formula X:
Wherein, tTFor the time at T moment, tT-1For the time at T-1 moment；
If Formula X is set up, illustrate that driver has the dangerous driving behavior of zig zag；Otherwise illustrate that driver does not take a sudden turn
Dangerous driving behavior.
In the present embodiment, it is obtained in above steptT-1=1s, tT=1.5s, thenGreater than f presetting in step 12fxp=0.5, therefore may determine that there is racing in T moment driver
Curved movement.
Step 23 judges to drive according to the angle of the real-time vector of driver's current time shoulder and the shoulder standard vector
Whether member currently has incorrect sitting posture, if the angle of the real-time vector of current time shoulder and the shoulder standard vector is greater than shoulder
Angle threshold value, then the current abnormal sitting posture of driver is true, and otherwise the current sitting posture of driver is correct；
Optionally, the current real-time vector of T moment shoulder of driver and the shoulder standard vector in the step 23
Angle theta1 TIt is obtained using Formula XII:
Wherein, VJSFor the real-time vector of shoulder at current T moment, VJS=(XT rightshoulder-XT leftshoulder,
YT rightshoulder-YT leftshoulder,ZT rightshoulder-ZT leftshoulder), (XT leftshoulder,YT leftshoulder,
ZT leftshoulder) be current T moment left shoulder joint three-dimensional coordinate, (XT rightshoulder,YT rightshoulder,ZT rightshoulder)
For the three-dimensional coordinate of current T moment right shoulder joint.
Therefore, the step 23 includes:
Step 231, the real-time vector V of shoulder that the current T moment is calculated using Formula XIJS:
VJS=(XT rightshoulder-XT leftshoulder,YT rightshoulder-YT leftshoulder,ZT rightshoulder-
ZT leftshoulder) Formula XI
Wherein, (XT leftshoulder,YT leftshoulder,ZT leftshoulder) be current T moment left shoulder joint three-dimensional coordinate,
(XT rightshoulder,YT rightshoulder,ZT rightshoulder) be current T moment right shoulder joint three-dimensional coordinate；
In the present embodiment, left shoulder coordinate is (XT leftshoulder,YT leftshoulder,ZT leftshoulder)=(- 5,0, -2),
Right shoulder coordinate is (XT rightshoulder,YT rightshoulder,ZT rightshoulder)=(5,0,2), then it can calculate outgoing vector VJS=
(10,0,4)。
Step 232, the shoulder angle theta that the current T moment is calculated using Formula XII1 T:
In the present embodiment, shoulder standard vector V has been obtained in step 13JB=(10,0,0).In previous step
In also obtained real-time vector VJS=(10,0,4), then available T moment shoulder angle theta1 T=21.8 °.
Step 233 judges current T moment shoulder angle theta1 TWhether shoulder angle threshold value is greater than, if current T moment shoulder folder
Angle θ1 TGreater than shoulder angle threshold value, then illustrate that the current abnormal sitting posture of driver is true；Otherwise, the current sitting posture of driver is correct.
In the present embodiment, defining shoulder angle threshold value is 20 °, and the current T moment of driver is judged in previous step
It is 21.8 ° that shoulder, which rotates angle, then may determine that the current T moment abnormal sitting posture of driver is true at this time.
Step 24 judges according to the angle of the real-time vector of driver's current time incidence and the incidence standard vector
Whether driver head has abnormal shaking, if the angle of the real-time vector of current time incidence and the incidence standard vector
Greater than neck threshold value, then driver's current head has abnormal shaking；Otherwise driver's current head does not occur abnormal shaking.
Optionally, the current real-time vector of T moment incidence of driver described in the step 24 and the incidence
The angle theta of standard vector2 TIt is obtained using Formula XIV:
Wherein, VNHSFor the real-time vector of incidence at current T moment, VNHS=(XT head-XT neck,YT head-YT neck,ZT head-
ZT neck), (XT head,YT head,ZT head) be current T moment head skeleton three-dimensional coordinate, (XT neck,YT neck,ZT neck) it is current T
The three-dimensional coordinate of moment neck bone.
Therefore, the step 24 includes:
Step 241 calculates the real-time vector V of T current time, incidence moment using Formula XIIIJS:
VNHS=(XT head-XT neck,YT head-YT neck,ZT head-ZT neck) Formula XIII
Wherein, (XT head,YT head,ZT head) be current T moment head skeleton three-dimensional coordinate, (XT neck,YT neck,ZT neck)
For the three-dimensional coordinate of current T moment neck bone；
In the present embodiment, in the three-dimensional coordinate (X of current T moment head skeletonT head,YT head,ZT head)=(- 5,5,
0), the three-dimensional coordinate (X of current T moment neck boneT neck,YT neck,ZT neck)=(0,0,0), then current T moment incidence is real
Shi Xiangliang VNHS=(- 5,5,0).
Step 242 calculates current T moment incidence angle theta using Formula XIV2 T:
In the present embodiment, because incidence standard vector V has been obtained at step 14NH=(0.0,5.0,0.0),
The real-time vector V of current T moment incidence has also been obtained in the previous stepNHS=(- 5.0,5.0,0.0), then can calculate and lift one's head
Neck angle theta2 T=30.0 °.
Step 243 judges current T moment incidence angle theta2 TWhether incidence angle threshold value is greater than, if current T moment head
Neck angle theta2 TGreater than incidence angle threshold value, then illustrate that driver's current head has abnormal shaking；Otherwise, driver is current
Head does not occur abnormal shaking.
In the present embodiment, incidence angle threshold value is 20 °, and the current T moment head of driver is judged in previous step
It is 30 ° that neck, which shakes angle, then may determine that driver's current head has abnormal shaking at this time.
Embodiment two
Verification experimental verification is carried out to driver's dangerous driving behavior detection method provided by the invention, subjects are bus
Driver specifically includes:
Step I, experiment scene is disposed, Kinect device is installed in bus driver's cabin；
Step II, driver gestures are calibrated, and are allowed driver to make the position of driver of standard and are kept for 3 seconds, system at this time
Driver's bone posture under standard gestures is recorded, and this posture is kept in structure B odyPoints.
Step III, the whether dangerous driving behavior of driver is detected, comprising:
(1) whether detection driver's hand leaves steering wheel
By being continuously increased, driver's hand leaves the number of steering wheel and whether frequency tests the method for the present invention can be quasi- in real time
Really detection.
1 driver's hand of table leaves direction looping test
| Frequency (5 minutes) | Number | Identify number | Discrimination |
| 25 | 50 | 50 | 100% |
| 50 | 50 | 50 | 100% |
| 100 | 50 | 50 | 100% |
| 300 | 50 | 48 | 96% |
| 600 | 50 | 46 | 92% |
Frequency indicates that driver's hand leaves the frequency of steering wheel in table 1, leaves steering wheel with 5 minutes for benchmark time hand
Frequency is gradually increasing, and when frequency is 25, is indicated that the every 12 seconds hands of driver leave a steering wheel, when frequency is 600, is indicated
The every 0.5 second hand of driver leaves a steering wheel.Number indicates to leave the testing time of steering wheel in current frequency assistant.Identification
Number indicates that the number that this method leaves in all testing times of steering wheel each hand, correctly identifies, discrimination are correct
Rate.
From table 1 it follows that is, 1 second driver is with regard to the handle side of leaving when the frequency that hand leaves steering wheel reaches 300 times
When to disk, this method still can achieve 96% discrimination；When the frequency that hand leaves steering wheel reaches 600 times, i.e., drive within 0.5 second
When the person of sailing leaves steering wheel with regard to handle, method provided by the invention can achieve 92% discrimination.
(2) whether detection driver has zig zag dangerous driving behavior
It since driver's zig zag is breakneck behavior, is not tested when vehicle travels, and allows driving
Member makes the movement of zig zag on the steering wheel.
Zig zag test result is shown in Table 2, in this experiment, still growth of the increase of usage frequency as cutting actions
Unit, the frequency of cutting actions is by occurring 20 times to increase to be 5 minutes 600 times in 5 minutes.When frequency is 20, driver's difference
Left racing and right cutting actions 20 times are made, method provided by the invention successfully has identified the movement of these zig zags.When frequency
When degree rises to 600, driver makes left racing and right cutting actions 40 times respectively, and method provided by the invention is correct respectively to be known
37 times and 38 times are not gone out, discrimination has reached 93%.
The zig zag test result of table 2
(3) whether detection driver has incorrect sitting posture
In this test, driver's left-right rotation body, double swerve body are detected using method provided by the invention
Whether driver has incorrect sitting posture.
Test result is shown in Table 3, and experiment frequency still rises to 5 minutes 600 times from 5 minutes 20 times.When frequency is 20, always
Body discrimination can achieve 100%, when Pingdu rises to 100, have a small number of movements are unrecognized to come out, when frequency is 600,
Overall discrimination still can achieve 96%.
The incorrect sitting posture testing result of 3 driver of table
(4) whether detection driver is inwholwe-hearted drives
In this test, applicant allows shaking head around driver to say with drive simulating person with people around
Words or other behaviors not driven attentively.
Test result is shown in Table 4, and experiment frequency still rises to 5 minutes 600 times from 5 minutes 20 times.When frequency is 20, always
Body discrimination can achieve 100%, when frequency is 100, has a small number of movements and fails to be identified, when frequency rises to 600
When, overall discrimination still can achieve 96%.
4 driver of table does not drive testing result attentively
Method provided by the invention efficiently and quickly can detect and identify common bus dangerous driving behavior,
And it is considerably less using the workload that this method carries out early-stage preparations, it can quickly start to examine the driving behavior of driver
It surveys, and testing conditions are not influenced by ambient light, largely improves driver's hazardous act detection of view-based access control model
Feasibility.
Claims (4)
1. a kind of driver's dangerous driving behavior detection method, which is characterized in that the method includes:
When the two sides that step 1, driver's both hands hold steering wheel keep standard posture, driver is acquired using Kinect camera
Bone posture at this time, establishes standard attitude mode；
Described establishes standard attitude mode, comprising:
Step 11 establishes the steering wheel standard enveloping space shown in formula I, and the steering wheel standard includes space in X-axis, Y-axis
And the unit of the coordinate on Z axis is unit length:
Wherein, (Xlefthand,Ylefthand,Zlefthand) it is left hand bone three-dimensional coordinate of the driver when keeping standard posture,
(Xrighthand,Yrighthand,Zrighthand) it is right hand bone three-dimensional coordinate of the driver when keeping standard posture, SWxminFor direction
Coordinate minimum value on disk standard enveloping space X axis, SWxmaxFor coordinate maximum value on steering wheel standard enveloping space X axis,
SWyminFor coordinate minimum value in steering wheel standard enveloping space Y-axis, SWymaxTo be sat in steering wheel standard enveloping space Y-axis
Mark maximum value, SWzminFor the upward coordinate minimum value of steering wheel standard enveloping space Z axis, SWzmaxFor steering wheel standard enveloping space Z
Coordinate maximum value in axial direction, LfxpFor projector distance of the length in the steering wheel standard enveloping space on X axis of steering wheel, Hfxp
For the high projector distance in the steering wheel standard enveloping space in Y-axis of steering wheel, WfxpFor steering wheel width in steering wheel
The upward projector distance of Z axis in the standard enveloping space；
Step 12 calculates steering wheel standard rotation frequency f using Formula Vfxp:
Wherein, RfxpFor projector distance of the radius in the steering wheel standard enveloping space on X axis of steering wheel,It is single
Position is unit length, SROIt is racing to detection sensitivity, SRO>0,TfxpFor the time that steering wheel normal direction of rotation half-turn needs, list
Position is s；
Step 13 calculates driver's shoulder standard vector V using Formula IVJB:
Wherein, (Xleftshoulder,Yleftshoulder,Zleftshoulder) it is left shoulder joint three of the driver when keeping standard posture
Tie up coordinate, (Xrightshoulder,Yrightshoulder,Zrightshoulder) it is right shoulder joint three of the driver when keeping standard posture
Tie up coordinate；
Step 14 calculates driver's incidence standard vector V using Formula VIINH:
VNH=(Xhead-Xneck,Yhead-Yneck,Zhead-Zneck) Formula VII
Wherein, (Xhead,Yhead,Zhead) it is head three-dimensional coordinate of the driver when keeping standard posture, (Xneck,Yneck,
Zneck) it is neck three-dimensional coordinate of the driver when keeping standard posture；
Step 2 repeats this step in driver's driving procedure, detects whether driver has danger in driving procedure
Driving behavior:
Using the bone posture of Kinect camera acquisition driver, compared with the standard attitude mode, detection is driven
Whether the person of sailing has dangerous driving behavior, comprising:
Step 21, according to the three-dimensional coordinate of driver's left hand bone and the three-dimensional coordinate of right hand bone whether in the enveloping space
Whether the interior hand for judging driver has left steering wheel, if the three-dimensional coordinate of the three-dimensional coordinate of left hand bone and right hand bone is in institute
In the enveloping space stated, then the hand of driver does not leave steering wheel；Otherwise the hand of driver has left steering wheel；
Whether step 22 is higher than steering wheel standard rotation frequency according to current steering wheel rotational frequency and judges whether driver has urgency
The dangerous driving behavior of turning, if current steering wheel rotational frequency is higher than steering wheel standard rotation frequency, driver has racing
Curved dangerous driving behavior；Otherwise, the dangerous driving behavior that driver does not take a sudden turn；
Step 23 judges that driver works as according to the real-time vector of driver's current time shoulder and the angle of the shoulder standard vector
It is preceding whether to have incorrect sitting posture, if the angle of the real-time vector of current time shoulder and the shoulder standard vector is greater than shoulder angle
Threshold value, then the current abnormal sitting posture of driver is true, and otherwise the current sitting posture of driver is correct；
Step 24 judges to drive according to the angle of the real-time vector of driver's current time incidence and the incidence standard vector
Whether member head has abnormal shaking, if the real-time vector of current time incidence and the angle of the incidence standard vector are greater than
Neck threshold value, then driver's current head has abnormal shaking；Otherwise driver's current head does not occur abnormal shaking.
2. driver's dangerous driving behavior detection method as described in claim 1, which is characterized in that in the step 22 when
Preceding T moment steering wheel rotational frequency fT fxpIt is obtained using Formula X:
Wherein, tTFor the time at current T moment, tT-1For the time at T-1 moment,For T-1 moment hand bone vector, For left hand bone T-1 moment X axis coordinate,For right hand bone T-1 moment X-axis
To coordinate；For current T moment hand bone vector, For left hand bone T moment X axis seat
Mark,For right hand bone T moment X axis coordinate.
3. driver's dangerous driving behavior detection method as described in claim 1, which is characterized in that driven in the step 23
The angle theta of the person of the sailing current real-time vector of T moment shoulder and the shoulder standard vector1 TIt is obtained using Formula XII:
Wherein, VJSFor the real-time vector of shoulder at current T moment, VJS=(XT rightshoulder-XT leftshoulder,YT rightshoulder-
YT leftshoulder,ZT rightshoulder-ZT leftshoulder), (XT leftshoulder,YT leftshoulder,ZT leftshoulder) be current T when
Carve the three-dimensional coordinate of left shoulder joint, (XT rightshoulder,YT rightshoulder,ZT rightshoulder) it is current T moment right shoulder joint
Three-dimensional coordinate.
4. driver's dangerous driving behavior detection method as described in claim 1, which is characterized in that institute in the step 24
The angle theta of the driver stated the current real-time vector of T moment incidence and the incidence standard vector2 TIt is obtained using Formula XIV:
Wherein, VNHSFor the real-time vector of incidence at current T moment, VNHS=(XT head-XT neck,YT head-YT neck,ZT head-
ZT neck), (XT head,YT head,ZT head) be current T moment head skeleton three-dimensional coordinate, (XT neck,YT neck,ZT neck) it is current T
The three-dimensional coordinate of moment neck bone.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811487007.5A CN109543651B (en) | 2018-12-06 | 2018-12-06 | Method for detecting dangerous driving behavior of driver |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811487007.5A CN109543651B (en) | 2018-12-06 | 2018-12-06 | Method for detecting dangerous driving behavior of driver |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109543651A true CN109543651A (en) | 2019-03-29 |
| CN109543651B CN109543651B (en) | 2022-11-22 |
Family
ID=65852982
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811487007.5A Active CN109543651B (en) | 2018-12-06 | 2018-12-06 | Method for detecting dangerous driving behavior of driver |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109543651B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111368696A (en) * | 2020-02-28 | 2020-07-03 | 淮阴工学院 | Dangerous chemical transport vehicle illegal driving behavior detection method and system based on visual cooperation |
| CN111449652A (en) * | 2020-05-06 | 2020-07-28 | 北方工业大学 | Construction safety monitoring method and device based on brain wave analysis |
| CN112287795A (en) * | 2020-10-22 | 2021-01-29 | 北京百度网讯科技有限公司 | Abnormal driving posture detection method, device, equipment, vehicle and medium |
| CN112339764A (en) * | 2020-11-04 | 2021-02-09 | 杨华勇 | New energy automobile driving posture analysis system based on big data |
| CN115471826A (en) * | 2022-08-23 | 2022-12-13 | 中国航空油料集团有限公司 | Method and device for judging safe driving behavior of aircraft refueling truck and safe operation and maintenance system |
| CN116965781A (en) * | 2023-04-28 | 2023-10-31 | 南京晓庄学院 | Method and system for monitoring vital signs and driving behaviors of driver |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105551182A (en) * | 2015-11-26 | 2016-05-04 | 吉林大学 | Driving state monitoring system based on Kinect human body posture recognition |
| JP2016200910A (en) * | 2015-04-08 | 2016-12-01 | 日野自動車株式会社 | Driver state determination device |
| CN108446678A (en) * | 2018-05-07 | 2018-08-24 | 同济大学 | A kind of dangerous driving behavior recognition methods based on skeleton character |
| US10081366B1 (en) * | 2015-05-04 | 2018-09-25 | Carnegie Mellon University | Sensor-based assessment of attention interruptibility |
-
2018
- 2018-12-06 CN CN201811487007.5A patent/CN109543651B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2016200910A (en) * | 2015-04-08 | 2016-12-01 | 日野自動車株式会社 | Driver state determination device |
| US10081366B1 (en) * | 2015-05-04 | 2018-09-25 | Carnegie Mellon University | Sensor-based assessment of attention interruptibility |
| CN105551182A (en) * | 2015-11-26 | 2016-05-04 | 吉林大学 | Driving state monitoring system based on Kinect human body posture recognition |
| CN108446678A (en) * | 2018-05-07 | 2018-08-24 | 同济大学 | A kind of dangerous driving behavior recognition methods based on skeleton character |
Non-Patent Citations (2)
| Title |
|---|
| 刘通等: "车辆典型危险行驶状态识别与检测研究进展", 《中国安全科学学报》 * |
| 肖献强等: "基于驾驶行为的汽车主动安全技术研究", 《中国机械工程》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111368696A (en) * | 2020-02-28 | 2020-07-03 | 淮阴工学院 | Dangerous chemical transport vehicle illegal driving behavior detection method and system based on visual cooperation |
| CN111449652A (en) * | 2020-05-06 | 2020-07-28 | 北方工业大学 | Construction safety monitoring method and device based on brain wave analysis |
| CN111449652B (en) * | 2020-05-06 | 2022-11-29 | 北方工业大学 | Construction safety monitoring method and device based on brain wave analysis |
| CN112287795A (en) * | 2020-10-22 | 2021-01-29 | 北京百度网讯科技有限公司 | Abnormal driving posture detection method, device, equipment, vehicle and medium |
| CN112287795B (en) * | 2020-10-22 | 2023-09-01 | 北京百度网讯科技有限公司 | Abnormal driving gesture detection method, device, equipment, vehicle and medium |
| CN112339764A (en) * | 2020-11-04 | 2021-02-09 | 杨华勇 | New energy automobile driving posture analysis system based on big data |
| CN115471826A (en) * | 2022-08-23 | 2022-12-13 | 中国航空油料集团有限公司 | Method and device for judging safe driving behavior of aircraft refueling truck and safe operation and maintenance system |
| CN115471826B (en) * | 2022-08-23 | 2024-03-26 | 中国航空油料集团有限公司 | Method and device for judging safe driving behavior of aviation fueller and safe operation and maintenance system |
| CN116965781A (en) * | 2023-04-28 | 2023-10-31 | 南京晓庄学院 | Method and system for monitoring vital signs and driving behaviors of driver |
| CN116965781B (en) * | 2023-04-28 | 2024-01-05 | 南京晓庄学院 | Method and system for monitoring vital signs and driving behaviors of driver |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109543651B (en) | 2022-11-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109543651A (en) | A kind of driver's dangerous driving behavior detection method | |
| CN104200192B (en) | Driver's gaze detection system | |
| EP1320830B1 (en) | Facial image processing system | |
| CN110913751B (en) | Wearable eye tracking system with slip detection and correction functions | |
| US8077217B2 (en) | Eyeball parameter estimating device and method | |
| García et al. | Driver monitoring based on low-cost 3-D sensors | |
| CN105847684A (en) | Unmanned aerial vehicle | |
| CN106625673A (en) | Narrow space assembly system and assembly method | |
| Selim et al. | AutoPOSE: Large-scale Automotive Driver Head Pose and Gaze Dataset with Deep Head Orientation Baseline. | |
| CN105868690A (en) | Method and apparatus for identifying mobile phone use behavior of driver | |
| Duan et al. | Robust cooperative target detection for a vision-based UAVs autonomous aerial refueling platform via the contrast sensitivity mechanism of eagle's eye | |
| Jha et al. | Probabilistic estimation of the driver's gaze from head orientation and position | |
| CN109934207A (en) | A kind of characteristic distance modification method of driver face based on facial expression fatigue driving detection algorithm | |
| Wang et al. | Automobile driver posture monitoring systems: A review | |
| CN111854620A (en) | Monocular camera-based actual pupil distance measuring method, device and equipment | |
| WO2017173480A1 (en) | Method and system of distinguishing between a glance event and an eye closure event | |
| Zhang et al. | Driver fatigue state detection based on facial key points | |
| WO2020237940A1 (en) | Fatigue detection method and device based on human eye state identification | |
| CN112766097B (en) | Sight line recognition model training method, sight line recognition device and sight line recognition equipment | |
| CN114356072A (en) | System and method for detecting spatial orientation of wearable device | |
| Liu et al. | Motion sickness modeling with visual vertical estimation and its application to autonomous personal mobility vehicles | |
| CN109409173A (en) | Driver's state monitoring method, system, medium and equipment based on deep learning | |
| Ribas et al. | In-Cabin vehicle synthetic data to test Deep Learning based human pose estimation models | |
| JP3686418B2 (en) | Measuring device and method | |
| Liu et al. | Gaze tracking technology for flight simulator |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |