CN106627529A - Intelligent monitoring method and device for vehicle - Google Patents
Intelligent monitoring method and device for vehicle Download PDFInfo
- Publication number
- CN106627529A CN106627529A CN201611110035.6A CN201611110035A CN106627529A CN 106627529 A CN106627529 A CN 106627529A CN 201611110035 A CN201611110035 A CN 201611110035A CN 106627529 A CN106627529 A CN 106627529A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- human
- human pilot
- behavior
- human eye
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60T—VEHICLE BRAKE CONTROL SYSTEMS OR PARTS THEREOF; BRAKE CONTROL SYSTEMS OR PARTS THEREOF, IN GENERAL; ARRANGEMENT OF BRAKING ELEMENTS ON VEHICLES IN GENERAL; PORTABLE DEVICES FOR PREVENTING UNWANTED MOVEMENT OF VEHICLES; VEHICLE MODIFICATIONS TO FACILITATE COOLING OF BRAKES
- B60T7/00—Brake-action initiating means
- B60T7/12—Brake-action initiating means for automatic initiation; for initiation not subject to will of driver or passenger
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Transportation (AREA)
- Mechanical Engineering (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention provides an intelligent monitoring method and device for a vehicle. The intelligent monitoring method and device for the vehicle are used for monitoring bad behaviors of a driver. The method includes the following steps: monitoring the driving behavior of the driver, and if the driver does not have a bad driving behavior, continuing to monitor; if the driver has the bad driving behavior, carrying out the following steps of obtaining an image of the driver, and by judging whether there is a mobile phone within a predetermined distance in the visibility range of the driver, judging whether the bad driving behavior is mobile phone operating or not; if it is judged that the bad driving behavior is mobile phone operating, shielding communication signals of the mobile phone by exerting electromagnetic interference so that the driver can not continue to use the mobile phone. The intelligent monitoring method and device for the vehicle can be used for monitoring a variety of bad driving behaviors including fatigue driving and giving a warning to the driver when the bad driving behaviors are detected, if the warning is invalid, disposition measures are further taken to actively control the vehicle to stop, the driving safety of the vehicle is improved, and traffic accidents are reduced.
Description
Technical field
The present invention relates to automobile intelligent monitoring field, and in particular to a kind of automobile intelligent monitoring method and device.
Background technology
Start to be widely used in automobile from three-point safety belt and air bag, the traffic safety of motorist has one
Fixed guarantee.But this kind of passive safe design, typically cannot anticipation accident emergency situations generation.The fact that simultaneously a large amount of
Illustrate, if can make before dangerous arriving forecasting in advance, or keep vehicle even running, safety coefficient will be significantly
Improve.Therefore, the concept of automobile active safety starts appearance, and the proposition of active safety, with intelligence, " is thought to automobile by autonomous
Examine ", countermeasure can in good time, be on one's own initiative taken, so as to reduce or remit the generation of traffic accident.Active safety technologies coordinate passive guarantor
Barrier design, can to greatest extent reduce traffic accident odds.
In the last few years, automobile was increasingly becoming the important selection of people's go off daily, and it is quick, comfortable to be that people bring
Traffic is experienced.However as increasing sharply for vehicles number in recent years, automobile also gives people to be brought greatly with relevant departments
Painful and worry.It is environmental pollution (as PM2.5 indexes rise) that increasingly not smooth traffic circulation, vehicle exhaust bring, serious
Road traffic accident takes place frequently, and all seriously stimulates the nerve of modern.But, the life security because directly threaten the mankind,
Traffic accident oneself become the serious problem of society today of a countries in the world facing.According to the data display of the World Health Organization, entirely
Annual at least 1,200,000 people in the world die from road accident, and 3000～50,000,000 people are injured；Circulate a notice of according to Public Security Department of China, entirely
State is informed of a case altogether the first half of the year in 2011 91,811 road traffic accidents for being related to casualties, cause 25,864 people it is dead, 106,
370 people are injured, 4.4 hundred million yuan of direct economic loss.
And in numerous illegal activities for causing traffic fatalities, it is an important inducement that driver fatigue drives.Together
From the as shown by data of the World Health Organization, in the world traffic accident accounts for total number of accident to sample caused by internal cause fatigue driving
20% or so, account for more than the 40% of especially big traffic accident.In China, equally exist than more serious fatigue driving situation, according to friendship
Logical management board's statistics, between 2008 to 2009 years, China is due directly to traffic accident number accounts for sum caused by fatigue driving
12.5%.
Car steering belongs to sensing type work, is a comprehensive very strong technical work.Its essence can be understood as one
The individual closed-loop system constituted including road environment, driver, three key elements of vehicle.Wherein the behavior pattern of driver is divided into sense again
Know information, analysis decision, implement operation three phases.Driver constantly perceives extraneous information, such as vehicle in driving procedure
Current state (including position, speed, direction, acceleration etc.), condition of road surface etc., require according to driving and condition of road surface are carried out
Rational operation, and adjusted accordingly according to feedback result, so circulation forms driving procedure.
Although fatigue is a kind of normal physiological phenomenon, when driver carries out driving work under fatigue state, its
Perception, situation judgement to surrounding environment and the manipulation ability to vehicle have different degrees of decline, therefore
Also just it is easy to affect traffic safety, traffic accident occurs.Once statistics shows, by three kinds of perception, decision-making and operation
Tired type causes the ratio of traffic accident to be respectively 52%, 38%, 10%.
Fatigue detecting (Fatigue Detecting) is, by monitoring the various fatigue characteristics of human body, tired shape to be found in time
State simultaneously provides early warning signal, is related to the multiple fields such as physiology, psychology, image procossing, motion tracking, pattern-recognition, is
One complicated and theoretical grind the problem that makes internal disorder or usurp with realistic price while having concurrently.
In the driving procedure of long period, the degree of fatigue of driver is gradually put aside, from the superficial to the deep.If can utilize
Technological means carries out real-time detection to driver's Jing Zhong states, once there is tired sign to occur, early warning, or Bian is just sent at once
The safety measures such as deceleration by force, parking are taken, then safe driving coefficient will be effectively improved.According to statistics, if driven
Reaction energy fast 0.5s of the people when traffic accident occurs, so that it may with 60% generation that traffic accident is reduced or avoided.Therefore, product
The research work about driver fatigue detection technique is carried out in pole, and the traffic environment to building safety has highly important reality
Meaning.
Except fatigue driving, also some other bad steering behaviors cause very big threat to traffic safety, such as
With the popularization of smart mobile phone, more and more bow race at every moment all hands not from mobile phone, eye not from mobile phone, even if drive
When, still there are many people to fail to lay down mobile phone, this carefree behavior defines huge hidden danger to traffic safety.Existing skill
Art seldom detects that, simultaneously for bad steering behavior, prior art is in inspection to the bad steering behavior beyond fatigue driving
All only it is to be reminded as survey is latter, this processing mode is excessively single, it is impossible to meet actual demand.
The content of the invention
At least part of solution problems of the prior art, the present invention proposes a kind of automobile intelligent monitoring method, uses
In the bad behavior of monitoring human pilot, methods described includes：
The driving behavior of monitoring human pilot,
If human pilot does not have bad steering behavior, continue to keep monitoring；
If human pilot has bad steering behavior, following steps are performed：
Obtain human pilot image；
By judging whether there is mobile phone in human pilot visual line of sight preset distance, judge whether bad steering behavior is behaviour
Make mobile phone；
When it is operating handset to judge bad steering behavior, by applying electromagnetic interference, the signal of communication of mobile phone is shielded, made
Obtaining human pilot cannot be continuing with mobile phone.
Preferably, the driving behavior of the monitoring human pilot includes：
Step S100, to human pilot Face detection is carried out；
Step S200, to human pilot human eye positioning is carried out；
Step S300, bad steering behavioral value；
Step S400, bad steering behavior warning；
Step S500, bad steering behavior is disposed；
Wherein, bad steering behavior dispose include for human pilot bow operating handset in the case of be disposed.
Preferably, Face detection of step S100 including the Face detection and infrared light under natural lighting according under, wherein,
Under natural lighting, face is positioned using the colour of skin, YCbCr color spaces are selected, in YCbCr color spaces
It is middle to determine satisfaction 109<cb<l40,135<cr<158 pixel region, thereby determines that out the position of face, and realization is determined face
Position.
Preferably, step S200 point is that human eye coarse positioning and human eye are accurately positioned, wherein,
Before human eye coarse positioning is carried out, first the hunting zone of human eye is further reduced, hunting zone is determined in face
Portion region from top to bottom 1/5 to 1/2 between；
For the facial image obtained after step S100 Face detection, using Gray Projection method to facial image from top to bottom
1/5 to 1/2 between region projected in the horizontal, obtain primary peak and secondary peak, first wave peak value is more than second
Crest value, by primary peak position human eye cross central line is defined as, and the trough between primary peak and secondary peak is determined
It is d for the distance between human eye top edge boundary, human eye cross central line and human eye top edge boundary, then, according to human eye pair
Title property, moves down apart from d from human eye cross central line, obtains eyes lower limb boundary, completes the coarse positioning of human eye；
Human eye area to coarse positioning, further carries out human eye and is accurately positioned：
Human eye template is initially set up, eyes are considered as into elliptical shape, its oval inscribed circle is eyeball, and remainder is eye
In vain, by oval and inscribed circle equation, current pixel position is drawn.I.e.：
If current pixel coordinate is (x, y)：
When pixel is located at eyeball：x2+y2＜ r2, wherein r is inradius,
When pixel is located at white of the eye region：And x2+y2＞ r2,
R represents oval major semiaxis length, and r represents oval semi-minor axis length, while be also the length of inradius,
After template is set up, slip scan is carried out using the template in the human eye area of coarse positioning, while to the white of the eye and eye
The gray scale difference of ball is calculated, in gray-scale map, when template covers position of human eye, due to the white of the eye it is bright with the gray scale difference of eyeball
It is aobvious, so the gray integration difference in the position is maximum, using the feature, by detecting respectively to the right and left, search
Where the left-right template gray scale difference value maximum found out, the exact position of two is identified as.
Preferably, step S300 includes：
Fatigue driving detection, in the given time, detects the closing time of human pilot eyes, if closing time exceedes
Predetermined threshold, then judge that fatigue driving behavior occurs in human pilot, and/or
Sight line deviation detection, in the given time, detects that human pilot sight line deviates the time of driving sight scope, if
Sight line deviates the time of driving sight scope more than predetermined threshold, then judge that human pilot occurs in that the driving row that sight line deviates
For.
Preferably, step S500 includes that the bad steering behavior number of times that ought be detected in the given time exceedes given threshold
Value, then take Disposal Measures to bad steering behavior, further includes：
Vehicle drive authority switches, and vehicle drive authority is switched to into automatic driving mode, under automatic driving mode, drives
The personnel of sailing cannot be controlled to vehicle；
Opening warning device, the alarming device on opening vehicle to cause other people attentions, alarming device include hazard way light,
Alarm；
Automatic retarding stops, and step by step ramp to stop, prevents by other vehicle rear-end collisions；
Preferably, automatic retarding stops and includes：
Positioning vehicle realizes the inspection to lane line and road surface by vehicle-mounted camera, vehicle GPS in the position of road
Survey, determine vehicle in the position of the track situation and road surface rightmost side roadside on the position of road and road surface；
By vehicle-mounted camera, determine vehicle front, right front, right side, right back whether have vehicle and with from
The distance between body vehicle；
The speed of other vehicles is determined by the distance between speed and other vehicles and vehicle of vehicle；
When right side does not have vehicle, and front, right front, the distance of right back vehicle and speed meet predetermined condition, then control
Vehicle processed track doubling lane change to the right；
When vehicle has not had road, ramp to stop in rightmost side roadside, i.e. vehicle right side.
The invention allows for a kind of Intelligent monitoring apparatus for automobiles, for monitoring the bad behavior of human pilot, the dress
Put including：
Monitoring device, for monitoring the driving behavior of human pilot, if there is no bad steering behavior in human pilot,
Continue to keep monitoring；
Processing meanss, if knowing that human pilot has bad steering behavior by detection means, obtain human pilot
Image, by judging whether there is mobile phone in human pilot visual line of sight preset distance, judges whether bad steering behavior is operation
Mobile phone, when it is operating handset to judge bad steering behavior, by applying electromagnetic interference, shields the signal of communication of mobile phone so that
Human pilot cannot be continuing with mobile phone.
Preferably, the monitoring device includes：
Face detection module, for carrying out Face detection to human pilot；
Human eye locating module, to human pilot human eye positioning is carried out；
Bad steering behavioral value module, for detecting to bad steering behavior；
Bad steering behavior alert module, for alerting bad steering behavior.
Wherein, bad steering behavior dispose module include for human pilot bow operating handset in the case of be disposed.
The present invention to detecting including the various bad steering behaviors including fatigue driving, and can detected not
After good driving behavior, alerted to human pilot, when alerting invalid, further taken Disposal Measures, active control vehicle
Stagnation of movement, improves the security of vehicle traveling, reduces the generation of traffic accident.
Description of the drawings
Fig. 1 is a kind of flow chart of automobile intelligent monitoring method of the invention；
Fig. 2 is a kind of block diagram of Intelligent monitoring apparatus for automobiles of the invention；
Specific embodiment
Below in conjunction with the accompanying drawing of the present invention, technical scheme is clearly and completely described.Here will be detailed
Carefully exemplary embodiment is illustrated, its example is illustrated in the accompanying drawings.When explained below is related to accompanying drawing, unless otherwise table
Show, the same numbers in different accompanying drawings represent same or analogous key element.Embodiment party described in following exemplary embodiment
Formula does not represent all embodiments consistent with the present invention.Conversely, they be only with as institute is in detail in appended claims
The example of the consistent apparatus and method of some aspects state, the present invention.
Referring to Fig. 1, the present invention proposes a kind of automobile intelligent monitoring method, including：
Step S100, to human pilot Face detection is carried out；
In order to realize to the Face detection under different illumination conditions, the step is divided into the Face detection under natural lighting and red
Face detection under outer illumination.
Under natural lighting, it is possible to use colour information is positioned to face, wherein the colour of skin is often utilizable
One of important information.Through statistical confirmation, the colour of skin difference of different ethnic groups is mainly affected by brightness, is hardly affected by colourity.
In order to effectively carry out colour of skin cluster, the present invention selects YCbCr color spaces.Many experiments show that the area of skin color of people exists
Approximate location in YCbCr space is:109<cb<l40,135<cr<158, accordingly the present invention is by YCbCr color spaces
It is determined that meeting 109<cb<l40,135<cr<158 pixel region, thus can determine that the position of face, realize to face
Positioning.
Can utilize without colouring information in the case where infrared light shines, so the method for color space is utilized as the colour of skin just
It is obstructed, only using the half-tone information of image.Since it is considered that in true driving environment, the head of driver is not always
It is motionless, but the moment occur to rock, it is contemplated that the method for motion detection Face detection out, here it is contemplated that
Positioned using context update.If, then present frame and initial frame are compared, will obtain without object in initial frame
Whole mobile object.Obviously, this method does not rely on the speed of object movement, and the method for this renewal is referred to as background more
Newly " basic ideas that dynamic background updates are:The region of variation in image is found out in time interval image difference method.Then will
The background in non-changing region is with current picture replacement.I.e. update background after this renewal has been done to background, the Hen great areas of background
Domain has been exactly a part for current picture, so go to carry out difference with next images with this background, in very big area
It is actually time interval image difference method in domain, therefore it not only inherits time interval image difference method to scene
Minor variations.Particularly the adaptability of light change, essentially eliminates accumulation property error, and and with fixed background difference side
The accurate advantage of method detection moving object.
The present invention can also realize Face detection using other various prior arts.
Step S200, to human pilot human eye positioning is carried out；
The step is divided into human eye coarse positioning again and human eye is accurately positioned.
Before human eye coarse positioning is carried out, first the hunting zone of human eye can further be reduced according to standard empirical, will
Hunting zone determine face area from top to bottom 1/5 to 1/2 between, can so reduce the amount of calculation of calculating process, make
System speed is accelerated, and meets the requirement of its real-time.
For the facial image obtained after step S100 Face detection, using Gray Projection method to facial image from top to bottom
1/5 to 1/2 between region projected in the horizontal, obtain primary peak and secondary peak, first wave peak value is more than second
Crest value, by primary peak position human eye cross central line is defined as, and the trough between primary peak and secondary peak is determined
It is d for the distance between human eye top edge boundary, human eye cross central line and human eye top edge boundary, then, according to human eye pair
Title property, moves down apart from d from human eye cross central line, obtains eyes lower limb boundary, completes the coarse positioning of human eye.
Human eye area to coarse positioning, further carries out human eye and is accurately positioned.
Human eye template is initially set up, generally, eyes can be considered as elliptical shape, its oval inscribed circle is eyeball,
Remainder is the white of the eye, by oval and inscribed circle equation, it can be deduced that current pixel position.I.e.：
If current pixel coordinate is (x, y)：
When pixel is located at eyeball：x2+y2＜ r2, wherein r is inradius,
When pixel is located at white of the eye region：And x2+y2＞ r2。
R represents oval major semiaxis length, and r represents oval semi-minor axis length, while being also the length of inradius.
After template is set up, slip scan is carried out using the template in the human eye area of coarse positioning, while to the white of the eye and eye
The gray scale difference of ball is calculated, and be can be seen that in gray-scale map, when template covers position of human eye, due to the white of the eye and the ash of eyeball
Degree difference is obvious, so the gray integration difference in the position is maximum, thus using the feature, by distinguishing the right and left
Detection, where the left-right template gray scale difference value maximum that search is found out, is identified as the exact position of two.
The present invention can also realize that human eye is positioned using other various prior arts.
Step S300, bad steering behavioral value；
Bad steering behavior includes fatigue driving, bowing sees the mobile phone or other sight lines long-time deviates driving sight scopes
Behavior.
Fatigue driving detection, in the given time, detects the closing time of human pilot eyes, if closing time exceedes
Predetermined threshold, then judge that fatigue driving behavior occurs in human pilot.
For normal driving behavior, the sight line of human pilot should be to watch vehicle front attentively within the most of time,
To guarantee to observe the surface conditions of vehicle front, human pilot is able to observe that the present invention road surface feelings of vehicle front
The visual line of sight of condition is referred to as driving sight scope.And work as human pilot and bow and see the mobile phone or do other actions unrelated with driving
When, visual line of sight at this moment can deviate from driving sight scope.
Sight line deviation detection, in the given time, detects that human pilot sight line deviates the time of driving sight scope, if
Sight line deviates the time of driving sight scope more than predetermined threshold, then judge that human pilot occurs in that the driving row that sight line deviates
For, such as be frequently see the mobile phone or do other influences driving behavior.
Step S400, bad steering behavior warning；
After detecting bad steering behavior, given a warning to human pilot by voice, to remind human pilot to note.
Step S500, bad steering behavior is disposed；
When the bad steering behavior number of times for detecting in the given time exceedes given threshold value, then bad steering behavior is adopted
Take Disposal Measures.
When human pilot is in extremely tired state or wallows under handset content, or extreme case, human pilot
Body occurs in that abnormal conditions, such as loses consciousness, and at this moment external alert mode may not cause the attention of human pilot, also
It is to say that the warning mode of at this moment external alert is invalid, at this moment needs Disposal Measures of taking the initiative, avoids traffic accident.
Bad steering behavior Disposal Measures include：
Vehicle drive authority switches, and vehicle drive authority is switched to into automatic driving mode, under automatic driving mode, drives
The personnel of sailing cannot be controlled to vehicle, and certainly, vehicle drive authority can be switched to manual drive mould by human pilot by hand
Formula.
Opening warning device, in order to remind other vehicles and behavior on road, the alarming device on opening vehicle is causing him
People notices that alarming device includes hazard way light, alarm etc..
Automatic retarding stop, in order to traffic hazard occurs in mode, need that vehicle stops, certainly, during parking need by
Level ramp to stop, to prevent by danger such as other vehicle rear-end collisions.
In order to not affect the traffic of other vehicles and pedestrian on road, it is preferred that automatic running is to the rightmost side road during parking
Stop on side.In order to realize that rightmost side roadside is stopped, it is necessary first to position vehicle in the position of road, this is related to lane line
Detection and pavement detection, can pass through the detection of vehicle-mounted camera, vehicle GPS realization to lane line and road surface, so that it is determined that itself
Vehicle is in the position of the track situation and road surface rightmost side roadside on the position of road and road surface.In order to by vehicle parking to road
Rightmost side roadside, vehicle needs to carry out doubling lane change, at this moment needs the situation for considering other vehicles on road surface, is taken the photograph by vehicle-mounted
As head, determine whether vehicle front, right front, right side, right back have vehicle and the distance between with vehicle.
When carrying out doubling lane change, distance is not only considered, in addition it is also necessary to consider speed, therefore also need to detect vehicle front, the right side
Front, right side, the speed of right back vehicle, can by between the speed of vehicle and other vehicles and vehicle away from
From determining the speed of other vehicles, such as, in preset time, other vehicles remain constant with the distance of vehicle, then
Other vehicles are equal with the speed of vehicle, if front or right front vehicle become remote with the distance of vehicle, before explanation
The speed of side or right front vehicle is more than the speed of vehicle, further according to the relation between distance and time, you can calculate
Front or the speed of right front vehicle, if right back vehicle becomes remote with the distance of vehicle, illustrate the car of right back vehicle
Speed is less than the speed of vehicle, further according to the relation between distance and time, you can calculate the speed of right back vehicle.When
Right side is without vehicle, and front, right front, the distance of right back vehicle and speed meet predetermined condition, then control vehicle to the right
The doubling lane change of side track.When vehicle has not had road in rightmost side roadside, i.e. vehicle right side, at this moment ramp to stop.
In view of the situation of road traffic signal lamp, it is preferred that the present invention also believes under automatic driving mode with traffic
Signal lamp detects function, and when traffic lights are detected, control vehicle deceleration is gone slowly, when detecting traffic lights for red light or
During amber light, vehicle deceleration stagnation of movement is controlled.
Different in view of the value of different brands vehicle, the loss for occurring to be produced during accident is different, and the present invention becomes in doubling
During road, also add carries out the identification function of brand and vehicle to other vehicles, by logo, the profile etc. that obtain other vehicles
Characteristic feature, then compare with the information that prestores in database, determine the brand and vehicle of vehicle.For different brands and car
The vehicle of type, sets different doubling lane change predetermined conditions.Such as very high vehicle is worth, lane change distance can be arranged
It is longer.
Preferably, for human pilot bow operating handset in the case of, can also further carry out special disposal.Pass through
Human pilot image is obtained, judges whether there is mobile phone in its visual line of sight preset distance, you can whether judge bad steering behavior
For operating handset.When it is operating handset to judge bad steering behavior, by applying electromagnetic interference, the communication letter of mobile phone is shielded
Number so that human pilot cannot be continuing with mobile phone.
The above-mentioned bad steering behavior Disposal Measures content of the present invention is that vehicle drive authority is being switched to into automatic Pilot
Implement under pattern.
Referring to Fig. 2, the invention allows for a kind of Intelligent monitoring apparatus for automobiles, including：
Face detection module, for carrying out Face detection to human pilot；
In order to realize to the Face detection under different illumination conditions, the module is divided into the Face detection module under natural lighting
Face detection module with infrared light according under.
Under natural lighting, it is possible to use colour information is positioned to face, wherein the colour of skin is often utilizable
One of important information.Through statistical confirmation, the colour of skin difference of different ethnic groups is mainly affected by brightness, is hardly affected by colourity.
In order to effectively carry out colour of skin cluster, the present invention selects YCbCr color spaces.Many experiments show that the area of skin color of people exists
Approximate location in YCbCr space is:109<cb<l40,135<cr<158, accordingly the present invention is by YCbCr color spaces
It is determined that meeting 109<cb<l40,135<cr<158 pixel region, thus can determine that the position of face, realize to face
Positioning.
Can utilize without colouring information in the case where infrared light shines, so the method for color space is utilized as the colour of skin just
It is obstructed, only using the half-tone information of image.Since it is considered that in true driving environment, the head of driver is not always
It is motionless, but the moment occur to rock, it is contemplated that the method for motion detection Face detection out, here it is contemplated that
Positioned using context update.If, then present frame and initial frame are compared, will obtain without object in initial frame
Whole mobile object.Obviously, this method does not rely on the speed of object movement, and the method for this renewal is referred to as background more
Newly " basic ideas that dynamic background updates are:The region of variation in image is found out in time interval image difference method.Then will
The background in non-changing region is with current picture replacement.I.e. update background after this renewal has been done to background, the Hen great areas of background
Domain has been exactly a part for current picture, so go to carry out difference with next images with this background, in very big area
It is actually time interval image difference method in domain, therefore it not only inherits time interval image difference method to scene
Minor variations.Particularly the adaptability of light change, essentially eliminates accumulation property error, and and with fixed background difference side
The accurate advantage of method detection moving object.
The present invention can also realize Face detection using other various prior arts.
Human eye locating module, to human pilot human eye positioning is carried out；
The module is divided into human eye coarse positioning module and human eye pinpoint module again.
Before human eye coarse positioning is carried out, first the hunting zone of human eye can further be reduced according to standard empirical, will
Hunting zone determine face area from top to bottom 1/5 to 1/2 between, can so reduce the amount of calculation of calculating process, make
System speed is accelerated, and meets the requirement of its real-time.
For the facial image obtained after Face detection, using Gray Projection method to facial image 1/5 arriving from top to bottom
Region is projected in the horizontal between 1/2, obtains primary peak and secondary peak, and first wave peak value is more than secondary peak value,
Primary peak position is defined as into human eye cross central line, the trough between primary peak and secondary peak is defined as on human eye
The distance between edge boundary, human eye cross central line and human eye top edge boundary are d, then, according to human eye symmetry, from people
Eye cross central line is moved down apart from d, obtains eyes lower limb boundary, completes the coarse positioning of human eye.
Human eye area to coarse positioning, human eye pinpoint module further carries out human eye and is accurately positioned.
Human eye template is initially set up, generally, eyes can be considered as elliptical shape, its oval inscribed circle is eyeball,
Remainder is the white of the eye, by oval and inscribed circle equation, it can be deduced that current pixel position.I.e.：
If current pixel coordinate is (x, y)：
When pixel is located at eyeball：x2+y2＜ r2, wherein r is inradius,
When pixel is located at white of the eye region：And x2+y2＞ r2。
R represents oval major semiaxis length, and r represents oval semi-minor axis length, while being also the length of inradius.
After template is set up, slip scan is carried out using the template in the human eye area of coarse positioning, while to the white of the eye and eye
The gray scale difference of ball is calculated, and be can be seen that in gray-scale map, when template covers position of human eye, due to the white of the eye and the ash of eyeball
Degree difference is obvious, so the gray integration difference in the position is maximum, thus using the feature, by distinguishing the right and left
Detection, where the left-right template gray scale difference value maximum that search is found out, is identified as the exact position of two.
The present invention can also realize that human eye is positioned using other various prior arts.
Bad steering behavioral value module, for detecting to bad steering behavior；
Bad steering behavior includes fatigue driving, bowing sees the mobile phone or other sight lines long-time deviates driving sight scopes
Behavior.
Fatigue driving detection, in the given time, detects the closing time of human pilot eyes, if closing time exceedes
Predetermined threshold, then judge that fatigue driving behavior occurs in human pilot.
For normal driving behavior, the sight line of human pilot should be to watch vehicle front attentively within the most of time,
To guarantee to observe the surface conditions of vehicle front, human pilot is able to observe that the present invention road surface feelings of vehicle front
The visual line of sight of condition is referred to as driving sight scope.And work as human pilot and bow and see the mobile phone or do other actions unrelated with driving
When, visual line of sight at this moment can deviate from driving sight scope.
Sight line deviation detection, in the given time, detects that human pilot sight line deviates the time of driving sight scope, if
Sight line deviates the time of driving sight scope more than predetermined threshold, then judge that human pilot occurs in that the driving row that sight line deviates
For, such as be frequently see the mobile phone or do other influences driving behavior.
Bad steering behavior alert module, for alerting bad steering behavior；
After detecting bad steering behavior, given a warning to human pilot by voice, to remind human pilot to note.
Module is disposed in bad steering behavior, for being disposed to bad steering behavior；
When the bad steering behavior number of times for detecting in the given time exceedes given threshold value, then bad steering behavior is adopted
Take Disposal Measures.
When human pilot is in extremely tired state or wallows under handset content, or extreme case, human pilot
Body occurs in that abnormal conditions, such as loses consciousness, and at this moment external alert mode may not cause the attention of human pilot, also
It is to say that the warning mode of at this moment external alert is invalid, at this moment needs Disposal Measures of taking the initiative, avoids traffic accident.
Bad steering behavior Disposal Measures include：
Vehicle drive authority switches, and vehicle drive authority is switched to into automatic driving mode, under automatic driving mode, drives
The personnel of sailing cannot be controlled to vehicle, and certainly, vehicle drive authority can be switched to manual drive mould by human pilot by hand
Formula.
Opening warning device, in order to remind other vehicles and behavior on road, the alarming device on opening vehicle is causing him
People notices that alarming device includes hazard way light, alarm etc..
Automatic retarding stop, in order to traffic hazard occurs in mode, need that vehicle stops, certainly, during parking need by
Level ramp to stop, to prevent by danger such as other vehicle rear-end collisions.
In order to not affect the traffic of other vehicles and pedestrian on road, it is preferred that automatic running is to the rightmost side road during parking
Stop on side.In order to realize that rightmost side roadside is stopped, it is necessary first to position vehicle in the position of road, this is related to lane line
Detection and pavement detection, can pass through the detection of vehicle-mounted camera, vehicle GPS realization to lane line and road surface, so that it is determined that itself
Vehicle is in the position of the track situation and road surface rightmost side roadside on the position of road and road surface.In order to by vehicle parking to road
Rightmost side roadside, vehicle needs to carry out doubling lane change, at this moment needs the situation for considering other vehicles on road surface, is taken the photograph by vehicle-mounted
As head, determine whether vehicle front, right front, right side, right back have vehicle and the distance between with vehicle.
When carrying out doubling lane change, distance is not only considered, in addition it is also necessary to consider speed, therefore also need to detect vehicle front, the right side
Front, right side, the speed of right back vehicle, can by between the speed of vehicle and other vehicles and vehicle away from
From determining the speed of other vehicles, such as, in preset time, other vehicles remain constant with the distance of vehicle, then
Other vehicles are equal with the speed of vehicle, if front or right front vehicle become remote with the distance of vehicle, before explanation
The speed of side or right front vehicle is more than the speed of vehicle, further according to the relation between distance and time, you can calculate
Front or the speed of right front vehicle, if right back vehicle becomes remote with the distance of vehicle, illustrate the car of right back vehicle
Speed is less than the speed of vehicle, further according to the relation between distance and time, you can calculate the speed of right back vehicle.When
Right side is without vehicle, and front, right front, the distance of right back vehicle and speed meet predetermined condition, then control vehicle to the right
The doubling lane change of side track.When vehicle has not had road in rightmost side roadside, i.e. vehicle right side, at this moment ramp to stop.
In view of the situation of road traffic signal lamp, it is preferred that the present invention also believes under automatic driving mode with traffic
Signal lamp detects function, and when traffic lights are detected, control vehicle deceleration is chronic, when detecting traffic lights for red light or
During amber light, vehicle deceleration stagnation of movement is controlled.
Different in view of the value of different brands vehicle, the loss for occurring to be produced during accident is different, and the present invention becomes in doubling
During road, also add carries out the identification function of brand and vehicle to other vehicles, by logo, the profile etc. that obtain other vehicles
Characteristic feature, then compare with the information that prestores in database, determine the brand and vehicle of vehicle.For different brands and car
The vehicle of type, sets different doubling lane change predetermined conditions.Such as very high vehicle is worth, lane change distance can be arranged
It is longer.
Preferably, for human pilot bow operating handset in the case of, can also further carry out special disposal.Pass through
Human pilot image is obtained, judges whether there is mobile phone in its visual line of sight preset distance, you can whether judge bad steering behavior
For operating handset.When it is operating handset to judge bad steering behavior, by applying electromagnetic interference, the communication letter of mobile phone is shielded
Number so that human pilot cannot be continuing with mobile phone.
The present invention to detecting including the various bad steering behaviors including fatigue driving, and can detected not
After good driving behavior, alerted to human pilot, when alerting invalid, further taken Disposal Measures, active control vehicle
Stagnation of movement, improves the security of vehicle traveling, reduces the generation of traffic accident.
Those skilled in the art will readily occur to its of the present invention after considering specification and putting into practice invention disclosed herein
Its embodiment.The application is intended to any modification of the present invention, purposes or adaptations, these modifications, purposes or
Person's adaptations follow the general principle of the present invention and including the undocumented common knowledge in the art of the present invention
Or conventional techniques.
It should be appreciated that the precision architecture for being described above and being shown in the drawings is the invention is not limited in, and
And can without departing from the scope carry out various modifications and changes.The scope of the present invention is only limited by appended claim.
Claims (9)
1. a kind of automobile intelligent monitoring method, for monitoring the bad behavior of human pilot, methods described includes：
The driving behavior of monitoring human pilot,
If human pilot does not have bad steering behavior, continue to keep monitoring；
If human pilot has bad steering behavior, following steps are performed：
Obtain human pilot image；
By judging whether there is mobile phone in human pilot visual line of sight preset distance, judge whether bad steering behavior is manipulator
Machine；
When it is operating handset to judge bad steering behavior, by applying electromagnetic interference, the signal of communication of mobile phone is shielded so that drive
The personnel of sailing cannot be continuing with mobile phone.
2. automobile intelligent monitoring method according to claim 1, the driving behavior of the monitoring human pilot includes：
Step S100, to human pilot Face detection is carried out；
Step S200, to human pilot human eye positioning is carried out；
Step S300, bad steering behavioral value；
Step S400, bad steering behavior warning；
Step S500, bad steering behavior is disposed；
Wherein, bad steering behavior dispose include for human pilot bow operating handset in the case of be disposed.
3. automobile intelligent monitoring method as claimed in claim 2, step S100 includes the Face detection under natural lighting and red
Face detection under outer illumination, wherein,
Under natural lighting, face is positioned using the colour of skin, select YCbCr color spaces, in YCbCr color spaces really
Surely 109 are met<cb<l40,135<cr<158 pixel region, thereby determines that out the position of face, realizes the positioning to face.
4. automobile intelligent monitoring method as claimed in claim 2, step S200 point is that human eye coarse positioning and human eye are accurately positioned,
Wherein,
Before human eye coarse positioning is carried out, first the hunting zone of human eye is further reduced, hunting zone is determined in facial regions
Domain from top to bottom 1/5 to 1/2 between；
For the facial image obtained after step S100 Face detection, using Gray Projection method to facial image from top to bottom 1/
Region is projected in the horizontal between 5 to 1/2, obtains primary peak and secondary peak, and first wave peak value is more than secondary peak
Value, by primary peak position human eye cross central line is defined as, and the trough between primary peak and secondary peak is determined be people
Eye top edge boundary, the distance between human eye cross central line and human eye top edge boundary are d, then, according to human eye symmetry,
Move down apart from d from human eye cross central line, obtain eyes lower limb boundary, complete the coarse positioning of human eye；
Human eye area to coarse positioning, further carries out human eye and is accurately positioned：
Human eye template is initially set up, eyes are considered as into elliptical shape, its oval inscribed circle is eyeball, and remainder is the white of the eye, led to
Oval and inscribed circle equation is crossed, current pixel position is drawn.I.e.：
If current pixel coordinate is (x, y)：
When pixel is located at eyeball：x2+y2＜ r2, wherein r is inradius,
When pixel is located at white of the eye region：And x2+y2＞ r2,
R represents oval major semiaxis length, and r represents oval semi-minor axis length, while be also the length of inradius,
After template is set up, slip scan is carried out using the template in the human eye area of coarse positioning, while to the white of the eye and eyeball
Gray scale difference is calculated, in gray-scale map, when template covers position of human eye, due to the white of the eye and eyeball gray scale difference substantially,
So the gray integration difference in the position is maximum, using the feature, by detecting respectively to the right and left, search is found out
Left-right template gray scale difference value maximum where, be identified as the exact position of two.
5. automobile intelligent monitoring method as claimed in claim 2, wherein, step S300 includes：
Fatigue driving detection, in the given time, detects the closing time of human pilot eyes, if closing time exceedes made a reservation for
Threshold value, then judge that fatigue driving behavior occurs in human pilot, and/or
Sight line deviation detection, in the given time, detects that human pilot sight line deviates the time of driving sight scope, if sight line
The time for deviateing driving sight scope exceedes predetermined threshold, then judge that human pilot occurs in that the driving behavior that sight line deviates.
6. automobile intelligent monitoring method as claimed in claim 2, step S500 includes detecting in the given time not
Good driving behavior number of times exceedes given threshold value, then take Disposal Measures to bad steering behavior, further includes：
Vehicle drive authority switches, and vehicle drive authority is switched to into automatic driving mode, under automatic driving mode, driver
Member cannot be controlled to vehicle；
Opening warning device, the alarming device on opening vehicle includes hazard way light, reports to the police to cause other people attentions, alarming device
Device；
Automatic retarding stops, and step by step ramp to stop, prevents by other vehicle rear-end collisions.
7. automobile intelligent monitoring method as claimed in claim 2, wherein, automatic retarding stops to be included：
Positioning vehicle realizes the detection to lane line and road surface in the position of road by vehicle-mounted camera, vehicle GPS,
Determine vehicle in the position of the track situation and road surface rightmost side roadside on the position of road and road surface；
By vehicle-mounted camera, determine vehicle front, right front, right side, right back whether have vehicle and with itself car
The distance between；
The speed of other vehicles is determined by the distance between speed and other vehicles and vehicle of vehicle；
When right side does not have vehicle, and front, right front, the distance of right back vehicle and speed meet predetermined condition, then control car
Track doubling lane change to the right；
When vehicle has not had road, ramp to stop in rightmost side roadside, i.e. vehicle right side.
8. a kind of Intelligent monitoring apparatus for automobiles, for monitoring the bad behavior of human pilot, described device includes：
Monitoring device, for monitoring the driving behavior of human pilot, if human pilot does not have bad steering behavior, continues
Keep monitoring；
Processing meanss, if knowing that human pilot has bad steering behavior by detection means, obtain human pilot image,
By judging whether there is mobile phone in human pilot visual line of sight preset distance, judge whether bad steering behavior is operating handset,
When it is operating handset to judge bad steering behavior, by applying electromagnetic interference, the signal of communication of mobile phone is shielded so that driver
Member cannot be continuing with mobile phone.
9. intelligent monitoring and controlling device according to claim 8, the monitoring device includes：
Face detection module, for carrying out Face detection to human pilot；
Human eye locating module, to human pilot human eye positioning is carried out；
Bad steering behavioral value module, for detecting to bad steering behavior；
Bad steering behavior alert module, for alerting bad steering behavior.
Wherein, bad steering behavior dispose module include for human pilot bow operating handset in the case of be disposed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611110035.6A CN106627529A (en) | 2016-12-06 | 2016-12-06 | Intelligent monitoring method and device for vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611110035.6A CN106627529A (en) | 2016-12-06 | 2016-12-06 | Intelligent monitoring method and device for vehicle |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106627529A true CN106627529A (en) | 2017-05-10 |
Family
ID=58819659
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611110035.6A Pending CN106627529A (en) | 2016-12-06 | 2016-12-06 | Intelligent monitoring method and device for vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106627529A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107862296A (en) * | 2017-11-20 | 2018-03-30 | 深圳市深视创新科技有限公司 | The monitoring method and system of driving behavior, computer-readable recording medium |
| TWI634454B (en) * | 2017-05-19 | 2018-09-01 | 致伸科技股份有限公司 | Human perception test system and method thereof |
| CN108608940A (en) * | 2018-04-27 | 2018-10-02 | 长沙硕铠电子科技有限公司 | A method of judge smart mobile phone whether in driver's seat region |
| CN108790816A (en) * | 2018-05-24 | 2018-11-13 | 安徽中科中涣防务装备技术有限公司 | A kind of bus driver's compartment safety comprehensive electronic enclosure system |
| CN109167874A (en) * | 2018-07-18 | 2019-01-08 | 奇酷互联网络科技（深圳）有限公司 | Mobile terminal and monitoring drive the method, apparatus of user action state |
| CN110765994A (en) * | 2019-10-10 | 2020-02-07 | 广东小天才科技有限公司 | Book content positioning method and device, electronic equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN2855840Y (en) * | 2005-12-01 | 2007-01-10 | 张秀云 | Electric wave shielding device used in automobile |
| CN104228679A (en) * | 2014-09-30 | 2014-12-24 | 浙江吉利控股集团有限公司 | Device for prompting phone call receiving and making in running process of vehicle |
| CN105160913A (en) * | 2015-08-17 | 2015-12-16 | 上海斐讯数据通信技术有限公司 | Method and apparatus for standardizing driving behaviors of drivers |
| CN105261153A (en) * | 2015-11-03 | 2016-01-20 | 北京奇虎科技有限公司 | Vehicle running monitoring method and device |
| CN105574487A (en) * | 2015-11-26 | 2016-05-11 | 中国第一汽车股份有限公司 | Facial feature based driver attention state detection method |
| CN105788028A (en) * | 2016-03-21 | 2016-07-20 | 上海仰笑信息科技有限公司 | Automobile data recorder with fatigue driving pre-warning function |
-
2016
- 2016-12-06 CN CN201611110035.6A patent/CN106627529A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN2855840Y (en) * | 2005-12-01 | 2007-01-10 | 张秀云 | Electric wave shielding device used in automobile |
| CN104228679A (en) * | 2014-09-30 | 2014-12-24 | 浙江吉利控股集团有限公司 | Device for prompting phone call receiving and making in running process of vehicle |
| CN105160913A (en) * | 2015-08-17 | 2015-12-16 | 上海斐讯数据通信技术有限公司 | Method and apparatus for standardizing driving behaviors of drivers |
| CN105261153A (en) * | 2015-11-03 | 2016-01-20 | 北京奇虎科技有限公司 | Vehicle running monitoring method and device |
| CN105574487A (en) * | 2015-11-26 | 2016-05-11 | 中国第一汽车股份有限公司 | Facial feature based driver attention state detection method |
| CN105788028A (en) * | 2016-03-21 | 2016-07-20 | 上海仰笑信息科技有限公司 | Automobile data recorder with fatigue driving pre-warning function |
Non-Patent Citations (2)
| Title |
|---|
| 刘洪榛: "基于机器视觉的疲劳驾驶检测算法研究", 《华南理工大学硕士学位论文》 * |
| 郑庆: "基于肤色的人脸检测与人眼定位", 《电子科技大学硕士学位论文》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI634454B (en) * | 2017-05-19 | 2018-09-01 | 致伸科技股份有限公司 | Human perception test system and method thereof |
| CN107862296A (en) * | 2017-11-20 | 2018-03-30 | 深圳市深视创新科技有限公司 | The monitoring method and system of driving behavior, computer-readable recording medium |
| CN108608940A (en) * | 2018-04-27 | 2018-10-02 | 长沙硕铠电子科技有限公司 | A method of judge smart mobile phone whether in driver's seat region |
| CN108790816A (en) * | 2018-05-24 | 2018-11-13 | 安徽中科中涣防务装备技术有限公司 | A kind of bus driver's compartment safety comprehensive electronic enclosure system |
| CN109167874A (en) * | 2018-07-18 | 2019-01-08 | 奇酷互联网络科技（深圳）有限公司 | Mobile terminal and monitoring drive the method, apparatus of user action state |
| CN110765994A (en) * | 2019-10-10 | 2020-02-07 | 广东小天才科技有限公司 | Book content positioning method and device, electronic equipment and storage medium |
| CN110765994B (en) * | 2019-10-10 | 2023-05-26 | 广东小天才科技有限公司 | Book content positioning method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106585629B (en) | A kind of control method for vehicle and device | |
| CN106627529A (en) | Intelligent monitoring method and device for vehicle | |
| CN101512617B (en) | Travel information providing device | |
| CN113223270B (en) | Fatigue driving monitoring, reminding and early warning method and system based on computer vision technology | |
| US6927694B1 (en) | Algorithm for monitoring head/eye motion for driver alertness with one camera | |
| CN108446600A (en) | A kind of vehicle driver's fatigue monitoring early warning system and method | |
| CN103886307B (en) | A kind of eye tracking and giving fatigue pre-warning method | |
| CN108407813A (en) | A kind of antifatigue safe driving method of vehicle based on big data | |
| CN108639055A (en) | The control method of the display system of vehicle and the display system of vehicle | |
| CN110826369A (en) | Driver attention detection method and system during driving | |
| CN106355838A (en) | Fatigue driving detection method and system | |
| CN102752458A (en) | Driver fatigue detection mobile phone and unit | |
| CN103714660A (en) | System for achieving fatigue driving judgment on basis of image processing and fusion between heart rate characteristic and expression characteristic | |
| CN110276968A (en) | A kind of fog-zone highway super long tunnel entrance security inducible system | |
| CN101783964A (en) | Auxiliary driving system for achromate or tritanope based on image identification technology | |
| CN107292251A (en) | A kind of Driver Fatigue Detection and system based on human eye state | |
| Liu et al. | Driver fatigue detection through pupil detection and yawing analysis | |
| CN104318237A (en) | Fatigue driving warning method based on face identification | |
| CN104751600A (en) | Anti-fatigue driving safety equipment based on iris recognition and application method thereof | |
| CN201716818U (en) | Driver doze alarm device | |
| JPH07329657A (en) | Warning display device for vehicle | |
| CN107599956A (en) | One kind meets at night safety control system | |
| CN106650635A (en) | Method and system for detecting rearview mirror viewing behavior of driver | |
| CN106652353A (en) | Traffic tool control method and device | |
| CN109866686A (en) | The intelligent active safety DAS (Driver Assistant System) and method analyzed in real time based on video |
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: 20170510 |