CN108230619A - Method for detecting fatigue driving based on multi-feature fusion - Google Patents
Method for detecting fatigue driving based on multi-feature fusion Download PDFInfo
- Publication number
- CN108230619A CN108230619A CN201611150273.XA CN201611150273A CN108230619A CN 108230619 A CN108230619 A CN 108230619A CN 201611150273 A CN201611150273 A CN 201611150273A CN 108230619 A CN108230619 A CN 108230619A
- Authority
- CN
- China
- Prior art keywords
- fatigue
- less
- values
- state
- yaw
- 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.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/02—Alarms for ensuring the safety of persons
- G08B21/06—Alarms for ensuring the safety of persons indicating a condition of sleep, e.g. anti-dozing alarms
-
- 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
-
- 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
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
- G06V40/165—Detection; Localisation; Normalisation using facial parts and geometric relationships
Abstract
The invention discloses a kind of method for detecting fatigue driving based on multi-feature fusion, include the following steps:S1:Face and road video image are acquired, carries out image preprocessing;S2:Eyes are carried out to facial image and face is accurately positioned, extract eyes and face picture;S3:Realize eyes opening and closing detection and face aperture measurement;S4:Using rectilinear stretch model, the detection of lane line is realized;S5:Fatigue is judged using PERLOS methods for the result of eyes opening and closing detection;For face aperture analysis of fatigue, fatigue is judged using aperture threshold value and duration threshold value;Track is yawed, fatigue is analyzed and determined to do track yaw using yaw rate δ;S6:Convergence analysis is done to features above, using multiple features data fusion method and the weighted average analysis of fusion multicharacteristic information, degree of fatigue is calculated;S7:According to deviation is detected, and have tired characterization on face and eye feature, corresponding early warning is just made.
Description
Technical field
The present invention relates to a kind of method for detecting fatigue driving based on multi-feature fusion.
Background technology
It is counted according to latest data, if driver occurs absent minded within the 3s times, about 80% friendship can be caused
Interpreter's event, mainly has deviation and rear-end collision, a large amount of traffic accidents have seriously affected our harmonious life.It grinds
Study carefully display, if the 1.5s before road traffic accident generation sends out early warning to driver, 90% this kind of accident can be avoided.Institute
There are these all to say " fatigue driving is suddenly in tiger ".So fatigue driving is in addition to reasonable arrangement running time, it is also necessary to effective section
Skill prior-warning device could utmostly reduce the generation of fatigue driving accident.According to detection and analysis it is found that the fatigue characteristic the simple bright
Aobvious, the algorithm actual effect for being identified and extracting to it is better, and anti-interference is stronger.Detection method based on machine vision simultaneously
It quickly grows, non-contact detection range is wide, scalability is strong.And the mainstream of vehicle-mounted detecting system is also by being based on multisensor
The fatigue driving intelligent detecting system of information fusion technology is instead of eye detecting device.So the present invention is proposed based on driver
Facial characteristics (eyes, face feature), lanes track characteristic information carry out the research method of comprehensive analysis.
Invention content
The technical problem to be solved in the present invention is to provide a kind of method for detecting fatigue driving based on multi-feature fusion.
Method for detecting fatigue driving based on multi-feature fusion, includes the following steps:
S1:Face and road video image are acquired, and carries out image preprocessing;
S2:Being accurately positioned, and extract eyes and face picture for eyes and face is carried out for treated facial image;
S3:Eyes opening and closing detection and face aperture measurement are realized with some image processing algorithms;
S4:According to the characteristics of road image using rectilinear stretch model, the detection of lane line is realized;
S5:Fatigue is judged using PERLOS methods for the result of eyes opening and closing detection;For face aperture fatigue point
Analysis, judges fatigue using aperture threshold value and duration threshold value;Track is yawed, track yaw analysis is done using yaw rate δ
Judge fatigue;
S6:Convergence analysis is done to features above, using multiple features data fusion method and is merged adding for multicharacteristic information
Weight average analytic approach, calculates degree of fatigue;
S7:According to detecting deviation, and there is fatigue characterization just to make corresponding early warning on face and eye feature.
Further, the PERLOS methods are as follows to judge the specific method of fatigue:
S2-1:Calculate f values:
1) in the data basis provided in eyes opening and closing detection algorithm, some simple filtering methods are used so that improve
The reliability of data;
2) when the state for detecting certain two frame changes to " 0 " from " 1 ", start frame count;
3) on the basis of step 2), stop frame count when algorithm detects that certain two frame state changes to " 1 " from " 0 ";
4) in the basis of step 3), multiply the frame period time using frame number, obtain the closed-eye time of this section;
5) time scale is calculated using the following formula, and gathers the tired rule of judgement and obtain experimenter's status;
Wherein, normal wink time length is Δ t, and the closed-eye time length measured in primary measurement of closing one's eyes is Δ
T', f are PERCLOS values;
6) frame count is reset, and is entered step 1);
S2-2:Tired judgment rule based on f values:
1) when f values are more than or equal to 0.8, it is believed that driver is in fatigue state;
2) when f values are less than 0.8 and more than or equal to 0.5, and occurred being more than three times in one minute, at this moment think to drive
Personnel are slightly tired;
3) when f values are less than 0.5 and more than 0.2, it is believed that in listless state, possibly into fatigue state;
4) when f values are less than 0.2, it is believed that in waking state;
S2-3:It calculates and continues closed-eye time TC values, calculation formula is as follows:
TC=te-ts,
Wherein, to detect the frame of eye closing as starting timing point ts(eye state is changed to the frame of " 0 " by " 1 "), and handle
Next detect eye opening as timing end point t againe(eye state is changed to the frame of " 1 " by " 0 ");
S2-4:Tired judgment rule based on TC values:
1) it is considered at major fatigue state when TC values are more than 1.5s;
2) it is considered at moderate fatigue state when TC values were less than 1.5s and more than 1 second;
3) it is considered at slight fatigue state when TC values are less than 1s and more than 0.5s;
4) spiritual kilter is in when TC values are less than 0.5s.
Further, judge that the specific method of fatigue is as follows using aperture threshold value and duration threshold value:
S3-1:Yawn threshold value λ=85, yawn duration threshold values Δ t=3s;
S3-2:Judge the specific steps of fatigue:
1) one-dimensional gaussian filtering is done on the opening curve obtained by being calculated in original detection;
2) current opening value and λ value are compared in curve, when there is opening value more than λ value, opens frame count, until
Opening value stops frame count when being less than λ value;
3) time obtained by frame count is calculated, and the time and Δ t times are compared, while reset frame count;
4) when the time for calculating gained being greater than or equal to Δ t=3s, it is believed that this is a yawn, and records this yawn
Time;
5) the yawn frequency in the unit of account time, and fatigue index is provided according to the size of this frequency;
6) return to 1) step start the detection of a new round;
S3-3:Based on lasting yawn duration YtTired judgment rule:
1) work as YtWhen value is more than or equal to 7s, it is believed that in major fatigue state;
2) work as YtWhen value is less than 7s more than or equal to 5.5s, it is believed that in moderate fatigue state;
3) work as YtWhen value is less than 5.5s more than or equal to 4.5s, it is believed that in listless state;
4) work as YtWhen value is less than 4.5s, it is believed that in clear state.
Further, analyze and determine that the specific method of fatigue is as follows using yaw rate δ to do track yaw:
S4-1:Definition yaw function δ=f (t), when t represents to yaw δ corresponding frames as independent variable in function f (t)
Between, δtIt is the function of frame time t, if the yaw rate kept during some normal vehicle operation is δc, definition yaw degree Δ δ=|
δt-δc|;
S4-2:Yaw degree is divided into level Four, it is specific as follows:
1) it is serious to yaw when Δ δ values are more than or equal to 0.8;
2) when Δ δ is less than 0.8 and more than or equal to 0.5, moderate yaw;
3) it is slight to yaw when Δ δ is less than 0.5 and more than or equal to 0.2;
4) when Δ δ is less than 0.2, normally travel does not calculate.
Further, it is as follows to carry out computational methods for the degree of fatigue:
S5-1:Degree of fatigue weight distribution distributes level of fatigue such as following table;
Level of fatigue | PERCLOS | Yt | TC | Δδ |
w0 | Less than 0.2 | Less than 4.5 | Less than 0.5 | Less than 0.2 |
w1 | 0.2-0.5 | 4.5-5.5 | 0.5-1 | 0.2-0.5 |
w2 | 0.5-0.8 | 5.5-7 | 1-1.5 | 0.5-0.8 |
w3 | More than 0.8 | More than 7 | More than 1.5 | More than 0.8 |
S5-2:Level of fatigue w0、w1、w2And w3Each key feature is represented respectively is in spiritual good, listless, slight
The situation of fatigue and major fatigue, then each fatigue data is normalized, it is as shown in the table;
Level of fatigue | PERCLOS | Yt | TC | Δδ |
w0 | 0.25 | 0.643 | 0.33 | 0.25 |
w1 | 0.25-0.625 | 0.64-0.79 | 0.33-0.67 | 0.25-0.625 |
w2 | 0.625-1 | 0.79-1 | 0.67-1 | 0.625-1 |
w3 | 1 | 1 | 1 | 1 |
S5-3:Then each information carrys out the degree of association difference that fatigue influences the weight distribution to each fatigue characteristic information
It is as shown in the table;
Level of fatigue | a0 | a1 | a2 | a3 |
Weight 1 | 0.6 | 0.6 | 0.7 | 0.6 |
Weight 2 | 0.65 | 0.68 | 0.7 | 0.6 |
S5-4:Definition fatigue strength is Fatigue, and the formula that definition calculates new fatigue strength Fatigue is as follows:
Fatigue=PERCLOS × a0+Yt × a1+TC × a2+ Δ δ × a3,
Wherein, ai represents level of fatigue weight, and the value of i is 0,1,2,3, corresponds to a0, a1, a2 and a3 respectively;
S5-5:The division such as following table of grade is done to fatigue strength again after convergence analysis.
Further, the giving fatigue pre-warning scheme is as follows:
1) level-one alerts:When it is listless to measure degree of fatigue, starts interior voice prompt, warn driver cannot
Fatigue driving;
2) two level alerts:When measuring degree of fatigue as slight fatigue, start interior voice prompt, warning driver just locates
During dangerous driving;Meanwhile start the outer voice warning of vehicle, and externally the dangerous anticollision letter of danger prompting blinking red lamp is met in transmitting
Number;
3) three-level alerts:When measuring degree of fatigue as severe fatigue, start interior voice prompt, order driver is immediately
Pulling over observing will turn off engine after 20 seconds;Meanwhile start the outer voice of vehicle and light warning system.
The beneficial effects of the invention are as follows:
The present invention selects PRECLOS, eye-closing period, yawn duration and track yaw rate according to the characteristics of three features as special
Value indicative does fusion three fatigue characteristics analysis using weighted mean method, and Detection accuracy improves a lot.
Specific embodiment
The present invention is further elaborated for specific examples below, but not as a limitation of the invention.
Method for detecting fatigue driving based on multi-feature fusion, includes the following steps:
S1:Face and road video image are acquired, and carries out image preprocessing;
S2:Being accurately positioned, and extract eyes and face picture for eyes and face is carried out for treated facial image;
S3:Eyes opening and closing detection and face aperture measurement are realized with some image processing algorithms;
S4:According to the characteristics of road image using rectilinear stretch model, the detection of lane line is realized;
S5:Fatigue is judged using PERLOS methods for the result of eyes opening and closing detection;For face aperture fatigue point
Analysis, judges fatigue using aperture threshold value and duration threshold value;Track is yawed, track yaw analysis is done using yaw rate δ
Judge fatigue;
S6:Convergence analysis is done to features above, using multiple features data fusion method and is merged adding for multicharacteristic information
Weight average analytic approach, calculates degree of fatigue;
S7:According to detecting deviation, and there is fatigue characterization just to make corresponding early warning on face and eye feature.
PERLOS methods are as follows to judge the specific method of fatigue:
S2-1:Calculate f values:
1) in the data basis provided in eyes opening and closing detection algorithm, some simple filtering methods are used so that improve
The reliability of data;
2) when the state for detecting certain two frame changes to " 0 " from " 1 ", start frame count;
3) on the basis of step 2), stop frame count when algorithm detects that certain two frame state changes to " 1 " from " 0 ";
4) in the basis of step 3), multiply the frame period time using frame number, obtain the closed-eye time of this section;
5) time scale is calculated using the following formula, and gathers the tired rule of judgement and obtain experimenter's status;
Wherein, normal wink time length is Δ t, and the closed-eye time length measured in primary measurement of closing one's eyes is Δ
T', f are PERCLOS values;
6) frame count is reset, and is entered step 1);
S2-2:Tired judgment rule based on f values:
1) when f values are more than or equal to 0.8, it is believed that driver is in fatigue state;
2) when f values are less than 0.8 and more than or equal to 0.5, and occurred being more than three times in one minute, at this moment think to drive
Personnel are slightly tired;
3) when f values are less than 0.5 and more than 0.2, it is believed that in listless state, possibly into fatigue state;
4) when f values are less than 0.2, it is believed that in waking state;
S2-3:It calculates and continues closed-eye time TC values, calculation formula is as follows:
TC=te-ts,
Wherein, to detect the frame of eye closing as starting timing point ts(eye state is changed to the frame of " 0 " by " 1 "), and handle
Next detect eye opening as timing end point t againe(eye state is changed to the frame of " 1 " by " 0 ");
S2-4:Tired judgment rule based on TC values:
1) it is considered at major fatigue state when TC values are more than 1.5s;
2) it is considered at moderate fatigue state when TC values were less than 1.5s and more than 1 second;
3) it is considered at slight fatigue state when TC values are less than 1s and more than 0.5s;
4) spiritual kilter is in when TC values are less than 0.5s.
Judge that the specific method of fatigue is as follows using aperture threshold value and duration threshold value:
S3-1:Yawn threshold value λ=85, yawn duration threshold values Δ t=3s;
S3-2:Judge the specific steps of fatigue:
1) one-dimensional gaussian filtering is done on the opening curve obtained by being calculated in original detection;
2) current opening value and λ value are compared in curve, when there is opening value more than λ value, opens frame count, until
Opening value stops frame count when being less than λ value;
3) time obtained by frame count is calculated, and the time and Δ t times are compared, while reset frame count;
4) when the time for calculating gained being greater than or equal to Δ t=3s, it is believed that this is a yawn, and records this yawn
Time;
5) the yawn frequency in the unit of account time, and fatigue index is provided according to the size of this frequency;
6) return to 1) step start the detection of a new round;
S3-3:Based on lasting yawn duration YtTired judgment rule:
1) work as YtWhen value is more than or equal to 7s, it is believed that in major fatigue state;
2) work as YtWhen value is less than 7s more than or equal to 5.5s, it is believed that in moderate fatigue state;
3) work as YtWhen value is less than 5.5s more than or equal to 4.5s, it is believed that in listless state;
4) work as YtWhen value is less than 4.5s, it is believed that in clear state.
Analyze and determine that the specific method of fatigue is as follows using yaw rate δ to do track yaw:
S4-1:Definition yaw function δ=f (t), when t represents 6 corresponding frame of yaw as independent variable in function f (t)
Between, δtIt is the function of frame time t, if the yaw rate kept during some normal vehicle operation is δc, definition yaw degree Δ δ=|
δt-δc|;
S4-2:Yaw degree is divided into level Four, it is specific as follows:
1) it is serious to yaw when Δ δ values are more than or equal to 0.8;
2) when Δ δ is less than 0.8 and more than or equal to 0.5, moderate yaw;
3) it is slight to yaw when Δ δ is less than 0.5 and more than or equal to 0.2;
4) when Δ δ is less than 0.2, normally travel does not calculate.
It is as follows that degree of fatigue carries out computational methods:
S5-1:Degree of fatigue weight distribution distributes level of fatigue such as following table;
Level of fatigue | PERCLOS | Yt | TC | Δδ |
w0 | Less than 0.2 | Less than 4.5 | Less than 0.5 | Less than 0.2 |
w1 | 0.2-0.5 | 4.5-5.5 | 0.5-1 | 0.2-0.5 |
w2 | 0.5-0.8 | 5.5-7 | 1-1.5 | 0.5-0.8 |
w3 | More than 0.8 | More than 7 | More than 1.5 | More than 0.8 |
S5-2:Level of fatigue w0、w1、w2And w3Each key feature is represented respectively is in spiritual good, listless, slight
The situation of fatigue and major fatigue, then each fatigue data is normalized, it is as shown in the table;
Level of fatigue | PERCLOS | Yt | TC | Δδ |
w0 | 0.25 | 0.643 | 0.33 | 0.25 |
w1 | 0.25-0.625 | 0.64-0.79 | 0.33-0.67 | 0.25-0.625 |
w1 | 0.625-1 | 0.79-1 | 0.67-1 | 0.625-1 |
w3 | 1 | 1 | 1 | 1 |
S5-3:Then each information carrys out the degree of association difference that fatigue influences the weight distribution to each fatigue characteristic information
It is as shown in the table;
Level of fatigue | a0 | a1 | a2 | a3 |
Weight 1 | 0.6 | 0.6 | 0.7 | 0.6 |
Weight 2 | 0.65 | 0.68 | 0.7 | 0.6 |
S5-4:Definition fatigue strength is Fatigue, and the formula that definition calculates new fatigue strength Fatigue is as follows:
Fatigue=PERCLOS × a0+Yt × a1+TC × a2+ Δ δ × a3,
Wherein, ai represents level of fatigue weight, and the value of i is 0,1,2,3, corresponds to a0, a1, a2 and a3 respectively;
S5-5:The division such as following table of grade is done to fatigue strength again after convergence analysis.
Giving fatigue pre-warning scheme is as follows:
1) level-one alerts:When it is listless to measure degree of fatigue, starts interior voice prompt, warn driver cannot
Fatigue driving;
2) two level alerts:When measuring degree of fatigue as slight fatigue, start interior voice prompt, warning driver just locates
During dangerous driving;Meanwhile start the outer voice warning of vehicle, and externally the dangerous anticollision letter of danger prompting blinking red lamp is met in transmitting
Number;
3) three-level alerts:When measuring degree of fatigue as severe fatigue, start interior voice prompt, order driver is immediately
Pulling over observing will turn off engine after 20 seconds;Meanwhile start the outer voice of vehicle and light warning system.
Claims (6)
1. method for detecting fatigue driving based on multi-feature fusion, which is characterized in that include the following steps:
S1:Face and road video image are acquired, and carries out image preprocessing;
S2:Being accurately positioned, and extract eyes and face picture for eyes and face is carried out for treated facial image;
S3:Eyes opening and closing detection and face aperture measurement are realized with some image processing algorithms;
S4:According to the characteristics of road image using rectilinear stretch model, the detection of lane line is realized;
S5:Fatigue is judged using PERLOS methods for the result of eyes opening and closing detection;For face aperture analysis of fatigue, adopt
Judge fatigue with aperture threshold value and duration threshold value;Track is yawed, it is tired that track yaw analytical judgment is done using yaw rate δ
Labor;
S6:Convergence analysis is done to features above, is put down using multiple features data fusion method and the weighting for merging multicharacteristic information
Equal analytic approach, calculates degree of fatigue;
S7:According to detecting deviation, and there is fatigue characterization just to make corresponding early warning on face and eye feature.
2. method for detecting fatigue driving according to claim 1, which is characterized in that the PERLOS methods judge fatigue
Specific method it is as follows:
S2-1:Calculate f values:
1) in the data basis provided in eyes opening and closing detection algorithm, some simple filtering methods are used so that improve data
Reliability;
2) when the state for detecting certain two frame changes to " 0 " from " 1 ", start frame count;
3) on the basis of step 2), stop frame count when algorithm detects that certain two frame state changes to " 1 " from " 0 ";
4) in the basis of step 3), multiply the frame period time using frame number, obtain the closed-eye time of this section;
5) time scale is calculated using the following formula, and gathers the tired rule of judgement and obtain experimenter's status;
Wherein, normal wink time length is Δ t, and the closed-eye time length measured in primary measurement of closing one's eyes is for Δ t', f
PERCLOS values;
6) frame count is reset, and is entered step 1);
S2-2:Tired judgment rule based on f values:
1) when f values are more than or equal to 0.8, it is believed that driver is in fatigue state;
2) when f values are less than 0.8 and more than or equal to 0.5, and occurred being more than three times in one minute, at this moment think driver
Slight fatigue;
3) when f values are less than 0.5 and more than 0.2, it is believed that in listless state, possibly into fatigue state;
4) when f values are less than 0.2, it is believed that in waking state;
S2-3:It calculates and continues closed-eye time TC values, calculation formula is as follows:
TC=te-ts,
Wherein, to detect the frame of eye closing as starting timing point ts(eye state is changed to the frame of " 0 " by " 1 "), and next
Detect eye opening as timing end point t againe(eye state is changed to the frame of " 1 " by " 0 ");
S2-4:Tired judgment rule based on TC values:
1) it is considered at major fatigue state when TC values are more than 1.5s;
2) it is considered at moderate fatigue state when TC values were less than 1.5s and more than 1 second;
3) it is considered at slight fatigue state when TC values are less than 1s and more than 0.5s;
4) spiritual kilter is in when TC values are less than 0.5s.
3. method for detecting fatigue driving according to claim 1, which is characterized in that using aperture threshold value and duration threshold value come
Judge that the specific method of fatigue is as follows:
S3-1:Yawn threshold value λ=85, yawn duration threshold values Δ t=3s;
S3-2:Judge the specific steps of fatigue:
1) one-dimensional gaussian filtering is done on the opening curve obtained by being calculated in original detection;
2) current opening value and λ value are compared in curve, when there is opening value more than λ value, frame count is opened, until aperture
Value stops frame count when being less than λ value;
3) time obtained by frame count is calculated, and the time and Δ t times are compared, while reset frame count;
4) when calculate gained time be greater than or equal to Δ t=3s when, it is believed that this is a yawn, and record this yawn when
Between;
5) the yawn frequency in the unit of account time, and fatigue index is provided according to the size of this frequency;
6) return to 1) step start the detection of a new round;
S3-3:Based on lasting yawn duration YtTired judgment rule:
1) work as YtWhen value is more than or equal to 7s, it is believed that in major fatigue state;
2) work as YtWhen value is less than 7s more than or equal to 5.5s, it is believed that in moderate fatigue state;
3) work as YtWhen value is less than 5.5s more than or equal to 4.5s, it is believed that in listless state;
4) work as YtWhen value is less than 4.5s, it is believed that in clear state.
4. method for detecting fatigue driving according to claim 1, which is characterized in that track yaw is done using yaw rate δ
Analyze and determine that the specific method of fatigue is as follows:
S4-1:Definition yaw function δ=f (t), t represents the corresponding frame times of yaw δ, δ as independent variable in function f (t)tIt is
The function of frame time t, if the yaw rate kept during some normal vehicle operation is δc, definition yaw degree Δ δ=| δt-δc|;
S4-2:Yaw degree is divided into level Four, it is specific as follows:
1) it is serious to yaw when Δ δ values are more than or equal to 0.8;
2) when Δ δ is less than 0.8 and more than or equal to 0.5, moderate yaw;
3) it is slight to yaw when Δ δ is less than 0.5 and more than or equal to 0.2;
4) when Δ δ is less than 0.2, normally travel does not calculate.
5. method for detecting fatigue driving according to claim 1, which is characterized in that the degree of fatigue carries out computational methods
It is as follows:
S5-1:Degree of fatigue weight distribution distributes level of fatigue such as following table;
S5-2:Level of fatigue w0、w1、w2And w3Represent respectively each key feature be in spirit it is good, listless, slight fatigue and
The situation of major fatigue, then each fatigue data is normalized, it is as shown in the table;
S5-3:Then each information is carried out the degree of association difference that fatigue influences as follows to the weight distribution of each fatigue characteristic information
Shown in table;
S5-4:Definition fatigue strength is Fatigue, and the formula that definition calculates new fatigue strength Fatigue is as follows:
Fatigue=PERCLOS × a0+Yt × a1+TC × a2+ Δ δ × a3,
Wherein, ai represents level of fatigue weight, and the value of i is 0,1,2,3, corresponds to a0, a1, a2 and a3 respectively;
S5-5:The division such as following table of grade is done to fatigue strength again after convergence analysis.
6. method for detecting fatigue driving according to claim 1, which is characterized in that the giving fatigue pre-warning scheme is as follows:
1) level-one alerts:When it is listless to measure degree of fatigue, start interior voice prompt, warn driver cannot fatigue
It drives;
2) two level alerts:When measuring degree of fatigue as slight fatigue, start interior voice prompt, warning driver is in danger
In dangerous driving procedure;Meanwhile start the outer voice warning of vehicle, and externally the dangerous anticollision signal of danger prompting blinking red lamp is met in transmitting;
3) three-level alerts:When measuring degree of fatigue as severe fatigue, start interior voice prompt, order driver keeps to the side immediately
Parking, engine is will turn off after 20 seconds;Meanwhile start the outer voice of vehicle and light warning system.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611150273.XA CN108230619A (en) | 2016-12-14 | 2016-12-14 | Method for detecting fatigue driving based on multi-feature fusion |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611150273.XA CN108230619A (en) | 2016-12-14 | 2016-12-14 | Method for detecting fatigue driving based on multi-feature fusion |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108230619A true CN108230619A (en) | 2018-06-29 |
Family
ID=62638151
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611150273.XA Withdrawn CN108230619A (en) | 2016-12-14 | 2016-12-14 | Method for detecting fatigue driving based on multi-feature fusion |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108230619A (en) |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109035718A (en) * | 2018-07-31 | 2018-12-18 | 苏州清研微视电子科技有限公司 | The dangerous driving behavior grading forewarning system method of multifactor fusion |
CN109272764A (en) * | 2018-09-30 | 2019-01-25 | 广州鹰瞰信息科技有限公司 | A kind of based reminding method and system of dangerous driving |
CN109367479A (en) * | 2018-08-31 | 2019-02-22 | 南京理工大学 | A kind of fatigue driving monitoring method and device |
CN109447025A (en) * | 2018-11-08 | 2019-03-08 | 北京旷视科技有限公司 | Fatigue detection method, device, system and computer readable storage medium |
CN109445393A (en) * | 2018-11-14 | 2019-03-08 | 重庆工业职业技术学院 | Light sensing automatic lighting system based on mobile terminal location |
CN109887239A (en) * | 2019-03-16 | 2019-06-14 | 南京英诺微盛光学科技有限公司 | It is a kind of for monitoring the wearable device and application method of fatigue driving |
CN110334600A (en) * | 2019-06-03 | 2019-10-15 | 武汉工程大学 | A kind of multiple features fusion driver exception expression recognition method |
CN110728824A (en) * | 2019-09-25 | 2020-01-24 | 东南大学 | Driver fatigue state detection and reminding method based on multi-source data |
CN110751011A (en) * | 2019-05-23 | 2020-02-04 | 北京嘀嘀无限科技发展有限公司 | Driving safety detection method, driving safety detection device and vehicle-mounted terminal |
CN110781872A (en) * | 2019-12-31 | 2020-02-11 | 南斗六星系统集成有限公司 | Driver fatigue grade recognition system with bimodal feature fusion |
CN113240885A (en) * | 2021-04-27 | 2021-08-10 | 宁波职业技术学院 | Method for detecting fatigue of vehicle-mounted driver |
CN113643512A (en) * | 2021-07-28 | 2021-11-12 | 北京中交兴路信息科技有限公司 | Fatigue driving detection method and device, electronic equipment and storage medium |
CN113838265A (en) * | 2021-09-27 | 2021-12-24 | 科大讯飞股份有限公司 | Fatigue driving early warning method and device and electronic equipment |
CN114132326A (en) * | 2021-11-26 | 2022-03-04 | 北京经纬恒润科技股份有限公司 | Method and device for processing fatigue driving |
CN116439710A (en) * | 2023-04-11 | 2023-07-18 | 中国人民解放军海军特色医学中心 | Ship driver fatigue detection system and method based on physiological signals |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102509418A (en) * | 2011-10-11 | 2012-06-20 | 东华大学 | Fatigue driving estimation and early-warning method and device of multi-sensor information fusion |
CN103400471A (en) * | 2013-08-12 | 2013-11-20 | 电子科技大学 | Detecting system and detecting method for fatigue driving of driver |
CN104112334A (en) * | 2013-04-16 | 2014-10-22 | 百度在线网络技术(北京)有限公司 | Fatigue driving early warning method and fatigue driving early warning system |
CN104112335A (en) * | 2014-07-25 | 2014-10-22 | 北京机械设备研究所 | Multi-information fusion based fatigue driving detecting method |
CN105719431A (en) * | 2016-03-09 | 2016-06-29 | 深圳市中天安驰有限责任公司 | Fatigue driving detection system |
-
2016
- 2016-12-14 CN CN201611150273.XA patent/CN108230619A/en not_active Withdrawn
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102509418A (en) * | 2011-10-11 | 2012-06-20 | 东华大学 | Fatigue driving estimation and early-warning method and device of multi-sensor information fusion |
CN104112334A (en) * | 2013-04-16 | 2014-10-22 | 百度在线网络技术(北京)有限公司 | Fatigue driving early warning method and fatigue driving early warning system |
CN103400471A (en) * | 2013-08-12 | 2013-11-20 | 电子科技大学 | Detecting system and detecting method for fatigue driving of driver |
CN104112335A (en) * | 2014-07-25 | 2014-10-22 | 北京机械设备研究所 | Multi-information fusion based fatigue driving detecting method |
CN105719431A (en) * | 2016-03-09 | 2016-06-29 | 深圳市中天安驰有限责任公司 | Fatigue driving detection system |
Cited By (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109035718A (en) * | 2018-07-31 | 2018-12-18 | 苏州清研微视电子科技有限公司 | The dangerous driving behavior grading forewarning system method of multifactor fusion |
CN109367479A (en) * | 2018-08-31 | 2019-02-22 | 南京理工大学 | A kind of fatigue driving monitoring method and device |
CN109272764A (en) * | 2018-09-30 | 2019-01-25 | 广州鹰瞰信息科技有限公司 | A kind of based reminding method and system of dangerous driving |
CN109447025A (en) * | 2018-11-08 | 2019-03-08 | 北京旷视科技有限公司 | Fatigue detection method, device, system and computer readable storage medium |
CN109447025B (en) * | 2018-11-08 | 2021-06-22 | 北京旷视科技有限公司 | Fatigue detection method, device, system and computer readable storage medium |
CN109445393A (en) * | 2018-11-14 | 2019-03-08 | 重庆工业职业技术学院 | Light sensing automatic lighting system based on mobile terminal location |
CN109887239A (en) * | 2019-03-16 | 2019-06-14 | 南京英诺微盛光学科技有限公司 | It is a kind of for monitoring the wearable device and application method of fatigue driving |
CN110751011A (en) * | 2019-05-23 | 2020-02-04 | 北京嘀嘀无限科技发展有限公司 | Driving safety detection method, driving safety detection device and vehicle-mounted terminal |
CN110334600A (en) * | 2019-06-03 | 2019-10-15 | 武汉工程大学 | A kind of multiple features fusion driver exception expression recognition method |
CN110728824A (en) * | 2019-09-25 | 2020-01-24 | 东南大学 | Driver fatigue state detection and reminding method based on multi-source data |
CN110728824B (en) * | 2019-09-25 | 2021-11-30 | 东南大学 | Driver fatigue state detection and reminding method based on multi-source data |
CN110781872A (en) * | 2019-12-31 | 2020-02-11 | 南斗六星系统集成有限公司 | Driver fatigue grade recognition system with bimodal feature fusion |
CN113240885A (en) * | 2021-04-27 | 2021-08-10 | 宁波职业技术学院 | Method for detecting fatigue of vehicle-mounted driver |
CN113643512A (en) * | 2021-07-28 | 2021-11-12 | 北京中交兴路信息科技有限公司 | Fatigue driving detection method and device, electronic equipment and storage medium |
CN113643512B (en) * | 2021-07-28 | 2023-07-18 | 北京中交兴路信息科技有限公司 | Fatigue driving detection method and device, electronic equipment and storage medium |
CN113838265A (en) * | 2021-09-27 | 2021-12-24 | 科大讯飞股份有限公司 | Fatigue driving early warning method and device and electronic equipment |
CN114132326A (en) * | 2021-11-26 | 2022-03-04 | 北京经纬恒润科技股份有限公司 | Method and device for processing fatigue driving |
CN116439710A (en) * | 2023-04-11 | 2023-07-18 | 中国人民解放军海军特色医学中心 | Ship driver fatigue detection system and method based on physiological signals |
CN116439710B (en) * | 2023-04-11 | 2023-10-20 | 中国人民解放军海军特色医学中心 | Ship driver fatigue detection system and method based on physiological signals |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108230619A (en) | Method for detecting fatigue driving based on multi-feature fusion | |
CN104183091B (en) | System for adjusting sensitivity of fatigue driving early warning system in self-adaptive mode | |
CN102509418B (en) | Fatigue driving estimation and early-warning method and device of multi-sensor information fusion | |
CN104637246B (en) | Driver multi-behavior early warning system and danger evaluation method | |
CN104794855B (en) | Driver attention's comprehensive evaluating device | |
CN102717765B (en) | Fatigue driving detection method and anti-fatigue driving auxiliary device | |
CN108446600A (en) | A kind of vehicle driver's fatigue monitoring early warning system and method | |
CN105719431A (en) | Fatigue driving detection system | |
JP6134112B2 (en) | Driving concentration determination method and system for vehicle operation state information base | |
CN104794856B (en) | Driver attention's compositive appraisement system | |
CN107972671A (en) | A kind of driving behavior analysis system | |
CN105303830A (en) | Driving behavior analysis system and analysis method | |
CN107697069A (en) | Fatigue of automobile driver driving intelligent control method | |
CN106080194A (en) | The method for early warning of anti-fatigue-driving and system | |
CN103198616B (en) | Based on method for detecting fatigue driving and the system of the identification of driver's neck moving characteristic | |
TW201321246A (en) | Driving behavior analysis and warning system and method thereof | |
CN106355838A (en) | Fatigue driving detection method and system | |
CN108021875A (en) | A kind of vehicle driver's personalization fatigue monitoring and method for early warning | |
CN102930693A (en) | Early warning system and method for safe driving | |
CN107563346A (en) | One kind realizes that driver fatigue sentences method for distinguishing based on eye image processing | |
CN109740477A (en) | Study in Driver Fatigue State Surveillance System and its fatigue detection method | |
CN105931430A (en) | Alarm sensitivity detection method and apparatus for driver state early warning system | |
CN110901385B (en) | Active speed limiting method based on fatigue state of driver | |
CN106408877A (en) | Rail traffic driver fatigue state monitoring method | |
CN212484555U (en) | Fatigue driving multi-source information detection system |
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 | ||
WW01 | Invention patent application withdrawn after publication | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20180629 |