CN109543655A - Method for detecting fatigue driving, device, computer equipment and storage medium - Google Patents
Method for detecting fatigue driving, device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN109543655A CN109543655A CN201811529598.8A CN201811529598A CN109543655A CN 109543655 A CN109543655 A CN 109543655A CN 201811529598 A CN201811529598 A CN 201811529598A CN 109543655 A CN109543655 A CN 109543655A
- Authority
- CN
- China
- Prior art keywords
- face
- feature information
- face feature
- fatigue
- image
- 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
- 238000000034 method Methods 0.000 title claims abstract description 40
- 238000003860 storage Methods 0.000 title claims abstract description 24
- 238000001514 detection method Methods 0.000 claims description 18
- 230000001815 facial effect Effects 0.000 claims description 18
- 238000000605 extraction Methods 0.000 claims description 12
- 238000013527 convolutional neural network Methods 0.000 claims description 11
- 238000005070 sampling Methods 0.000 claims description 11
- 239000000284 extract Substances 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 8
- 206010016256 fatigue Diseases 0.000 description 93
- 230000009471 action Effects 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 230000033228 biological regulation Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 210000003128 head Anatomy 0.000 description 3
- 230000014759 maintenance of location Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 238000010009 beating Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000000151 deposition Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000008921 facial expression Effects 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 230000029058 respiratory gaseous exchange Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/59—Context or environment of the image inside of a vehicle, e.g. relating to seat occupancy, driver state or inner lighting conditions
- G06V20/597—Recognising the driver's state or behaviour, e.g. attention or drowsiness
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The present invention relates to a kind of method for detecting fatigue driving, device, computer equipment and storage mediums, method includes the face-image for acquiring current driver's, face feature information is extracted from face-image, store the face feature information currently extracted, it compares the face feature information currently extracted and whether the preceding face feature information stored several times is all consistent, if all consistent, determine current driver's for fatigue driving.It is compared face characteristic is carried out with the preceding face feature information stored several times from the characteristic information of current face image zooming-out, judge whether ultra-long time drives or rests in the preset time whether there is or not abundant current driver's, to identify current driver's whether fatigue driving, corresponding early warning can effectively be made, and the operation of driver is not influenced, sufficient safety assurance is carried out for the trip of driver.
Description
Technical field
The present invention relates to fatigue driving detection field, in particular to a kind of method for detecting fatigue driving, device, computer are set
Standby and storage medium.
Background technique
Studies have shown that caused by 80%-90% is human factor in traffic accident, and fatigue driving be then principal element it
One.Wherein, ultra-long time driving is one of the main reason for leading to fatigue driving.Currently when detecting fatigue driving, detection
The DAS2000 type road that method mainly has the doze driver detective instrument, draw drum monitoring arrangement, Japan of U.S.'s development to develop
Face alerts instrument and tester PVT when reacting, and the sensor of the detection in these detection methods is all contact, in practical row
It often easily causes during vehicle driver uncomfortable or influences driver's operation, and can not be compared in continuous driving procedure for a long time
Objectively to detect current driver's, whether violation ultra-long time is driven.
Summary of the invention
Based on this, it is necessary to be easy to influence driver's operation for fatigue driving detection, in continuous driving procedure for a long time
In can not objective detection current driver's the problem of whether violation ultra-long time drives, provide a kind of method for detecting fatigue driving,
Device, computer equipment and storage medium.
A kind of method for detecting fatigue driving, comprising the following steps: acquire and work as from camera shooting and video according to prefixed time interval
The face-image of preceding driver；Face feature information is extracted from face-image, stores the face feature information currently extracted；Than
It is whether all consistent to the face feature information and the preceding face feature information stored several times currently extracted；If all consistent,
Then determine current driver's for fatigue driving.
In one embodiment, described to compare the face feature information currently extracted and the preceding facial characteristics stored several times
After whether information is all consistent, further includes: if the face feature information currently extracted and the preceding any one side stored several times
Portion's characteristic information is inconsistent, then fatigue characteristic information is extracted from current face image；Judging current fatigue characteristic information is
No is more than preset fatigue threshold, if current fatigue characteristic information determines current driver's more than preset fatigue threshold
For fatigue driving.
In one embodiment, the fatigue threshold is preset with multiple, for distinguishing the degree of fatigue of driver.It is described to sentence
Whether the fatigue characteristic information for breaking current is more than preset fatigue threshold, if current fatigue characteristic information is more than preset fatigue
Threshold value, then after determining current driver's for fatigue driving, further includes: according to current fatigue characteristic information, with each threshold in fatigue
Value is reference, judges the degree of fatigue of current driver's, and the degree of fatigue according to current driver's issues corresponding warning.
In one embodiment, in the face for acquiring current driver's from camera shooting and video according to prefixed time interval
After image, the method also includes following steps: carrying out gray processing processing to the driver's face-image currently acquired, obtains
Gray scale face-image, and gray scale face-image is enhanced, expression are as follows:
Wherein, f (x, y) indicates that the gray value of former face-image, g (x, y) indicate the ash of the face-image after grey level enhancement
Angle value, h1 indicate to meetWhen maximum gradation value, h2 indicate meetWhen minimum gradation value, hist [] indicate grey level enhancement rear face figure
The grey value histograms of picture, N indicate total pixel of grey level enhancement rear face image.
Correspondingly, described extract face feature information from face-image, comprising: from the face-image after grey level enhancement
Extract face feature information.
In one embodiment, the face feature information of extracting from face-image includes: passing through convolutional neural networks
Convolutional layer sliding sampling is carried out to an at least pattern portion picture with identical weight parameter；Pass through the secondary of convolutional neural networks
Sample level carries out de-redundancy processing to the data of sliding sampling, obtains face feature information.
In one embodiment, described that face feature information is extracted from face-image, it is special to store the face currently extracted
Reference breath, including being stored by Circular buffer area to the face feature information currently extracted；
It is described compare the face feature information currently extracted and the preceding face feature information stored several times whether all one
Before cause, including the face feature information stored several times before being read from the Circular buffer area.
In one embodiment, the Circular buffer area includes 1 write pointer write_pos, M read pointer read_pos
(i), M is the natural number more than or equal to 1,1≤i≤M；Wherein, read pointer read_pos (1) follow write pointer write_pos it
Afterwards, after read pointer read_pos (i+1) follows read pointer read_pos (i)；The size of the memory space in Circular buffer area is
QUEUE_SIZE, QUEUE_SIZE >=M+1；The face feature information currently extracted is written to by write pointer write_pos
In the corresponding memory space in Circular buffer area；Circular buffer area is arrived into preceding storage several times by M read pointer read_pos (i)
Each face feature information in corresponding memory space is read out one by one.
A kind of fatigue driving detection device, including facial image acquisition module, face feature information extraction module, face are special
Levy information storage module, face feature information comparison module, wherein
The facial image acquisition module is set as several pattern portions of the continuous acquisition current driver's from camera shooting and video
Image.
The face feature information extraction module, is set as extracting face feature information from face-image.
The face feature information memory module is set as storing the face feature information currently extracted.
The face feature information comparison module is set as comparing the face feature information currently extracted and deposit several times with preceding
Whether the face feature information of storage is all consistent, if all consistent, determines current driver's for fatigue driving.
A kind of computer equipment, including memory and processor are stored with computer-readable instruction in the memory, institute
When stating computer-readable instruction and being executed by the processor, so that the processor executes the step of above-mentioned method for detecting fatigue driving
Suddenly.
A kind of storage medium being stored with computer-readable instruction, the computer-readable instruction are handled by one or more
When device executes, so that the step of one or more processors execute above-mentioned method for detecting fatigue driving.
Above-mentioned method for detecting fatigue driving, device, computer equipment and storage medium, by within a preset time to preset
Time interval acquisition driver face-image, from the characteristic information of current face image zooming-out and preceding will store several times
Face feature information is compared, judge current driver's whether ultra-long time drive or whether there is or not abundant in the preset time
Rest, thus identify it is prolonged continuously in driving procedure current driver's whether fatigue driving；In addition, being worked as by identification
The facial fatigue characteristic information of preceding driver, to judge whether driver enters fatigue state；Corresponding early warning can be effectively made to mention
Show, and do not influence the operation of driver, carries out sufficient safety assurance for the trip of driver.
Detailed description of the invention
Fig. 1 is a kind of flow diagram of method for detecting fatigue driving in one embodiment of the invention；
Fig. 2 is the method flow schematic diagram that face feature information is extracted in one embodiment of the invention；
Fig. 3 is a kind of structural schematic diagram of fatigue driving detection device in one embodiment of the invention.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
Those skilled in the art of the present technique are appreciated that unless expressly stated, singular " one " used herein, " one
It is a ", " described " and "the" also may include plural form.It is to be further understood that used in specification of the invention
Wording " comprising " refers to that there are the feature, program, step, operation, element and/or component, but it is not excluded that in the presence of or add
Add other one or more features, program, step, operation, element, component and/or their group.
As shown in Figure 1, a kind of method for detecting fatigue driving, includes the following steps S1-S6:
S1, acquisition face-image: the face-image of current driver's is acquired from camera shooting and video according to prefixed time interval.
In any position (such as steering wheel or front-seat glass) peace for monitoring driver's facial expression of automobile
A photographic device is filled, the face-image of driver is obtained by the photographic device.
In one embodiment, step S1, which is also comprised the steps of:, carries out ash to the driver's face-image currently acquired
Degreeization processing, obtains gray scale face-image, and enhance gray scale face-image, specifically uses following formula:
Wherein, f (x, y) indicates that the gray value of former face-image, g (x, y) indicate the ash of the face-image after grey level enhancement
Angle value,
H1 indicates to meetWhen maximum gradation value, h2 indicate meetWhen minimum gradation value, hist [] indicate grey level enhancement rear face figure
The grey value histograms of picture, N indicate total pixel of grey level enhancement rear face image.Enhanced face-image is for extracting face
Characteristic information.
S2, extraction and storage face feature information: extracting face feature information from face-image, what storage was currently extracted
Face feature information.
As shown in Fig. 2, in one embodiment, extraction face feature information includes from face-image described in step S2
Following steps:
S21: by the volume base of convolutional neural networks (CNN) with identical weight parameter to an at least pattern portion picture into
Row sliding sampling；
The convolution window of default 5*5 width, through convolution window since the initial position of facial picture, gradually to facial picture
Data are traversed, and are sampled to facial picture.
S22: de-redundancy processing is carried out to the data of sliding sampling by the double sampling layer of convolutional neural networks, obtains face
Portion's characteristic information.
The convolutional neural networks extract the different characteristic of picture by convolutional layer, by double sampling layer to extracting
Feature is sampled, with removal since there are the letters of spatial redundancy caused by stronger correlation between adjacent pixel inside image
It ceases (may include multiple convolutional layers, double sampling layer in a convolutional neural networks).
The present embodiment uses convolutional neural networks, carries out facial feature extraction to the facial picture of acquisition, then by extraction
Each face feature information, which is together in series, constitutes final complete sub-pictures feature, is used for face recognition.It is same in convolutional neural networks
Weight parameter in one layer can be shared, because each different zones of same picture have certain similitude, solve in this way
It has determined full connection calculation amount problems of too.
S3, compare face feature information: comparing the face feature information currently extracted and the preceding face stored several times is special
Whether reference breath is all consistent, if it has, then determining that current driver's are fatigue driving, and executes S6；If it has not, then executing
S4。
By the way that the face feature information of the driver currently acquired and the preceding face feature information stored several times are carried out
It compares, judges whether the driver in the face-image acquired in a period of time is same people with this." traffic safety law implements item
Example " the 62nd article of regulation, vehicle driver must not continuous operating motor vehicles be more than rest of not stopping for 4 hours, or stop and rest
Time is less than 20 minutes.That is, by judging whether the driver in the face-image acquired in 4 hours is same
People, can identify in time driver whether fatigue driving.
Therefore, the detection maximum time interval of each fatigue driving is arranged within 4 hours than convenient.With front and back two
The detection time of secondary fatigue driving was set to for 4 hours, compared the face feature information currently extracted and 4 hours
Whether the face feature information of preceding extraction is all consistent, if the face feature information that front and back is extracted is consistent, assert small at 4
When interior vehicle driver be same people, current driver's are fatigue driving.In this way, although the number of fatigue detecting is few,
Be to be easy to generate such problems: current driver and the driver at moment before 4 hours are same driver, but period
The driver is likely to drive in turn with other people, easily causes erroneous judgement when detecting in this way.It is therefore desirable to reasonably be arranged
The detection time interval of front and back fatigue driving twice.
When can be seen that the detection of front and back fatigue driving twice according to " traffic safety law implementing regulations " the 62nd article regulation
Between be arranged at intervals on 20 minutes or within it is the most accurate, both can detecte out whether current driver's continuously drive more than 4 in this way
A hour can also identify that the driver is rested every time whether more than 20 in the case where current driver's have and drive in turn
Minute.By taking the detection time of front and back fatigue driving twice is set to 20 minutes as an example: then needing to detect 13 times in 4 hours
(the 13rd time is when previous detection).In this way, the face feature information of the face feature information currently extracted and first 12 times storages
It is compared one by one；If each comparison result is consistent, very big probability thinks that current driver's driving has been more than 4 small
When；Once once comparison result is different, then illustrate that driver has rest within 4 hours, and the time of having a rest is more than 20 points
Clock.Certainly, it also has days off by turns a few minutes, a possibility that the sum of time repeatedly rested was more than 20 minutes, will lead to nothing in this way
Whether method accurate judgement current driver's rested more than 20 minutes.It is apparent, however, that saying from convention, this possibility is very
It is small, even across repeatedly having days off by turns more than 20 minutes, it is also difficult to relieve fatigue, therefore can ignore.
In one embodiment, the detection time of front and back fatigue driving twice is set to 1 hour, in this way, at 4
In hour, the face feature information that the face feature information currently extracted is stored with first 4 times is compared one by one, identifies driver
Whether fatigue driving.In this way, can both reduce operand, the time of the driving and rest that are also more conform with driver is accustomed to.
It is of course also possible to according to whether being that same driver drives within the fatigable degree setting identification how long of driver
Vehicle is sailed, such as is set as in 2 hours, it is primary every 20 minutes fatigue detectings, then it needs to detect 7 times the (the 7th in 2 hours
It is secondary for when previous detection), in this way, the face feature informations of the face feature information currently extracted and first 6 times storages one by one into
Row compares；If each comparison result is consistent, very big probability thinks that current driver's driving has been more than 2 hours.
S4, it extracts fatigue characteristic information: extracting fatigue characteristic information from current face image.
Certainly, only in accordance with whether abiding by legal driving time judges whether fatigue driving is not sufficient to current driver's
Judge whether current driver's are tired, therefore, the application also changes from the face action of driver to identify that driver is abiding by
Whether there is fatigue driving in the case where legal driving time.
By the contour detecting of face, the organ sites on head in continuous each face-image are identified, such as eyes, mouth
Deng head capable of being easily identified out the information at the position of fatigue state as interested fatigue characteristic information, such as blink
Time, amplitude of nodding play movement etc. of breathing out.
Judgement using the specific region amplitude between continuous face-image, as identification driver's face action variation
Foundation is based on the width specifically, seeking the amplitude of the specific region of the continuous interframe of face-image, and predetermined amplitude threshold value
Degree and amplitude threshold identify the variation of driver's face action.For example, the nodding action that successive frame face-image is shown
When amplitude variation is more than preset first amplitude threshold, then it is considered as driver and is dozing off, otherwise, be then considered as regular event.Even
When the amplitude variation of the continuous frame face-image mouth shown and facial action is more than preset second amplitude threshold, then it is considered as driving
Member breathes out beating, and otherwise, is then considered as regular event.And to the duration of some movements, counted, sequential frame image it
Between sampling interval duration be it is the same, determine it is a certain movement occur in the face-image of how many frame, also just have identified this
The duration of movement, for example, being counted using this method to a wink time.
S5, judge fatigue characteristic information: judging whether current fatigue characteristic information is more than preset fatigue threshold；If
It is then to determine that current driver's for fatigue driving, and execute S6.
Fatigue threshold can set multiple grades, preset fatigue driving evaluation criterion, according to the micro- expression of face and head
Portion's movement, is configured fatigue level, as shown in table 1 below, the fatigue threshold provided with three-level.According to the current tired of acquisition
Labor characteristic information is reference with each fatigue threshold, judges the degree of fatigue of current driver's.
| Fatigue level | The driving condition of driver |
| Level-one | Nodding action frequency and amplitude are more than threshold value respectively |
| Second level | Single closed-eye time is more than normal wink time (0.5 second) |
| Three-level | It closes one's eyes (eye-closing period is more than one second) and frequently nods for a long time |
Table 1
S6, warning fatigue driving: fatigue driving warning is issued.
When determining current driver's for fatigue driving, early warning movement is made in time.Can foundation, level of fatigue, to driving
The person of sailing or other related personnel issue prompt, such as shown in table 2:
Table 2
In one embodiment, described that face feature information is extracted from face-image, it is special to store the face currently extracted
Reference breath, including being stored by Circular buffer area to the face feature information currently extracted；
It is described compare the face feature information currently extracted and the preceding face feature information stored several times whether all one
Before cause, including the face feature information stored several times before being read from the Circular buffer area.
Circular buffer area is a kind of end to end queue data structure, it then follows first in first out.Circular buffer area
Write-in process can be buffered with the process of reading using the model of " producer and consumer ", thus facilitate the use of caching with
Management.The amount of capacity in Circular buffer area can be arranged according to the size for uploading data.
The data from queue heads to rear of queue are successively stored in Circular buffer area with the storage unit of one group of continuation address, setting
Write pointer write_pos, read pointer read_pos, are respectively directed to writing position and reading by write_pos and read_pos
Position.
In the application, Circular buffer area needs to design 1 write pointer write_pos, M read pointer read_pos (i), M
For the natural number more than or equal to 1,1≤i≤M.Wherein, after read pointer read_pos (1) follows write pointer write_pos, together
Reason, read pointer read_pos (i+1) are followed after read pointer read_pos (i).The size of Circular buffer area memory space is
QUEUE_SIZE, QUEUE_SIZE >=M+1.
When initializing queue, write_pos=M, read_pos (i)=M-i are enabled.When have data be written Circular buffer area
When, write pointer write_pos increases 1；When there is data to read from Circular buffer area, read pointer read_pos (i) increases 1.
That is, will currently extract face feature information storage to write pointer after currently extracting face feature information
In Circular buffer area pointed by write_pos, write pointer write_pos increases 1；Then, each read pointer read_ is read respectively
Extraction face feature information in Circular buffer area pointed by pos (i), i.e., the face feature information stored first M times are distinguished
It reads out, each read pointer read_pos (i) increases 1 respectively；Then the face feature information and first M times storage currently extracted are compared
Face feature information it is whether all consistent.
Above-described embodiment acquires the face-image of driver at predetermined intervals within a preset time, will be from current
The characteristic information that face-image extracts is compared with the preceding face feature information stored several times, whether judges current driver's
Ultra-long time drives or whether there is or not abundant rests in the preset time, thus identify current driver's whether fatigue driving；
In addition, by the facial fatigue characteristic information of identification current driver's, to judge whether driver enters fatigue state；It can be effective
Corresponding early warning is made, carries out sufficient safety assurance for the trip of driver.
As shown in figure 3, in one embodiment it is proposed that a kind of fatigue driving detection device, including facial image acquisition
Module, face feature information is extracted and memory module, face feature information comparison module, fatigue characteristic information extraction modules, tired
Labor characteristic information judgment module, fatigue driving alarm module.
The facial image acquisition module is set as several pattern portions of the continuous acquisition current driver's from camera shooting and video
Image；
The face feature information extraction module, is set as extracting face feature information from face-image；
The face feature information memory module is set as storing the face feature information currently extracted；
The face feature information comparison module is set as comparing the face feature information currently extracted and deposit several times with preceding
Whether the face feature information of storage is all consistent, if all consistent, determines current driver's for fatigue driving；
The fatigue characteristic information extraction modules are set as storing several times in the face feature information currently extracted with preceding
Any one face feature information it is inconsistent in the state of, from current face image extract fatigue characteristic information；
The fatigue characteristic signal judgement module is set as judging whether current fatigue characteristic information is more than preset tired
Labor threshold value；If current fatigue characteristic information is more than preset fatigue threshold, determine current driver's for fatigue driving
The fatigue driving alarm module is set as issuing tired in the state of determining current driver's for fatigue driving
Please warning is sailed.
Above-described embodiment acquires the face-image of driver at predetermined intervals within a preset time, will be from current
The characteristic information that face-image extracts is compared with the preceding face feature information stored several times, whether judges current driver's
Ultra-long time drives or whether there is or not abundant rests in the preset time, thus identify current driver's whether fatigue driving；
In addition, by the facial fatigue characteristic information of identification current driver's, to judge whether driver enters fatigue state；It can be effective
Corresponding early warning is made, carries out sufficient safety assurance for the trip of driver.
Based on the same technical idea, the invention also provides a kind of computer equipment, the computer equipment includes depositing
Reservoir, processor and it is stored in the computer-readable instruction that can be run on the memory and on the processor, the place
The step of reason device executes the method for detecting fatigue driving in the various embodiments described above when executing the computer-readable instruction.
Based on the same technical idea, the present invention also provides a kind of storage medium for being stored with computer-readable instruction,
When the computer-readable instruction is executed by one or more processors, so that one or more processors execute above-mentioned each implementation
Example in the method for detecting fatigue driving the step of.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, which can be stored in a computer-readable storage and be situated between
In matter, the program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein, storage medium above-mentioned can be
The non-volatile memory mediums such as magnetic disk, CD, read-only memory (Read-Only Memory, ROM) or random storage note
Recall body (Random Access Memory, RAM) etc..
Each technical characteristic of embodiment described above can be combined arbitrarily, for simplicity of description, not to above-mentioned reality
It applies all possible combination of each technical characteristic in example to be all described, as long as however, the combination of these technical characteristics is not deposited
In contradiction, all should be considered as described in this specification.
The embodiments described above only express several embodiments of the present invention, and the description thereof is more specific and detailed, but simultaneously
Limitations on the scope of the patent of the present invention therefore cannot be interpreted as.It should be pointed out that for those of ordinary skill in the art
For, without departing from the inventive concept of the premise, various modifications and improvements can be made, these belong to guarantor of the invention
Protect range.Therefore, the scope of protection of the patent of the invention shall be subject to the appended claims.
Claims (10)
1. a kind of method for detecting fatigue driving, which comprises the following steps:
The face-image of current driver's is acquired from camera shooting and video according to prefixed time interval；
Face feature information is extracted from face-image, stores the face feature information currently extracted；Compare the face currently extracted
Whether portion's characteristic information and the preceding face feature information stored several times are all consistent, if all consistent, determine current drive
Member is fatigue driving.
2. method for detecting fatigue driving as described in claim 1, which is characterized in that described to compare the facial characteristics currently extracted
After whether information and the preceding face feature information stored several times are all consistent, further includes:
If the face feature information currently extracted and the preceding any one face feature information stored several times are inconsistent, from current
Fatigue characteristic information is extracted in face-image；
Judge whether current fatigue characteristic information is more than preset fatigue threshold, if current fatigue characteristic information is more than default
Fatigue threshold, then determine current driver's for fatigue driving.
3. method for detecting fatigue driving as claimed in claim 2, which is characterized in that
The fatigue threshold be preset with it is multiple, for distinguishing the degree of fatigue of driver；
It is described to judge whether current fatigue characteristic information is more than preset fatigue threshold, if current fatigue characteristic information is more than
Preset fatigue threshold, then after determining current driver's for fatigue driving, further includes:
It is reference with each fatigue threshold according to current fatigue characteristic information, judges the degree of fatigue of current driver's, foundation is worked as
The degree of fatigue of preceding driver issues corresponding warning.
4. method for detecting fatigue driving as claimed in claim 1 or 2, which is characterized in that described according to prefixed time interval
After the face-image for acquiring current driver's in camera shooting and video, the method also includes following steps:
Gray processing processing is carried out to the face-image of the driver currently acquired, obtains gray scale face-image, and to gray scale face
Image is enhanced, expression are as follows:
Wherein, f (x, y) indicates that the gray value of former face-image, g (x, y) indicate the gray value of the face-image after grey level enhancement,
H1 indicates to meetWhen maximum gradation value, h2 indicate meetWhen minimum gradation value, hist [] indicate grey level enhancement rear face figure
The grey value histograms of picture, N indicate total pixel of grey level enhancement rear face image；
Correspondingly, described extract face feature information from face-image, comprising: extracted from the face-image after grey level enhancement
Face feature information.
5. method for detecting fatigue driving as claimed in claim 1 or 2, which is characterized in that described to extract face from face-image
Portion's characteristic information, comprising:
Sliding sampling is carried out to an at least pattern portion picture with identical weight parameter by the volume base of convolutional neural networks；
De-redundancy processing is carried out to the data of sliding sampling by the double sampling layer of convolutional neural networks, obtains facial characteristics letter
Breath.
6. method for detecting fatigue driving as claimed in claim 1 or 2, which is characterized in that described to store the face currently extracted
Characteristic information, including being stored by Circular buffer area to the face feature information currently extracted；
It is described compare the face feature information currently extracted and the preceding face feature information stored several times whether all it is consistent it
Before, including the face feature information stored several times before being read from the Circular buffer area.
7. method for detecting fatigue driving as claimed in claim 6, which is characterized in that
The Circular buffer area includes 1 write pointer write_pos, and M read pointer read_pos (i), M is more than or equal to 1
Natural number, 1≤i≤M, wherein after read pointer read_pos (1) follows write pointer write_pos, read pointer read_pos (i
+ 1) after following read pointer read_pos (i)；The size of the memory space in Circular buffer area is QUEUE_SIZE, QUEUE_
SIZE≥M+1；
The face feature information currently extracted is written to the corresponding memory space in Circular buffer area by write pointer write_pos
It is interior；
Each face in the corresponding memory space in Circular buffer area is arrived into preceding storage several times by M read pointer read_pos (i)
Portion's characteristic information is read out one by one.
8. a kind of fatigue driving detection device characterized by comprising facial image acquisition module, face feature information are extracted
Module, face feature information memory module, face feature information comparison module, wherein
The facial image acquisition module is set as several pattern portions figure of the continuous acquisition current driver's from camera shooting and video
Picture；
The face feature information extraction module, is set as extracting face feature information from face-image；
The face feature information memory module is set as storing the face feature information currently extracted；
The face feature information comparison module is set as comparing the face feature information currently extracted and preceding stores several times
Whether face feature information is all consistent, if all consistent, determines current driver's for fatigue driving.
9. a kind of computer equipment, including memory and processor, it is stored with computer-readable instruction in the memory, it is described
When computer-readable instruction is executed by the processor, so that the processor executes such as any one of claims 1 to 7 right
It is required that the step of method for detecting fatigue driving.
10. a kind of storage medium for being stored with computer-readable instruction, the computer-readable instruction is handled by one or more
When device executes, so that one or more processors execute the fatigue driving as described in any one of claims 1 to 7 claim and examine
The step of survey method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811529598.8A CN109543655A (en) | 2018-12-14 | 2018-12-14 | Method for detecting fatigue driving, device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811529598.8A CN109543655A (en) | 2018-12-14 | 2018-12-14 | Method for detecting fatigue driving, device, computer equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109543655A true CN109543655A (en) | 2019-03-29 |
Family
ID=65855294
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811529598.8A Pending CN109543655A (en) | 2018-12-14 | 2018-12-14 | Method for detecting fatigue driving, device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109543655A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110246304A (en) * | 2019-06-27 | 2019-09-17 | 北京三一智造科技有限公司 | Tired system for prompting and method |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2013235341A (en) * | 2012-05-07 | 2013-11-21 | Nikon Corp | Fatigue confirmation device |
| CN104598934A (en) * | 2014-12-17 | 2015-05-06 | 安徽清新互联信息科技有限公司 | Monitoring method for smoking behavior of driver |
| CN105788176A (en) * | 2016-05-25 | 2016-07-20 | 厦门理工学院 | Fatigue driving monitoring and prompting method and system |
| CN106314149A (en) * | 2016-08-25 | 2017-01-11 | 乐视控股（北京）有限公司 | Method and device for controlling fatigue driving |
| CN106415594A (en) * | 2014-06-16 | 2017-02-15 | 北京市商汤科技开发有限公司 | A method and a system for face verification |
| CN107644209A (en) * | 2017-09-21 | 2018-01-30 | 百度在线网络技术（北京）有限公司 | Method for detecting human face and device |
| CN107697069A (en) * | 2017-10-31 | 2018-02-16 | 上海汽车集团股份有限公司 | Fatigue of automobile driver driving intelligent control method |
| CN108294759A (en) * | 2017-01-13 | 2018-07-20 | 天津工业大学 | A kind of Driver Fatigue Detection based on CNN Eye state recognitions |
| CN108932207A (en) * | 2017-05-23 | 2018-12-04 | 珠海全志科技股份有限公司 | SDIO-WIFI data transmission method and system with buffer area |
-
2018
- 2018-12-14 CN CN201811529598.8A patent/CN109543655A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2013235341A (en) * | 2012-05-07 | 2013-11-21 | Nikon Corp | Fatigue confirmation device |
| CN106415594A (en) * | 2014-06-16 | 2017-02-15 | 北京市商汤科技开发有限公司 | A method and a system for face verification |
| CN104598934A (en) * | 2014-12-17 | 2015-05-06 | 安徽清新互联信息科技有限公司 | Monitoring method for smoking behavior of driver |
| CN105788176A (en) * | 2016-05-25 | 2016-07-20 | 厦门理工学院 | Fatigue driving monitoring and prompting method and system |
| CN106314149A (en) * | 2016-08-25 | 2017-01-11 | 乐视控股（北京）有限公司 | Method and device for controlling fatigue driving |
| CN108294759A (en) * | 2017-01-13 | 2018-07-20 | 天津工业大学 | A kind of Driver Fatigue Detection based on CNN Eye state recognitions |
| CN108932207A (en) * | 2017-05-23 | 2018-12-04 | 珠海全志科技股份有限公司 | SDIO-WIFI data transmission method and system with buffer area |
| CN107644209A (en) * | 2017-09-21 | 2018-01-30 | 百度在线网络技术（北京）有限公司 | Method for detecting human face and device |
| CN107697069A (en) * | 2017-10-31 | 2018-02-16 | 上海汽车集团股份有限公司 | Fatigue of automobile driver driving intelligent control method |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110246304A (en) * | 2019-06-27 | 2019-09-17 | 北京三一智造科技有限公司 | Tired system for prompting and method |
| CN110246304B (en) * | 2019-06-27 | 2023-08-15 | 北京三一智造科技有限公司 | Fatigue reminding system and method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11783601B2 (en) | Driver fatigue detection method and system based on combining a pseudo-3D convolutional neural network and an attention mechanism | |
| Sigari et al. | A driver face monitoring system for fatigue and distraction detection | |
| CN108309311A (en) | A kind of real-time doze of train driver sleeps detection device and detection algorithm | |
| You et al. | A fatigue driving detection algorithm based on facial motion information entropy | |
| CN103700220A (en) | Fatigue driving monitoring device | |
| Milan | DRIVER’S DROWSINESS DETECTION | |
| CN115393830A (en) | Fatigue driving detection method based on deep learning and facial features | |
| Biju et al. | Drowsy driver detection using two stage convolutional neural networks | |
| Kavitha et al. | A novel approach for driver drowsiness detection using deep learning | |
| Neeraja et al. | DL-Based Somnolence Detection for Improved Driver Safety and Alertness Monitoring | |
| CN109543655A (en) | Method for detecting fatigue driving, device, computer equipment and storage medium | |
| Mašanović et al. | Driver monitoring using the in-vehicle camera | |
| Priyanka et al. | A Review on Drowsiness Prediction System using Deep Learning Approaches | |
| Shaykha et al. | FEER: Non-intrusive facial expression and emotional recognition for driver's vigilance monitoring | |
| Abdullah et al. | Driver fatigue detection | |
| Arora et al. | Real time safety alert system for car | |
| CN202875355U (en) | Real-time monitoring device for mental state of driver | |
| Singh et al. | Real-time Driver Drowsiness Detection System using Cascaded ConvNet Framework | |
| Adarsh et al. | Drowsiness Detection System in Real Time Based on Behavioral Characteristics of Driver using Machine Learning Approach | |
| Dalve et al. | Real Time Prevention of Driver Fatigue Using Deep Learning and MediaPipe | |
| Kumral et al. | Temporal Analysis Based Driver Drowsiness Detection System Using Deep Learning Approaches | |
| Mohan et al. | Identification of Driver Drowsiness Detection using a Regularized Extreme Learning Machine | |
| Priya et al. | Machine Learning-Based System for Detecting and Tracking Driver Drowsiness | |
| Shang et al. | A Novel Method for Unsafe Driving States Detection | |
| Wijaya et al. | A Fatigue Detection Model Based on Convolutional Neural Network |
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 |