CN106448050B - A kind of smart mobile phone traffic hazard method for early warning and system based on vehicle noise - Google Patents
A kind of smart mobile phone traffic hazard method for early warning and system based on vehicle noise Download PDFInfo
- Publication number
- CN106448050B CN106448050B CN201611005062.7A CN201611005062A CN106448050B CN 106448050 B CN106448050 B CN 106448050B CN 201611005062 A CN201611005062 A CN 201611005062A CN 106448050 B CN106448050 B CN 106448050B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- mobile phone
- vehicle noise
- pedestrian
- data
- 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.)
- Active
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
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
- G10L25/51—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination
Abstract
The present invention relates to a kind of smart mobile phone traffic hazard method for early warning and system based on vehicle noise, it is characterized in that, the system includes data acquisition module based on smart mobile phone, characteristic extracting module, central processing unit, background data base, control module and warning module, in background data base matching relationship and model are established using the training sample of acquisition, pass through the real-time collection vehicle noise data group of smart mobile phone, feature extraction is carried out to vehicle noise data group, call the matching relationship stored in background data base and model output result, and result is analyzed and determined, it controls smart mobile phone and triggers first early warning and urgent early warning, reminding in time has dysaudia, it sees the mobile phone while walking roadside and wallowing in, the trip crowd of mobile phone music is listened with earphone, improve the safety of road traveler, effectively reduce the accident rate on road.
Description
Technical field
The present invention relates to traffic safety and field of acoustics, more particularly to a kind of smart mobile phone traffic danger based on vehicle noise
Dangerous method for early warning and system.
Background technology
Nowadays, also there is potential traffic safety hidden danger while bringing great convenience in smart mobile phone,
The phenomenon that mobile phone is listened in side with earphone is walked or cycle in road top to be seen everywhere, and is seen the mobile phone because wallowing on the road or is listened with earphone
Mobile phone and the traffic accident caused by road vehicle situation can not be noticed often by media report；In addition, in non-mandrel roller
Intersection or crossing at, some have the special population of dysaudia because self disorder can not predict intersection or people in advance
Whether row lateral road blind area has arrival vehicle, and is easy to be ignored by driver to lead to traffic accident.For it is existing these
Traffic safety hidden danger, if it is possible to remind pedestrian to make emergency danger-avoiding to arrival vehicle in advance, so that it may effectively to reduce on road
Accident rate, protect special population, improve safe trip rate.
Social intelligence's mobile phone has become the indispensable communication of society and converter tools now, has spread to people
The every aspect lived, and smart mobile phone itself is already installed with various inductors and hardware device, using smart mobile phone as
It is a kind of that the tool to give warning in advance is made to arrival vehicle, for improve safety be undoubtedly a kind of very feasible and facility can
The method and measure leaned on.
The present invention proposes a kind of smart mobile phone traffic hazard method for early warning and system based on vehicle noise, and vehicle is in road
On the noise that sends out when driving it is obvious and be easy to differentiate, using vehicle noise remind pedestrian be undoubtedly it is a kind of conveniently
Mode.
Invention content
Based on background above, it is an object of the invention to protect the row for wallowing on road and seeing the mobile phone or listening mobile phone with earphone
People and specific group reduce road accident rate, achieve the purpose that the pedestrian that gives warning in advance.
The present invention proposes a kind of smart mobile phone traffic hazard early warning system based on vehicle noise, which is characterized in that the intelligence
Energy mobile phone traffic hazard early warning system includes acquisition module, characteristic extracting module, central processing unit, background data base, control
The input terminal of module and warning module, the characteristic extracting module connects acquisition module, and output end connects central processing unit,
The central processing unit output end is connect with control module, and is mutually communicated with background data base, the control mould
Block is responsible for receiving the control instruction of central processing unit, and controls the working condition of warning module.
The smart mobile phone traffic hazard early warning system is the application developed on intelligent mobile phone platform as one
Program APP is present in mobile phone main interface.
The acquisition module, including it is placed in the microphone inside smart mobile phone, it is responsible for incuding in real time and acquiring surrounding
Vehicle noise data, the vehicle noise data include vehicle noise sound intensity data and audio data；The feature extraction
Module is responsible for extracting vehicle noise feature from vehicle noise array of data, and is transmitted to central processing unit, described
Vehicle noise feature includes vehicle noise sound intensity variation characteristic and audible spectrum feature；
The central processing unit is the decision package of whole system, is communicated with background data base, is responsible for basis
Vehicle noise feature calls the matching relationship and model stored in background data base, and current vehicle noise number is inputted into model
According to output is as a result, analyzed and determined, and send out instruction to control module；
There are pre-set control program in the control module, inside, are responsible for receiving the instruction of central processing unit,
Control the working condition of prior-warning device；
The warning module, that is, the usual hardware response apparatus being placed in inside smart mobile phone, including mobile phone screen, voice
Equipment and vibratory equipment are responsible for that smart phone user is reminded to avoid vehicle in time.
Steps are as follows for the method for building up of the matching relationship and model that are stored in the background data base：
B1. training sample is acquired：The training sample of acquisition is stored in background data base, and training sample includes vehicle noise
Data, type of vehicle, speed, vehicle and pedestrian's spacing and vehicle and pedestrian's relative distance relationship；The type of vehicle packet
Include large car, in-between car and compact car three types；The vehicle and pedestrian's relative distance relationship include vehicle approaching row
People and sail out of two kinds of relative distance relationships of pedestrian；
B2. vehicle noise feature is extracted：Vehicle approaching pedestrian and two kinds of pedestrian is being sailed out of respectively, is analyzing and extract
Go out three kinds of type of vehicle with different speeds common vehicle noise sound intensity variation characteristic when driving, in approaching pedestrian, point
Analyse and extract respectively from vehicle noise audible spectrum feature common under the corresponding different speed of operation of three types vehicle；
B3. matching relationship is established：The matching relationship includes two kinds of matching relationships, i.e., the first is the vehicle noise sound intensity
Matching relationship between variation characteristic and vehicle and pedestrian's relative distance relationship, for according to vehicle noise sound intensity variation characteristic come
Judge that vehicle is approaching pedestrian or sails out of pedestrian, is for second between vehicle noise audible spectrum feature and type of vehicle
With relationship, for going out type of vehicle according to vehicle noise audible spectrum characteristic matching；
B4. model is established：Do not consider the case where vehicle sails out of pedestrian, is only established and three in the case of vehicle approaching pedestrian
Kind type of vehicle distinguishes the artificial nerve network model of corresponding vehicle noise data and speed, vehicle and pedestrian's spacing, and
A kind of type of vehicle corresponds to a kind of execution program of artificial nerve network model；The artificial nerve network model makes an uproar to vehicle
Sound data and speed, vehicle and pedestrian's spacing carry out concurrent collaborative processing, by adjusting mutual between internal node in network
Connection relation finds the corresponding relation between vehicle noise data and speed, vehicle and pedestrian's spacing；The artificial neural network
Network model uses the BP neural network model of simple-type, and input layer is vehicle noise data, and output layer is speed, vehicle and row
The human world away from.
The storage form of the training sample of the acquisition is：Two storages are established according to vehicle and pedestrian's relative distance relationship
File is deposited, then establishes the subfile of three type of vehicle in the store files of vehicle and pedestrian's relative distance relationship respectively,
By above-mentioned speed sample value, vehicle with pedestrian's spacing sample value and collected corresponding vehicle noise data with classification matrix
Form is stored in subfile.
Matching relationship between the vehicle noise sound intensity variation characteristic and vehicle and pedestrian's relative position relation is：Vehicle
The variation of the noise sound intensity sails out of pedestrian with vehicle in the feature of reduction trend and matches, in the feature and vehicle approaching of raising trend
Pedestrian matches.
A kind of smart mobile phone traffic hazard method for early warning based on vehicle noise, steps are as follows for this method：
S1. the vehicle noise data in ambient enviroment are acquired in real time using smart mobile phone microphone, constitute vehicle noise number
According to array, the vehicle noise array of data includes vehicle noise sound intensity array of data and audio data array；
S2. vehicle noise feature extraction, the vehicle noise feature are carried out to collected vehicle noise array of data
Including vehicle noise sound intensity variation characteristic and audible spectrum feature；
S3. according to vehicle noise sound intensity variation characteristic, call the vehicle noise sound intensity variation stored in background data base special
Vehicle of seeking peace match obtaining a result with the matching relationship of pedestrian's relative distance relationship；
If S4. matching result is that vehicle is in approaching pedestrian's state, the first early warning of smart mobile phone is triggered immediately, it is no
Then, smart mobile phone is still in real-time acquisition state；
S5. under the premise of smart mobile phone has triggered first early warning, number of units after being called according to vehicle noise audible spectrum feature
According to the matching relationship of the vehicle noise audible spectrum feature and type of vehicle that are stored in library, current vehicle type is matched, to
Call the artificial nerve network model corresponding to the type of vehicle；
S6. the last one data component is extracted from vehicle noise sound intensity array of data and audio data array respectively, i.e.,
Current vehicle noise data, are input in artificial nerve network model, output current vehicle speed, vehicle and pedestrian's spacing；
S7. distance threshold is determined by current vehicle speed, the distance threshold refer to vehicle braking distance, driver reaction away from
From and system response sum of the distance；
S8. compare the size of current vehicle and pedestrian's spacing and distance threshold, if current vehicle is smaller than with pedestrian
In distance threshold, then the urgent early warning of smart mobile phone is triggered immediately, otherwise, still in the real-time acquisition state of vehicle noise data.
The vehicle noise feature extracting method includes as follows：
S21. the vehicle noise data in smart mobile phone microphone timely collection ambient enviroment constitute vehicle noise number
According to array, the vehicle noise array of data includes vehicle noise sound intensity array of data and audio data array；
S22. sound intensity data value discrete in sound intensity array of data is done into continuous treatment, obtains function curve variation diagram,
Extract plots changes feature, as sound intensity variation characteristic；
S23. vehicle noise audio data array is done into Fourier transformation and obtains noise spectrum, then extract noise audio frequency
Frequency, amplitude and the phase property of spectrum, as audible spectrum feature.
The first early warning is carried to the vibration and voice of mobile phone user after smart mobile phone detects vehicle noise
It wakes up, which does not carry out any interference to smart mobile phone other applications and control.
The urgent early warning is after vehicle enters threshold value with pedestrian's spacing, and system control smart mobile phone carries out vibration and carries
It wakes up, interrupts mobile phone current task, automatically switches mobile phone screen and the mandatory insertion voice prompt into earphone jack, work as danger
After elimination, then original working condition is responded immediately to.
Based on above technical scheme, the present invention using the smart mobile phone of vehicle noise give warning in advance method and system have with
Lower advantage：
The present invention need not install other hardware devices on smart mobile phone or elsewhere, and cost is few and convenient for taking
Band.
The present invention makes vehicle by acquiring the method that training sample data establish two kinds of matching relationships and BP neural network model
Noise is changed into the alerting signal of smart mobile phone, plays mobile phone personnel or special population is provided and given warning in advance to wallow in, more has
Profit protects out administrative staff, reduces accident rate.
The method of the present invention and system are a popularizations of the application and development to smart mobile phone, before having more wide development
Scape.
Description of the drawings
Fig. 1 is the smart mobile phone traffic hazard early warning system schematic diagram based on vehicle noise of the present invention；
Fig. 2 be the present invention background data base in the flow chart of matching relationship and method for establishing model that stores；
Fig. 3 is the job network schematic diagram of the BP neural network model employed in the present invention；
Fig. 4 is the flow chart of the smart mobile phone traffic hazard method for early warning based on vehicle noise of the present invention；
Fig. 5 is the flow chart of the vehicle noise feature extracting method of the present invention.
Specific implementation mode
The present invention is described in further detail below in conjunction with the accompanying drawings.
As shown in Figure 1, being the schematic diagram of the smart mobile phone traffic hazard early warning system the present invention is based on vehicle noise：
A kind of smart mobile phone traffic hazard early warning system based on vehicle noise of the present invention, including acquisition module (10),
Characteristic extracting module (20), central processing unit (30), background data base (40), control module (50) and warning module (60),
The input terminal connection acquisition module (10) of the characteristic extracting module (20), output end connect central processing unit (30), institute
Central processing unit (30) output end stated is connect with control module (50), and is mutually communicated with background data base (40), institute
The control module (50) stated is responsible for receiving the control instruction of central processing unit (30), and controls the working condition of warning module.
A kind of smart mobile phone traffic hazard early warning system based on vehicle noise of the present invention can be developed as one
Application program be present in intelligent mobile phone platform, with common hardware device cooperating in smart mobile phone, user is reached
The effect of danger early warning, inside include module it is as follows：
Acquisition module (10) is responsible for incuding and acquire surrounding vehicles noise data, including vehicle noise sound intensity data in real time
And audio data can also be voluntarily arranged by user wherein the frequency acquired can be arranged with default system, acquire m respectively every time
A vehicle noise sound intensity data and audio data constitute vehicle noise array of data, and wherein vehicle noise array of data includes vehicle
Noise sound intensity array of data and audio data array；
Characteristic extracting module (20) is responsible for extracting vehicle noise feature, the vehicle from vehicle noise array of data
Feature of noise includes noise sound intensity variation characteristic and audible spectrum feature, and is transmitted to central processing unit (30)；
Central processing unit (30) is the decision package of whole system, is communicated with background data base (40), and root is responsible for
It obtains a result according to the matching relationship and model that are stored in vehicle noise feature and vehicle noise data call background data base (40)
It is analyzed and determined, and instruction is sent out to control module (50) according to judging result.Wherein, changed according to the vehicle noise sound intensity special
It levies and matches with the relative distance relationship of the vehicle and pedestrian that are stored in background data base (40), judge that vehicle is approaching or sails
From pedestrian, if the nearly pedestrian of vehicle, vehicle is matched in background data base (40) using vehicle noise audible spectrum feature
Type, and then the corresponding artificial nerve network model of the type of vehicle is called to handle current vehicle noise data, it exports
Current vehicle speed and vehicle and pedestrian's spacing, and then show that threshold distance and vehicle carry out danger judgement with pedestrian's spacing；
There are pre-set control program in control module (50), inside, are responsible for receiving the finger of central processing unit (30)
It enables, controls the working condition of prior-warning device；
Source of early warning (60), i.e., provisioned usual hardware equipment on smart mobile phone, be responsible for reminding smart phone user and
When avoid vehicle, including mobile phone vibratory equipment, speech ciphering equipment, when source of early warning (60) receives the instruction of control module (50)
When, vibration signal is generated for first early warning and the voice messaging prerecorded reminds mobile phone user, simultaneously for urgent early warning
With displaying information on screen, vibration signal and into earphone jack, mandatory insertion voice prompting message reminds hand to switch mobile phone screen
Machine user's Emergency avoidance.
As shown in Fig. 2, the method for building up step of the matching relationship and model that are stored in above-mentioned background data base (40) is such as
Under：
B1. training sample is acquired：The training sample to be acquired include vehicle noise data, type of vehicle, speed, vehicle with
Then the spacing of pedestrian and the relative position relation of the two are stored in database to training sample according to certain storage mode
In, wherein the vehicle noise data include vehicle noise sound intensity data and audio data；
Above-mentioned vehicle noise training sample include p speed, g vehicle and pedestrian's spacing, three kinds of type of vehicle and
Two kinds of vehicles are with the total 6pg under pedestrian position relationship to vehicle noise training sample data；
Above-mentioned storage mode is：Two store files are established according to the relative distance relationship of vehicle and pedestrian, are then divided
The subfile for not establishing three type of vehicle in the store files of vehicle and the relative distance relationship of pedestrian, by above-mentioned speed sample
This value, spacing sample value and collected corresponding vehicle noise data are stored in subfile, wherein in three kinds of type of vehicle
Subfile in speed sample value, vehicle with pedestrian's spacing sample value and collected corresponding vehicle noise data with square of classifying
The form of battle array is stored, and matrix content see the table below shown：Wherein Vi(i=1,2 ..., be p) p speed training sample value, Sj
(j=1,2 ..., be g) vehicle and pedestrian's spacing training sample value, FijIt is vehicle with vehicle velocity ViIn distance values SjWhen vehicle make an uproar
Speech intensity values, QijIt is vehicle with vehicle velocity ViIn distance values SjWhen vehicle noise audio value；
Above-mentioned type of vehicle includes compact car, in-between car and large car, and wherein compact car includes car, buggy
And 7 station wagons below etc.；In-between car includes medium truck, 7 to 40 middle buses etc.；Large car include trailer,
Engineering truck, 40 or more motor buses, truck and container car etc.；
The principle of classification of the relative distance relationship of above-mentioned vehicle and pedestrian be according to the distance change trend with pedestrian into
Row classification is divided into approaching pedestrian and sails out of two kinds of situations of pedestrian, acquires the row of each type vehicle in both cases respectively
Sail speed, vehicle and pedestrian's spacing and vehicle noise data；
Above-mentioned vehicle and pedestrian's spacing are able to detect that in the maximum magnitude D of vehicle noise in smart mobile phone, by this
The obtained g distance values of the uniform decile of maximum magnitude；
Above-mentioned speed sample value is uniformly to choose p respectively in car speed section (0,100km/h) based on experience value
A speed is as training sample；
B2. vehicle noise feature is extracted：Approaching and two kinds of pedestrian is being sailed out of respectively first, is analyzing and extract three
Kind of type of vehicle is with different speeds common vehicle noise sound intensity variation characteristic when driving；In approaching pedestrian, analysis is simultaneously
Extraction respectively from vehicle noise audible spectrum feature common under the corresponding different speeds of the vehicle of three types；
B3. matching relationship is established：The matching relationship includes two kinds, the first is vehicle and the relative distance pass of pedestrian
Matching relationship between system and vehicle noise sound intensity variation characteristic, for judging that vehicle is just still sailing out of pedestrian in approaching pedestrian；
Matching relationship between the vehicle and the relative distance relationship and vehicle noise sound intensity variation characteristic of pedestrian is：Vehicle noise sound
Strong variation sails out of pedestrian with vehicle in the feature of reduction trend and matches, and is in the feature and vehicle approaching pedestrian's phase of raising trend
Match；It is for second the matching relationship between vehicle noise audible spectrum feature and type of vehicle, large car, in-between car and compact car
Match respectively with three kinds of vehicle noise audible spectrum features, for according to vehicle noise audible spectrum feature recognition vehicle class
Type；
B4. model is established：Do not consider the case where vehicle sails out of pedestrian, only establishes each in the case of vehicle approaching pedestrian
The artificial nerve network model of vehicle noise data and speed, vehicle and pedestrian's spacing corresponding to type of vehicle；Wherein, described
Artificial nerve network model in vehicle noise data and speed, vehicle and pedestrian's spacing are carried out at distributed parallel information
Reason, by adjusting the interconnected relationship between internal node find vehicle noise data and speed, vehicle and pedestrian's spacing it
Between corresponding relation.
Above-mentioned artificial nerve network model is classified according to type of vehicle, call artificial nerve network model when first
Go out type of vehicle according to vehicle noise audible spectrum feature recognition, it is defeated into the corresponding artificial nerve network model of the type of vehicle
Enter current vehicle noise data, the number of plies by changing hidden layer is trained, and current vehicle speed, vehicle and row are obtained in output layer
The human world away from.
Above-mentioned artificial neural network uses the BP neural network of simple-type, as shown in figure 3, the network model establishment step
It is as follows：
Wherein, xiExpression vehicle noise data, i=1,2, x1、x2Vehicle noise sound intensity data value and audio are indicated respectively
Data value；ωjiIndicate that hidden layer implies node for j-th to the weighted value between i-th of input node of input layer, j=1,
2,...,h；γ (x) indicates the excitation function of hidden layer；θjIndicate j-th of threshold value for implying node of hidden layer；WkjIndicate output
K-th of output node of layer is to j-th of weighted value implied between node of hidden layer；η (x) indicates the excitation function of output layer；bk
The threshold value of expression k-th of output node of output layer, k=1,2.
Communication process of the signal in BP neural network model is as follows：
J-th of input value for implying node of hidden layer
J-th of output valve for implying node of hidden layer
The input value of k-th of output node of output layer
The output valve y of k-th of output node of output layerk, i.e. y1And y2Speed and vehicle and pedestrian's spacing are indicated respectively
Value：
As shown in figure 4, a kind of smart mobile phone traffic hazard method for early warning based on vehicle noise, steps are as follows for this method：
S1. the vehicle noise data in smart mobile phone microphone timely collection ambient enviroment are utilized, vehicle is constituted and makes an uproar
Sound array of data, the vehicle noise array of data include vehicle noise sound intensity array of data (F1,F2,...,Fm) and audio
Array of data (Q1,Q2,...,Qm)；
S2. vehicle noise feature extraction, the vehicle noise feature are carried out to collected vehicle noise array of data
Including vehicle noise sound intensity variation characteristic and audible spectrum feature；
S3. according to vehicle noise sound intensity variation characteristic, the vehicle noise sound intensity stored in background data base (40) is called to become
Change feature and vehicle match obtaining a result with the matching relationship of pedestrian's relative distance relationship；
If S4. matching result is that vehicle is in approaching pedestrian's state, the first early warning of smart mobile phone is triggered immediately, it is no
Then, smart mobile phone is still in real-time acquisition state；
S5. under the premise of smart mobile phone has triggered first early warning, background data base (40) is called according to audible spectrum feature
The audible spectrum feature of middle storage and the matching relationship of type of vehicle match current vehicle type, to call the type of vehicle
Corresponding artificial nerve network model；
S6. the last one component is extracted from vehicle noise sound intensity array of data and audio data array respectively, i.e., currently
Vehicle noise data, be input in artificial nerve network model, export current vehicle speed V and vehicle and pedestrian's interval S；
S7. determine that distance threshold L, the distance threshold L refer to vehicle braking distance L by current vehicle speed V1, driver
Reaction distance L2And system response distance L3The sum of, wherein the relationship between speed and vehicle braking distance isFormula
For middle a values in the empirical value of large, medium and small type vehicle, default value of the empirical value as system can also be by user according to oneself
Wish be voluntarily arranged；
S8. compare the size of distance threshold L and current vehicle and pedestrian's interval S, if current vehicle is less than with pedestrian's interval S
Equal to distance threshold L, then the urgent early warning of smart mobile phone is triggered immediately, otherwise, still in the real-time acquisition shape of vehicle noise data
State.
Further, the vehicle noise feature includes vehicle noise sound intensity variation characteristic and audible spectrum feature, such as
Shown in Fig. 5, feature extracting method includes as follows：
S21. the real-time inductive pick-up vehicle noise sound intensity array of data of smart mobile phone microphone and audio data array；
S22. sound intensity data value discrete in sound intensity array of data is done into continuous treatment, obtains function curve variation diagram,
Extract plots changes feature, as sound intensity variation characteristic；
S23. vehicle noise audio data array is done into Fourier transformation, obtains noise spectrum, then extract noise audio
Frequency, amplitude and the phase property of frequency spectrum, as audible spectrum feature.
Above-mentioned first early warning is the vibration to mobile phone user and voice after smart mobile phone detects that vehicle travels noise
It reminds, which does not carry out any interference to smart mobile phone other applications and control.
Above-mentioned urgent early warning is after vehicle enters threshold value with pedestrian's spacing, and system control smart mobile phone carries out vibration and carries
It wakes up, interrupt mobile phone task, automatically switch mobile phone screen and be inserted into mandatory voice prompt into earphone jack, when danger is eliminated
Afterwards, then original working condition is responded immediately to.
The smart mobile phone traffic hazard early warning system based on vehicle noise of the present invention can be used as one in smart mobile phone
Item software application, you can provide free download installation procedure with the smart phone user for each platform, user downloads and pacifies
After filling corresponding program, danger early warning can be carried out using smart mobile phone, improves the safety coefficient of trip.
Above-mentioned smart mobile phone traffic hazard early warning system is just constantly in open state, beats after smart mobile phone booting
After opening the system, it is provided on system interface and selects one-to-one virtual key, the selection of each function to divide into function
Surely there are multiple rational options, can be arranged with default system, can also be voluntarily arranged for user.
Using these virtual keys, user can select the frequency of smart mobile phone collection vehicle noise data, and every time
Constitute the component number m of vehicle noise array of data；The option set according to system, user can be background data base
(40) distance threshold is calculated in, desired braking acceleration a is set, or desired threshold is directly set in early warning system
Be worth distance x, the braking acceleration a option values being arranged in system in the braking acceleration empirical value of each usual public vehicles,
The option of threshold distance x is the value taken in vehicle and the safe distance of pedestrian, if user is arranged desired braking and adds simultaneously
Speed a and threshold distance x, then the threshold distance x of the automatic default user setting of system is effectively setting.
Meanwhile selecting virtual key, user that desired early warning alert interval and prompting form can be set by function,
Prompting form option includes that vibrating alert, voice reminder, screens switch are reminded, can select it is one or more, but must be at least
One kind is selected, whole selected states of system default can also be kept.
Based on above technical scheme, the present invention has using the smart mobile phone traffic hazard method for early warning of vehicle noise with system
Standby following advantage：
The present invention need not install other hardware devices on smart mobile phone or elsewhere, and cost is few and convenient for taking
Band takes the method for establishing matching relationship and BP neural network model that vehicle noise is made to be changed by acquiring training sample data
The alerting signal of smart mobile phone plays mobile phone or listens the pedestrian of mobile phone or special population to provide danger on road with earphone to wallow in
Danger gives warning in advance, more favorable to protect out administrative staff, reduces accident rate；In addition, the method for the present invention and system are pair
One popularization of the application and development of smart mobile phone, has more wide development prospect.
It should be noted last that above technical scheme describes the invention in detail, those skilled in the art couple
This programme is modified or replaced equivalently, and without departure from the spirit and scope of the technical program, should all be covered in the present invention
Right in.
Claims (8)
1. a kind of smart mobile phone traffic hazard early warning system based on vehicle noise, which is characterized in that the smart mobile phone traffic is endangered
Dangerous early warning system includes data acquisition module, characteristic extracting module, central processing unit, background data base, control module and pre-
Alert module；The input terminal of the characteristic extracting module connects acquisition module, and output end connects central processing unit；In described
Central Processing Unit output end is connect with control module, and is mutually communicated with background data base；The control module is responsible for connecing
The control instruction of central processing unit is received, and controls the working condition of warning module；The smart mobile phone traffic hazard early warning
System is that the application APP developed on intelligent mobile phone platform as one is present in mobile phone main interface；
The data acquisition module, including it is placed in the microphone inside smart mobile phone, the vehicle being responsible for around acquisition in real time is made an uproar
Sound data, the vehicle noise data include vehicle noise sound intensity data and audio data；
The characteristic extracting module is responsible for extracting vehicle noise feature from vehicle noise array of data, and is transmitted to
Central processing unit, the vehicle noise feature include vehicle noise sound intensity variation characteristic and audible spectrum feature；
The central processing unit is the decision package of whole system, is communicated with background data base, is responsible for according to vehicle
Feature of noise calls the matching relationship and model stored in background data base, and current vehicle noise data is inputted into model, defeated
Go out as a result, being analyzed and determined, and instruction is sent out to control module；
There are pre-set control program in the control module, inside, are responsible for receiving the instruction of central processing unit, control
The working condition of prior-warning device；
The warning module, that is, the usual hardware response apparatus being placed in inside smart mobile phone, including mobile phone screen, speech ciphering equipment
And vibratory equipment, it is responsible for that smart phone user is reminded to avoid vehicle in time.
2. a kind of smart mobile phone traffic hazard early warning system based on vehicle noise according to claim 1, feature exist
In steps are as follows for the method for building up of the matching relationship and model that are stored in the background data base：
B1. training sample is acquired：The training sample of acquisition is stored in background data base, and training sample includes vehicle noise number
According to, type of vehicle, speed, vehicle and pedestrian's spacing and vehicle and pedestrian's relative distance relationship, the type of vehicle include
Large car, in-between car and compact car three types；The vehicle and pedestrian's relative distance relationship include vehicle approaching pedestrian
With sail out of two kinds of relative distance relationships of pedestrian；
B2. vehicle noise feature is extracted：Vehicle approaching pedestrian and two kinds of pedestrian is being sailed out of respectively, is analyzing and extract three
Kind of type of vehicle is with different speeds common vehicle noise sound intensity variation characteristic when driving；In approaching pedestrian, analysis is simultaneously
Extract respectively from vehicle noise audible spectrum feature common under the corresponding different speed of operation of three types vehicle；
B3. matching relationship is established：The matching relationship includes two kinds of matching relationships, i.e., the first is the variation of the vehicle noise sound intensity
Matching relationship between feature and vehicle and pedestrian's relative distance relationship, for being judged according to vehicle noise sound intensity variation characteristic
Vehicle is approaching pedestrian or sails out of pedestrian, is for second that the matching between vehicle noise audible spectrum feature and type of vehicle is closed
System, for going out type of vehicle according to vehicle noise audible spectrum characteristic matching；
B4. model is established：Do not consider the case where vehicle sails out of pedestrian, is only established and three kinds of vehicles in the case of vehicle approaching pedestrian
Type distinguishes the artificial nerve network model of corresponding vehicle noise data and speed, vehicle and pedestrian's spacing, and a kind of
Type of vehicle corresponds to a kind of execution program of artificial nerve network model；The artificial nerve network model is to vehicle noise number
Concurrent collaborative processing is carried out according to speed, vehicle and pedestrian's spacing, by adjusting the interconnection between internal node in network
Relationship finds the corresponding relation between vehicle noise data and speed, vehicle and pedestrian's spacing；The artificial neural network mould
Type uses the BP neural network model of simple-type, and input layer is vehicle noise data, and output layer is between speed, vehicle and pedestrian
Away from.
3. a kind of smart mobile phone traffic hazard early warning system based on vehicle noise according to claim 2, feature exist
In the storage form of the training sample of the acquisition is：Two storage texts are established according to vehicle and pedestrian's relative distance relationship
Then part establishes the subfile of three type of vehicle in the store files of vehicle and pedestrian's relative distance relationship respectively, will be upper
Speed sample value, vehicle are stated with pedestrian's spacing sample value and collected corresponding vehicle noise data in the form of classification matrix
It is stored in subfile.
4. a kind of smart mobile phone traffic hazard early warning system based on vehicle noise according to claim 2, feature exist
In the matching relationship between the vehicle noise sound intensity variation characteristic and vehicle and pedestrian's relative distance relationship is：Vehicle is made an uproar
The strong variation of speech sails out of pedestrian with vehicle in the feature of reduction trend and matches, in feature and the vehicle approaching pedestrian of raising trend
Match.
5. a kind of smart mobile phone traffic hazard method for early warning based on vehicle noise, which is characterized in that program includes six modules：
Data acquisition module, characteristic extracting module, central processing unit, background data base, control module and warning module, this method step
It is rapid as follows：
S1. the vehicle noise data in ambient enviroment are acquired in real time using smart mobile phone microphone, constitute vehicle noise data number
Group, the vehicle noise array of data include vehicle noise sound intensity array of data and audio data array；
S2. vehicle noise feature extraction is carried out to collected vehicle noise array of data, the vehicle noise feature includes
Vehicle noise sound intensity variation characteristic and audible spectrum feature；
S3. according to vehicle noise sound intensity variation characteristic, call the vehicle noise sound intensity variation characteristic stored in background data base and
Vehicle match obtaining a result with the matching relationship of pedestrian's relative distance relationship；
If S4. matching result is that vehicle is in approaching pedestrian's state, the first early warning of smart mobile phone, otherwise, intelligence are triggered immediately
Energy mobile phone is still in real-time acquisition state；
S5. under the premise of smart mobile phone has triggered first early warning, background data base is called according to vehicle noise audible spectrum feature
The vehicle noise audible spectrum feature of middle storage and the matching relationship of type of vehicle, match current vehicle type, to call
Artificial nerve network model corresponding to the type of vehicle；
S6. the last one data component is extracted from vehicle noise sound intensity array of data and audio data array respectively, i.e., currently
Vehicle noise data, be input in artificial nerve network model, output current vehicle speed, vehicle and pedestrian's spacing；
S7. distance threshold is determined by current vehicle speed, the distance threshold refer to vehicle braking distance, driver reaction distance with
And system response sum of the distance；
S8. compare the size of current vehicle and pedestrian's spacing and distance threshold, if current vehicle and pedestrian be smaller than equal to away from
From threshold value, then the urgent early warning of smart mobile phone is triggered immediately, otherwise, still in the real-time acquisition state of vehicle noise data.
6. a kind of smart mobile phone traffic hazard method for early warning based on vehicle noise according to claim 5, feature exist
In vehicle noise feature extracting method includes as follows：
S21. the vehicle noise data in smart mobile phone microphone timely collection ambient enviroment constitute vehicle noise data number
Group, the vehicle noise array of data include vehicle noise sound intensity array of data and audio data array；
S22. sound intensity data value discrete in sound intensity array of data is done into continuous treatment, obtains function curve variation diagram, extracted
Plots changes feature, as sound intensity variation characteristic；
S23. vehicle noise audio data array is done into Fourier transformation and obtains noise spectrum, then extract noise audible spectrum
Frequency, amplitude and phase property, as audible spectrum feature.
7. a kind of smart mobile phone traffic hazard method for early warning based on vehicle noise according to claim 5, feature exist
In the first early warning is after smart mobile phone detects vehicle noise, and the vibration to mobile phone user and voice reminder, this is first
Secondary early warning does not carry out any interference to smart mobile phone other applications and controls.
8. a kind of smart mobile phone traffic hazard method for early warning based on vehicle noise according to claim 5, feature exist
In, the urgent early warning be after vehicle and pedestrian's spacing enter threshold value, system control smart mobile phone carry out vibrating alert, in
Cut off the hands machine current task, automatically switch mobile phone screen and the mandatory insertion voice prompt into earphone jack, when danger is eliminated
Afterwards, then original working condition is responded immediately to.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611005062.7A CN106448050B (en) | 2016-11-15 | 2016-11-15 | A kind of smart mobile phone traffic hazard method for early warning and system based on vehicle noise |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611005062.7A CN106448050B (en) | 2016-11-15 | 2016-11-15 | A kind of smart mobile phone traffic hazard method for early warning and system based on vehicle noise |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106448050A CN106448050A (en) | 2017-02-22 |
| CN106448050B true CN106448050B (en) | 2018-08-28 |
Family
ID=58207198
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611005062.7A Active CN106448050B (en) | 2016-11-15 | 2016-11-15 | A kind of smart mobile phone traffic hazard method for early warning and system based on vehicle noise |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106448050B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107171685A (en) * | 2017-04-24 | 2017-09-15 | 努比亚技术有限公司 | A kind of Wearable, terminal and safety instruction method |
| CN108022596A (en) * | 2017-11-28 | 2018-05-11 | 湖南海翼电子商务股份有限公司 | Audio signal processing method and vehicle electronic device |
| CN109961532A (en) * | 2017-12-25 | 2019-07-02 | 大连楼兰科技股份有限公司 | Dangerous information early warning and transmission storage system |
| EP3550856A1 (en) * | 2018-04-04 | 2019-10-09 | Ningbo Geely Automobile Research & Development Co. Ltd. | Vehicle alert device and method |
| CN110021141B (en) * | 2019-03-21 | 2021-11-02 | 深圳市科迈爱康科技有限公司 | Blind guiding early warning method, device and system |
| CN109887271A (en) * | 2019-03-21 | 2019-06-14 | 深圳市科迈爱康科技有限公司 | Pedestrains safety method for early warning, apparatus and system |
| CN110502980B (en) * | 2019-07-11 | 2021-12-03 | 武汉大学 | Method for identifying scene behaviors of pedestrians playing mobile phones while crossing roads |
| CN112172836B (en) * | 2020-09-28 | 2022-05-13 | 广州小鹏汽车科技有限公司 | Information pushing method and device, vehicle and readable medium |
| CN112437372B (en) * | 2020-10-28 | 2022-03-11 | 头领科技（昆山）有限公司 | Danger alarm system chip for closed earphone |
| CN112671977A (en) * | 2020-12-25 | 2021-04-16 | 深圳市九洲电器有限公司 | Safety early warning method and device, storage medium and electronic equipment |
| CN113975110A (en) * | 2021-10-20 | 2022-01-28 | 北京声智科技有限公司 | Blind guiding safety prompting method, blind guiding equipment, blind guiding device and electronic equipment |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8942662B2 (en) * | 2012-02-16 | 2015-01-27 | The United States of America, as represented by the Secretary, Department of Health and Human Services, Center for Disease Control and Prevention | System and method to predict and avoid musculoskeletal injuries |
| KR101609914B1 (en) * | 2014-02-19 | 2016-04-07 | 이동원 | the emergency situation sensing device responding to physical and mental shock and the emergency situation sensing method using the same |
| CN204887366U (en) * | 2015-07-19 | 2015-12-16 | 段太发 | Can monitor bluetooth headset of environment sound |
| CN105741856A (en) * | 2016-04-08 | 2016-07-06 | 王美金 | Earphone capable of prompting environmental crisis sounds in listening to music state |
| CN106060643A (en) * | 2016-06-28 | 2016-10-26 | 乐视控股（北京）有限公司 | Method and device for playing multimedia file and earphones |
-
2016
- 2016-11-15 CN CN201611005062.7A patent/CN106448050B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN106448050A (en) | 2017-02-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106448050B (en) | A kind of smart mobile phone traffic hazard method for early warning and system based on vehicle noise | |
| Albert et al. | Which smartphone's apps may contribute to road safety? An AHP model to evaluate experts' opinions | |
| CN107316436A (en) | Dangerous driving state processing method, electronic equipment and storage medium | |
| CN108989541A (en) | Session initiation device, system, vehicle and method based on situation | |
| US20100305806A1 (en) | Portable Multi-Modal Emergency Situation Anomaly Detection and Response System | |
| CN108735203A (en) | Voice interactive method, terminal and computer-readable medium | |
| CN105282703B (en) | Contact the method and system of vehicle and interior mobile terminal | |
| CN101364408A (en) | Sound image combined monitoring method and system | |
| CN110366107A (en) | Vehicle communication method and the device for using this method | |
| CN110060484A (en) | A kind of real-time early warning system and method violating the regulations of the highway passenger transportation based on block chain | |
| CN204087490U (en) | A kind of giving fatigue pre-warning system based on machine vision | |
| CN108305623A (en) | Electric control method and device | |
| CN206931362U (en) | A kind of freeway tunnel safety monitoring system | |
| CN112071309B (en) | Network appointment vehicle safety monitoring device and system | |
| CN108437993A (en) | A kind of road anger vehicle drive automatically control system and method | |
| CN107315519A (en) | OS switching methods, device and mobile terminal under driving condition | |
| CN204890010U (en) | Driver system of testing and assessing | |
| CN106713633A (en) | Deaf people prompt system and method, and smart phone | |
| CN108492819A (en) | Language exercise method, apparatus, intelligent vehicle mounted terminal and storage medium | |
| CN108694407A (en) | A kind of driving behavior recognition methods based on mobile terminal | |
| CN109451385A (en) | A kind of based reminding method and device based on when using earphone | |
| CN102568158A (en) | Alarming system for automobile vehicle fatigue driving | |
| CN106904122A (en) | Vehicle speech sound alarming method and apparatus | |
| CN104361887A (en) | Quick acoustic event detection system under traveling noise environment | |
| CN107867295A (en) | Be in danger accidents early warning method, storage device and the car-mounted terminal of probability based on vehicle |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |