CN103928026B - Automobile voice command acquiring and processing system and method - Google Patents
Automobile voice command acquiring and processing system and method Download PDFInfo
- Publication number
- CN103928026B CN103928026B CN201410198669.6A CN201410198669A CN103928026B CN 103928026 B CN103928026 B CN 103928026B CN 201410198669 A CN201410198669 A CN 201410198669A CN 103928026 B CN103928026 B CN 103928026B
- Authority
- CN
- China
- Prior art keywords
- signal
- noise reduction
- data signal
- main channel
- voice signal
- 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
Landscapes
- Fittings On The Vehicle Exterior For Carrying Loads, And Devices For Holding Or Mounting Articles (AREA)
- Soundproofing, Sound Blocking, And Sound Damping (AREA)
Abstract
The invention discloses an automobile voice command acquiring and processing system and method. A first denoising unit of the system conducts analog-digital conversion on a left channel sound signal and a right channel sound signal to obtain a left channel digital signal and a right channel digital signal. A second denoising unit conducts analog-digital conversion on a main channel sound signal to obtain a main channel digital signal. A signal processing unit conducts denoising on the main channel digital signal according to the left channel digital signal and the right channel digital signal to obtain a denoised digital signal. The second denoising unit further conducts denoising on the main channel sound signal according to a superposed signal of the left channel sound signal and the right channel sound signal and the denoised digital signal to obtain a denoised voice command. According to the automobile voice command acquiring and processing system and method, engine noise, outside noise and air-conditioning system and passenger seat noise can be filtered out effectively, the recognition rate of control commands of a driver is increased, and traveling security of automobiles is guaranteed.
Description
Technical field
The present invention relates to signal processing technology field, more particularly to a kind of automobile speech instruction acquisition processing system and side
Method.
Background technology
With the development of science and technology, the noise pollution in modern transportation instrument passenger cabin becomes topic of concern.How
Reduce as far as possible due to engine operating, air-flow and vehicle body friction, air-conditioning system and the passenger side in the environment of this relative closure
Sail the noise pollution that sound etc. is brought such that it is able to accurately identify the control instruction that driver sends, become research both at home and abroad
Focus.
Traditional phonetic order acquisition processing system includes the main wheat for picking up driver's control instruction (voice signal)
Gram wind and the signal processing module for being processed voice signal, signal processing module is used for making an uproar in voice signal
Acoustical signal is filtered.The defect of said system is：When extraneous and in-car interference sound is less, signal processing module can be by language
The noise filtering adulterated in message number, system can also play a role；When driver opens vehicle window and opens air-conditioning etc., due to
The noise type adulterated in voice signal is more, and system can not targetedly filter noise, so as to the system is to driver's
The discrimination of control instruction will have a greatly reduced quality, and the recognition result of mistake is easily caused maloperation, so as to not only not offer convenience, instead
And some unnecessary troubles can be brought, or even the security driven a vehicle can be affected.
The content of the invention
Asking for the outer noise pollution of in-car is not considered during driver's phonetic order is recognized in order to solve prior art
Topic, the invention provides a kind of automobile speech instruction acquisition processing system and method.
For achieving the above object, the technical scheme is that：
A kind of automobile speech instruction acquisition processing system, including：
Main channel sound pickup device, for obtaining main channel voice signal；
Left channel sound pickup device, for obtaining left channel voice signal；
Right channel sound pickup device, for obtaining right channel voice signal；
First noise reduction unit, for the left channel voice signal that sends to the left channel sound pickup device and institute
Stating the right channel voice signal of right channel sound pickup device transmission carries out respectively analog-to-digital conversion, obtains left channel numeral
Signal and right channel data signal；
Second noise reduction unit, for carrying out modulus to the main channel voice signal that the main channel sound pickup device sends
Conversion, obtains main channel data signal；
Signal processing unit, for the left channel data signal that sent according to first noise reduction unit and right channel
Data signal, to the main channel data signal that second noise reduction unit sends noise reduction process is carried out, and obtains the numeral after noise reduction
Signal；
First noise reduction unit, is additionally operable to enter the left channel voice signal and the right channel voice signal
Row superposition, the voice signal after being superimposed；
Second noise reduction unit, be additionally operable to according to first noise reduction unit send superposition after voice signal and
Data signal after the noise reduction that the signal processing unit sends, to the main channel voice signal noise reduction process is carried out, and is obtained
Phonetic order after noise reduction；
The left channel pick device is located at the left side of the main channel sound pickup device and near car engine
The position of machine, the right channel pick device is located at the right side of the main channel sound pickup device and just copilot is being taken advantage of
The position of visitor；
Wherein, signal processing unit obtains the data signal after noise reduction includes：
(1) three railway digital signals are read in, is labeled as V0, V1, V, be corresponding in turn to left channel data signal, right side and lead to
Road data signal and main channel data signal, wherein main channel data signal are from the closest of driver, and voice signal is most bright
It is aobvious；
(2) to three railway digital signal V0, V1, V carry out sub-frame processing simultaneously, and frame length is 256 points, and it is 128 points that frame is moved, not enough
Mend 0；
(3) to three road signal V0, V1, V carries out Hamming filtering；
(4) sound end identification is carried out using short-time average energy algorithm to signal V；
(5) simultaneously Fast Fourier Transform (FFT) FFT is carried out to three road voice signals；
(6) voice of signal V and the power spectrum of noise are asked for；
(7) signal of signal V and the frequency domain power ratio of noise are asked for, recycles look-up table to obtain the noise reduction coefficient of signal V0
The noise reduction coefficient β of α and signal V1；
(8) using power spectrum-subtraction：E (V)-α E (V0)-β E (V1), calculate the power of the signal V after noise reduction；
(9) Fast Fourier Transform Inverse is carried out to the phase place of the signal V before the power and noise reduction of the signal V after noise reduction
IFFT, the data value after conversion retains real part；
(10) Hamming window is removed；
(11) framing signal is converted to into voice signal, modulation output signal, so as to obtain noise reduction after data signal.
Preferably, first noise reduction unit and second noise reduction unit are the broadband voice that model is R262 and catch
Catch noise reduction integrated chip.
Preferably, the main channel sound pickup device, the left channel sound pickup device and the right side lead to
Road sound pickup device is arranged on Auto Instrument desk；
The main channel sound pickup device is just arranged to driver, and the left channel pick device is logical positioned at the master
The left side of road sound pickup device and the position of close automobile engine, the right channel pick device is logical positioned at the master
The right side of road sound pickup device and just to the position of copilot passenger.
Preferably, the signal processing unit is single-chip microcomputer.
A kind of automobile speech instruction acquisition processing method, including：
Obtain main channel voice signal, left channel voice signal and right channel voice signal；
Carry out modulus respectively to the main channel voice signal, left channel voice signal and right channel voice signal to turn
Change, obtain main channel data signal, left channel data signal and right channel data signal；
According to the left channel data signal and right channel data signal, the main channel data signal is dropped
Make an uproar process, obtain the data signal after noise reduction；
The left channel voice signal and the right channel voice signal are overlapped, the sound after being superimposed
Signal；
The main channel voice signal is entered according to the data signal after the voice signal after the superposition and the noise reduction
Row noise reduction process, obtains the phonetic order after noise reduction；
The left channel voice signal is by positioned at the left side of pick device of the main channel voice signal and close
The pick device pickup of the position of automobile engine, the right channel voice signal is by positioned at the main channel voice signal
The right side of pick device and just to the pick device pickup of the position of copilot passenger；
Wherein, it is described according to the left channel data signal and right channel data signal, to main channel numeral
Signal carries out noise reduction process, obtains the data signal after noise reduction, including：
(1) three railway digital signals are read in, is labeled as V0, V1, V, be corresponding in turn to left channel data signal, right side and lead to
Road data signal and main channel data signal, wherein main channel data signal are from the closest of driver, and voice signal is most bright
It is aobvious；
(2) to three railway digital signal V0, V1, V carry out sub-frame processing simultaneously, and frame length is 256 points, and it is 128 points that frame is moved, not enough
Mend 0；
(3) to three road signal V0, V1, V carries out Hamming filtering；
(4) sound end identification is carried out using short-time average energy algorithm to signal V；
(5) simultaneously Fast Fourier Transform (FFT) FFT is carried out to three road voice signals；
(6) voice of signal V and the power spectrum of noise are asked for；
(7) signal of signal V and the frequency domain power ratio of noise are asked for, recycles look-up table to obtain the noise reduction coefficient of signal V0
The noise reduction coefficient β of α and signal V1；
(8) using power spectrum-subtraction：E (V)-α E (V0)-β E (V1), calculate the power of the signal V after noise reduction；
(9) Fast Fourier Transform Inverse is carried out to the phase place of the signal V before the power and noise reduction of the signal V after noise reduction
IFFT, the data value after conversion retains real part；
(10) Hamming window is removed；
(11) framing signal is converted to into voice signal, modulation output signal, so as to obtain noise reduction after data signal.
Preferably, according to the left channel data signal and right channel data signal, by power spectrum-subtraction pair
The main channel data signal carries out noise reduction process, obtains the data signal after noise reduction.
The beneficial effects of the present invention is, embodiment of the present invention automobile speech instruction acquisition processing system and method can
Engine noise, pass-by noise, air-conditioning system and copilot noise are effectively filtered out, the identification of driver's control instruction is improve
Rate, it is ensured that the traffic safety of vehicle.
Description of the drawings
Fig. 1 shows the structural representation of embodiment of the present invention automobile speech instruction acquisition processing system；
Fig. 2 shows the circuit diagram of embodiment of the present invention automobile speech instruction acquisition processing system；
Fig. 3 shows the flow chart of embodiment of the present invention automobile speech instruction acquisition processing method.
Description of reference numerals
MC1 left channel sound pickup devices
MC2 right channel sound pickup devices
MC3 main channels sound pickup device
The noise reduction units of U1 first
The noise reduction units of U2 second
U3 signal processing units
U4 reset units
The memory cell of U5 first
The memory cell of U6 second
Specific embodiment
Embodiments of the invention are described below in detail, the example of the embodiment is shown in the drawings, wherein from start to finish
Same or similar label represents same or similar element or the element with same or like function.Below with reference to attached
The embodiment of figure description is exemplary, is only used for explaining the present invention, and is not construed as limiting the claims.
The present invention is not considering the outer noise pollution situation of in-car for phonetic order acquisition processing system of the prior art
Under, it is low to the discrimination of the control instruction of driver, or even the problem of the security that can affect to drive a vehicle, there is provided a kind of automobile language
Sound instruction acquisition processing system and method, fully take into account engine noise, pass-by noise, air-conditioning system and copilot and make an uproar
Sound, improves the discrimination of driver's control instruction, it is ensured that the traffic safety of vehicle.
Fig. 1 shows the structural representation of embodiment of the present invention automobile speech instruction acquisition processing system, the automobile language
Sound instruction acquisition processing system, including main channel sound pickup device MC3, left channel sound pickup device MC1, right channel
Sound pickup device MC2, the first noise reduction unit U1, the second noise reduction unit U2 and signal processing unit U3, wherein, left channel sound
Sound pick device MC1 and right channel sound pickup device MC2 are electrically connected with the first noise reduction unit U1, main channel voice pickup
Device MC3 is electrically connected with the second noise reduction unit U2, and first noise reduction unit U1 and the second noise reduction unit U2 are and signal transacting
Unit U3 is electrically connected, and the first noise reduction unit U1 and the second noise reduction unit U2 are electrically connected to each other.
Specifically, first：Main channel sound pickup device MC3, for obtaining main channel voice signal；Left channel sound
Pick device MC1, for obtaining left channel voice signal；Right channel sound pickup device MC2, for obtaining right channel
Voice signal.
Here, these three sound pickup devices are all disposed within Auto Instrument desk, wherein, for picking up main channel sound letter
Number main channel sound pickup device MC3 (such as main channel microphone) be arranged on just to the position of driver, for picking up a left side
The left channel sound pickup device MC1 (such as left channel microphone) of wing passage voice signal is located at the main channel sound
The left side of pick device MC3 and the position of close automobile engine, right channel pick device (such as right channel wheat
Gram wind) it is located at the right side of the main channel sound pickup device MC3 and just to the position of copilot passenger.By above-mentioned to each
The position of individual sound pickup device is arranged, it is possible to achieve：In the voice signal that main channel sound pickup device MC3 is picked up, drive
Proportion shared by the control instruction of the person of sailing is relatively large, i.e. main channel sound pickup device MC3 is mainly used in picking up driver's
Control instruction；In addition, in the voice signals that pick up of left channel sound pickup device MC1, engine noise and pass-by noise
Shared proportion is relatively large, i.e. left channel sound pickup device MC1 is mainly used in picking up engine noise and pass-by noise；
Further, in the voice signals that pick up of right channel sound pickup device MC2, Noise in Air Conditioning System and copilot passenger noise institute
The proportion for accounting for is relatively large, i.e. right channel sound pickup device MC2 is mainly used in picking up Noise in Air Conditioning System and copilot is taken advantage of
Objective noise.
Second：First noise reduction unit U1, for the left channel sound sent to the left channel sound pickup device MC1
The right channel voice signal that message number and the right channel sound pickup device MC2 send carries out respectively analog-to-digital conversion (will
Analog signal is converted to data signal), obtain left channel data signal and right channel data signal.The first noise reduction list
First U1, is additionally operable to be overlapped the left channel voice signal and the right channel voice signal, after being superimposed
Voice signal, and the voice signal after being superimposed is sent to into the second noise reduction unit U2, for subsequent treatment.
Second noise reduction unit U2, for entering to the main channel voice signal that the main channel sound pickup device MC3 sends
Row analog-to-digital conversion, obtains main channel data signal.
Here, the first noise reduction unit U1 and the second noise reduction unit U2 are both preferably the research and development of ON Semiconductor Corporation of the U.S.
Model is that the broadband voice of R262 catches noise reduction integrated chip, and the integrated chip had both possessed speech capturing processing function but also with drop
Make an uproar function.The left channel voice signal for getting is sent to the first noise reduction list by the left channel sound pickup device MC1
First U1, the first noise reduction unit U1 carries out analog-to-digital conversion to the left channel voice signal, obtains corresponding left channel numeral letter
Number；Similarly, the right channel voice signal for getting is sent to the first noise reduction by the right channel sound pickup device MC2
Unit U1, the first noise reduction unit U1 carries out analog-to-digital conversion to the right channel voice signal, obtains corresponding right channel numeral
Signal.The main channel voice signal for getting is sent to the second noise reduction unit U2 by the main channel sound pickup device MC3, the
Two noise reduction units U2 carry out analog-to-digital conversion to the main channel voice signal, obtain corresponding main channel data signal.
3rd, left channel data signal and right channel data signal are sent to signal transacting by the first noise reduction unit U1
Unit U3, while main channel data signal is also sent to signal processing unit U3, the signal transacting by the second noise reduction unit U2
Unit U3 can be selected as the one kind in central processing unit, flush bonding processor, programmable logic controller (PLC) and single-chip microcomputer, at this
In one preferred embodiment of invention, signal processing unit U3 is preferably single-chip microcomputer, specially Microchip. semiconductor companies
The single-chip microcomputer of model PIC18F2682 of research and development.
Signal processing unit U3 is using left channel data signal and right channel data signal to main channel data signal
Carry out noise reduction process, obtain the data signal after noise reduction, the noise reduction algorithm specifically used be those skilled in the art frequently with
Power spectrum-subtraction, the method is mainly included the following steps that：
(1) three railway digital signals are read in, is labeled as V0, V1, V, be corresponding in turn to left channel data signal, right side and lead to
Road data signal and main channel data signal, wherein main channel data signal are from the closest of driver, and voice signal is most bright
It is aobvious；
(2) to three railway digital signal V0, V1, V carry out sub-frame processing simultaneously, and frame length is 256 points, and it is 128 points that frame is moved, not enough
Mend 0；
(3) to three road signal V0, V1, V carries out Hamming filtering；
(4) sound end identification is carried out using short-time average energy algorithm to signal V；
(5) simultaneously Fast Fourier Transform (FFT) FFT is carried out to three road voice signals；
(6) voice of signal V and the power spectrum of noise are asked for；
(7) signal of signal V and the frequency domain power ratio of noise are asked for, recycles look-up table to obtain the noise reduction coefficient of signal V0
The noise reduction coefficient β of α and signal V1；
(8) using power spectrum-subtraction：E (V)-α E (V0)-β E (V1), calculate the power of the signal V after noise reduction；
(9) Fast Fourier Transform Inverse is carried out to the phase place of the signal V before the power and noise reduction of the signal V after noise reduction
IFFT, the data value after conversion retains real part；
(10) Hamming window is removed；
(11) framing signal is converted to into voice signal, modulation output signal, so as to obtain noise reduction after data signal.
Data signal after the noise reduction for getting is returned to the second noise reduction unit U2 by the 4th, signal processing unit U3, the
Voice signal and the signal processing unit after the superposition that two noise reduction units U2 are sent according to first noise reduction unit U1
Data signal after the noise reduction that U3 sends, to the main channel voice signal noise reduction process is carried out, and is obtained the voice after noise reduction and is referred to
Order.
Here, the voice signal after superposition can regard set (engine noise and pass-by noise, the Yi Jikong of noise as
Adjusting system noise and copilot passenger's noise), the data signal after noise reduction can be regarded as through signal processing unit U3 process
Echo signal afterwards, using the data signal after the voice signal and noise reduction after the superposition, larger ratio is accounted for control instruction
The method that the main channel voice signal of weight carries out noise reduction process, is the algorithm of those skilled in the art's routine employing, and the algorithm shows
The inside of the second noise reduction unit U2 has been integrated in, so as to the second noise reduction unit U2 can be by the algorithm of its internal storage, profit
With the data signal after the voice signal and noise reduction after superposition, noise reduction process is carried out to the main channel voice signal, to left side
The noise signal of passage and right channel is decayed and is filtered, and obtains the phonetic order after noise reduction.Export more clearly
Driver's sound instruction signal, reaches the purpose for lifting sound instruction discrimination.
To sum up, embodiment of the present invention automobile speech instruction acquisition processing system, can effectively filter out outside engine noise, car
Noise, air-conditioning system and copilot noise, improve the discrimination of driver's control instruction, it is ensured that the driving peace of vehicle
Entirely.
Further, it is left as shown in Fig. 2 being the circuit diagram of embodiment of the present invention automobile speech instruction acquisition processing system
Wing passage sound pickup device MC1 is delivered to the G1 ends of the first noise reduction unit U1 by left channel voice signal by electric capacity C9 couplings
Mouthful, analog-to-digital conversion is carried out, obtain left channel data signal；Similarly, right channel sound pickup device MC2 passes through electric capacity
Right channel voice signal is delivered to the e1 port of the first noise reduction unit U1 for C10 couplings, carries out analog-to-digital conversion, obtains right channel
Data signal；First noise reduction unit U1 also does overlap-add procedure to left channel voice signal and right channel voice signal, obtains
Voice signal after superposition, and the voice signal after superposition is sent by D6 ports to the G1 ports of the second noise reduction unit U2,
In case subsequent treatment.
Main channel voice signal is delivered to the second noise reduction unit by main channel sound pickup device MC3 by electric capacity C14 couplings
The e1 port of U2, carries out analog-to-digital conversion, obtains main channel data signal.
Left channel data signal and right channel data signal are passed through its C7 port by the first noise reduction unit U1 respectively, are sent
To 11 ports of signal processing unit U3；Simultaneously main channel data signal is also passed through its C7 port by the second noise reduction unit U2, is sent
To 12 ports of signal processing unit U3.Signal processing unit U3 is using left channel data signal and right channel numeral letter
Number, noise reduction process is carried out to main channel data signal with reference to power spectrum-subtraction, obtain the data signal after noise reduction, and by its 5
Port is sent to the C3 ports of the second noise reduction unit U2, while by its 4 port to the second noise reduction unit U2 tranmitting data register signal.
Voice signal and its C3 signal after the superposition that second noise reduction unit U2 is received using its G1 port is received
Noise reduction after data signal, with reference to the noise reduction algorithm of the second noise reduction unit U2 storage inside, its e1 port is received
Main channel voice signal carries out noise reduction process, the noise signal of left channel or right channel is decayed and is filtered, finally
The phonetic order after noise reduction is obtained, that is, obtains more clearly driver's sound instruction signal, reach lifting sound instruction identification
The purpose of rate.
In addition, CAT25640 is referring again to the model of Fig. 2, the first memory cell U5 and the second memory cell U6, point
Do not provide storage information for the first noise reduction unit U1 and the second noise reduction unit U2, the data I/O port 1 of the first memory cell U5,
2,5,6 respectively with the I/O ports C9 of the first noise reduction unit U1, B8, A9, A7 is connected, the data I/O end of the second memory cell U6
Mouthfuls 1,2,5,6 respectively with the I/O ports C9 of the second noise reduction unit U2, B8, A9, A7 is connected, electric capacity C1 to C6, C16 to C20 point
Not Wei the first noise reduction unit U1 and the second noise reduction unit U2 anti-interference capacitor, be mainly used in filtering spike interference, the second drop
Make an uproar unit U2 D6I/O ports and electric capacity C13 composition analogue audio frequency output circuit, reset unit U4 be ON Semiconductor Corporation
The electrification reset integrated circuit of model MAX809 of research and development, the 2nd pin of reset unit U4 connects the 1st of signal processing unit U3
Pin, constitutes electrification reset circuit, and electric capacity CX1 is the crystal oscillator electric capacity of signal processing unit U3, C25, C26 and electric capacity CX1
The clock circuit of composition signal processing unit U3, electric capacity C8, C27 and C28 composition signal processing unit U3, the first noise reduction unit U1
Power filter capacitor with the power supply of the second noise reduction unit U2, signal processing unit U3 is the research and development of Microchip. semiconductor companies
The single-chip microcomputer of model PIC18F2682, signal processing unit U3 is mainly entered by its 4 pin and 5 pin to the second noise reduction unit U2
Row programming Control, particularly as extension, signal processing unit U3 can also be by its 2 pin and 3 pin to the first noise reduction unit U1
It is programmed control.
Correspondingly, the embodiment of the present invention also provides a kind of automobile speech instruction acquisition processing method, as shown in figure 3, being this
The flow chart of method, the automobile speech instruction acquisition processing method is comprised the following steps：
Step 301：Obtain main channel voice signal, left channel voice signal and right channel voice signal.
Step 302：The main channel voice signal, left channel voice signal and right channel voice signal are entered respectively
Row analog-to-digital conversion, obtains main channel data signal, left channel data signal and right channel data signal.
Step 303：According to the left channel data signal and right channel data signal, main channel numeral is believed
Number noise reduction process is carried out, obtain the data signal after noise reduction.
Step 304：The left channel voice signal and the right channel voice signal are overlapped, are superimposed
Voice signal afterwards.
Step 305：According to the data signal after the voice signal after the superposition and the noise reduction to the main channel sound
Message number carries out noise reduction process, obtains the phonetic order after noise reduction.
Embodiment of the present invention automobile speech instruction acquisition processing method, can effectively filter out engine noise, pass-by noise,
Air-conditioning system and copilot noise, improve the discrimination of driver's control instruction, it is ensured that the traffic safety of vehicle.
The specific embodiment of each step of said method can refer to the description in the above system of the embodiment of the present invention,
This is repeated no more.
It should be noted that system embodiment described above is only schematic, wherein described as separation unit
The unit of part explanation can be or may not be physically separate, can be as the part that unit shows or also may be used
Not being physical location, you can with positioned at a place, or can also be distributed on multiple NEs.Can be according to reality
The purpose for needing to select some or all of module therein to realize this embodiment scheme.Those of ordinary skill in the art exist
In the case of not paying creative work, you can to understand and implement.
Construction, feature and the action effect of the present invention, above institute is described in detail according to the embodiment shown in schema above
Only presently preferred embodiments of the present invention is stated, but the present invention is not to limit practical range shown in drawing, it is every according to structure of the invention
Want made change, or be revised as the Equivalent embodiments of equivalent variations, still without departing from specification and diagram covered it is spiritual when,
All should be within the scope of the present invention.
Claims (6)
1. a kind of automobile speech instruction acquisition processing system, it is characterised in that include：
Main channel sound pickup device, for obtaining main channel voice signal；
Left channel sound pickup device, for obtaining left channel voice signal；
Right channel sound pickup device, for obtaining right channel voice signal；
First noise reduction unit, for the left channel voice signal that sends to the left channel sound pickup device and the right side
The right channel voice signal that wing passage sound pickup device sends carries out respectively analog-to-digital conversion, obtains left channel data signal
With right channel data signal；
Second noise reduction unit, turns for carrying out modulus to the main channel voice signal that the main channel sound pickup device sends
Change, obtain main channel data signal；
Signal processing unit, the left channel data signal and right channel for being sent according to first noise reduction unit is digital
Signal, to the main channel data signal that second noise reduction unit sends noise reduction process is carried out, and obtains the data signal after noise reduction；
First noise reduction unit, is additionally operable to be folded the left channel voice signal and the right channel voice signal
Plus, the voice signal after being superimposed；
Second noise reduction unit, the voice signal and described being additionally operable to after the superposition sent according to first noise reduction unit
Data signal after the noise reduction that signal processing unit sends, to the main channel voice signal noise reduction process is carried out, and obtains noise reduction
Phonetic order afterwards；
The left channel sound pickup device is located at the left side of the main channel sound pickup device and near car engine
The position of machine, the right channel sound pickup device is located at the right side of the main channel sound pickup device and just to the passenger side
Sail the position of passenger；
Wherein, signal processing unit obtains the data signal after noise reduction includes：
(1) three railway digital signals are read in, is labeled as V0, V1, V, be corresponding in turn to left channel data signal, right channel number
Word signal and main channel data signal, wherein main channel data signal are from the closest of driver, and voice signal is most obvious；
(2) to three railway digital signal V0, V1, V carry out sub-frame processing simultaneously, and frame length is 256 points, and it is 128 points that frame is moved, and deficiency mends 0；
(3) to three road signal V0, V1, V carries out Hamming filtering；
(4) sound end identification is carried out using short-time average energy algorithm to signal V；
(5) simultaneously Fast Fourier Transform (FFT) FFT is carried out to three road voice signals；
(6) voice of signal V and the power spectrum of noise are asked for；
(7) ask for the signal of signal V and the frequency domain power ratio of noise, recycle look-up table obtain signal V0 noise reduction coefficient α and
The noise reduction coefficient β of signal V1；
(8) using power spectrum-subtraction：E (V) α E (V0) β E (V1), calculate the power of the signal V after noise reduction；
(9) Fast Fourier Transform Inverse IFFT is carried out to the phase place of the signal V before the power and noise reduction of the signal V after noise reduction, is become
Data value after changing retains real part；
(10) Hamming window is removed；
(11) framing signal is converted to into voice signal, modulation output signal, so as to obtain noise reduction after data signal.
2. automobile speech instruction acquisition processing system according to claim 1, it is characterised in that first noise reduction unit
The broadband voice that model is R262 is with second noise reduction unit catch noise reduction integrated chip.
3. automobile speech instruction acquisition processing system according to claim 1 and 2, it is characterised in that the main channel sound
Sound pick device, the left channel sound pickup device and the right channel sound pickup device are arranged at automobile instrument
On platform；
The main channel sound pickup device is just arranged to driver.
4. automobile speech instruction acquisition processing system according to claim 1 and 2, it is characterised in that the signal transacting
Unit is single-chip microcomputer.
5. a kind of automobile speech instruction acquisition processing method, it is characterised in that include：
Obtain main channel voice signal, left channel voice signal and right channel voice signal；
Analog-to-digital conversion is carried out respectively to the main channel voice signal, left channel voice signal and right channel voice signal,
Obtain main channel data signal, left channel data signal and right channel data signal；
According to the left channel data signal and right channel data signal, the main channel data signal is carried out at noise reduction
Reason, obtains the data signal after noise reduction；
The left channel voice signal and the right channel voice signal are overlapped, the sound letter after being superimposed
Number；
The main channel voice signal is dropped according to the data signal after the voice signal after the superposition and the noise reduction
Make an uproar process, obtain the phonetic order after noise reduction；
Left side and close automobile of the left channel voice signal by the pick device positioned at the main channel voice signal
The pick device pickup of the position of engine, the right channel voice signal is by the pickup positioned at the main channel voice signal
The right side of device and just to the pick device pickup of the position of copilot passenger；
Wherein, it is described according to the left channel data signal and right channel data signal, to the main channel data signal
Noise reduction process is carried out, the data signal after noise reduction is obtained, including：
(1) three railway digital signals are read in, is labeled as V0, V1, V, be corresponding in turn to left channel data signal, right channel number
Word signal and main channel data signal, wherein main channel data signal are from the closest of driver, and voice signal is most obvious；
(2) to three railway digital signal V0, V1, V carry out sub-frame processing simultaneously, and frame length is 256 points, and it is 128 points that frame is moved, and deficiency mends 0；
(3) to three road signal V0, V1, V carries out Hamming filtering；
(4) sound end identification is carried out using short-time average energy algorithm to signal V；
(5) simultaneously Fast Fourier Transform (FFT) FFT is carried out to three road voice signals；
(6) voice of signal V and the power spectrum of noise are asked for；
(7) ask for the signal of signal V and the frequency domain power ratio of noise, recycle look-up table obtain signal V0 noise reduction coefficient α and
The noise reduction coefficient β of signal V1；
(8) using power spectrum-subtraction：E (V) α E (V0) β E (V1), calculate the power of the signal V after noise reduction；
(9) Fast Fourier Transform Inverse IFFT is carried out to the phase place of the signal V before the power and noise reduction of the signal V after noise reduction, is become
Data value after changing retains real part；
(10) Hamming window is removed；
(11) framing signal is converted to into voice signal, modulation output signal, so as to obtain noise reduction after data signal.
6. automobile speech instruction acquisition processing method according to claim 5, it is characterised in that according to the left channel
Data signal and right channel data signal, noise reduction process is carried out by power spectrum-subtraction to the main channel data signal, is obtained
Data signal to after noise reduction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410198669.6A CN103928026B (en) | 2014-05-12 | 2014-05-12 | Automobile voice command acquiring and processing system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410198669.6A CN103928026B (en) | 2014-05-12 | 2014-05-12 | Automobile voice command acquiring and processing system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103928026A CN103928026A (en) | 2014-07-16 |
| CN103928026B true CN103928026B (en) | 2017-04-12 |
Family
ID=51146225
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410198669.6A Active CN103928026B (en) | 2014-05-12 | 2014-05-12 | Automobile voice command acquiring and processing system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103928026B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6754184B2 (en) * | 2014-12-26 | 2020-09-09 | パナソニック インテレクチュアル プロパティ コーポレーション オブ アメリカＰａｎａｓｏｎｉｃ Ｉｎｔｅｌｌｅｃｔｕａｌ Ｐｒｏｐｅｒｔｙ Ｃｏｒｐｏｒａｔｉｏｎ ｏｆ Ａｍｅｒｉｃａ | Voice recognition device and voice recognition method |
| CN107481730B (en) * | 2016-10-14 | 2020-09-18 | 宝沃汽车（中国）有限公司 | Noise reduction method and system for automobile pickup system |
| CN106796790B (en) * | 2016-11-16 | 2020-11-10 | 深圳达闼科技控股有限公司 | Robot voice instruction recognition method and related robot device |
| CN110211579B (en) * | 2019-04-28 | 2021-12-24 | 北京蓦然认知科技有限公司 | Voice instruction recognition method, device and system |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1768555A (en) * | 2003-04-08 | 2006-05-03 | 皇家飞利浦电子股份有限公司 | Method and apparatus for reducing an interference noise signal fraction in a microphone signal |
| CN1841500A (en) * | 2005-03-30 | 2006-10-04 | 松下电器产业株式会社 | Method and apparatus for resisting noise based on adaptive nonlinear spectral subtraction |
| CN1971709A (en) * | 2004-01-07 | 2007-05-30 | 株式会社电装 | Noise elimination system, voice identification system and automobile navigation system |
| CN102637438A (en) * | 2012-03-23 | 2012-08-15 | 同济大学 | Voice filtering method |
| CN102819009A (en) * | 2012-08-10 | 2012-12-12 | 汽车零部件研究及发展中心有限公司 | Driver sound localization system and method for automobile |
| CN102938254A (en) * | 2012-10-24 | 2013-02-20 | 中国科学技术大学 | Voice signal enhancement system and method |
| CN103200496A (en) * | 2012-01-05 | 2013-07-10 | 立锜科技股份有限公司 | Recording device and method for reducing noise |
| CN103533489A (en) * | 2013-10-24 | 2014-01-22 | 安徽江淮汽车股份有限公司 | Vehicle-mounted noise reduction module |
-
2014
- 2014-05-12 CN CN201410198669.6A patent/CN103928026B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1768555A (en) * | 2003-04-08 | 2006-05-03 | 皇家飞利浦电子股份有限公司 | Method and apparatus for reducing an interference noise signal fraction in a microphone signal |
| CN1971709A (en) * | 2004-01-07 | 2007-05-30 | 株式会社电装 | Noise elimination system, voice identification system and automobile navigation system |
| CN1841500A (en) * | 2005-03-30 | 2006-10-04 | 松下电器产业株式会社 | Method and apparatus for resisting noise based on adaptive nonlinear spectral subtraction |
| CN103200496A (en) * | 2012-01-05 | 2013-07-10 | 立锜科技股份有限公司 | Recording device and method for reducing noise |
| CN102637438A (en) * | 2012-03-23 | 2012-08-15 | 同济大学 | Voice filtering method |
| CN102819009A (en) * | 2012-08-10 | 2012-12-12 | 汽车零部件研究及发展中心有限公司 | Driver sound localization system and method for automobile |
| CN102938254A (en) * | 2012-10-24 | 2013-02-20 | 中国科学技术大学 | Voice signal enhancement system and method |
| CN103533489A (en) * | 2013-10-24 | 2014-01-22 | 安徽江淮汽车股份有限公司 | Vehicle-mounted noise reduction module |
Non-Patent Citations (1)
| Title |
|---|
| 《BelaSignaR262：麦克风音频降噪解决方案》;无;《世界电子元器件》;20130630(第6期);第15-16页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103928026A (en) | 2014-07-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103928026B (en) | Automobile voice command acquiring and processing system and method | |
| US9731668B2 (en) | Multi-vehicle settings | |
| CN104332159B (en) | Vehicular voice-operated system man-machine interaction method and device | |
| DE102019107624A1 (en) | System and method for fulfilling a voice request | |
| CN102324035A (en) | Method and system of applying lip posture assisted speech recognition technique to vehicle navigation | |
| CN103471202B (en) | A kind of vehicle intelligent air purification control system based on fuzzy control and method | |
| CN207045104U (en) | A kind of vehicle-mounted glass color change system to change colour with mood | |
| CN105931650B (en) | A kind of self-adaptation noise reduction method based on audio feature extraction | |
| DE112012006617T5 (en) | Onboard information device | |
| DE102017218542A1 (en) | Method for voice control of vehicle functions of a motor vehicle from outside the motor vehicle, roof antenna element for a motor vehicle and motor vehicle with a roof antenna element | |
| CN106347260B (en) | A kind of vehicle control system and its control method based on speech recognition | |
| CN106328133A (en) | Voice response responding device | |
| CN108437993A (en) | A kind of road anger vehicle drive automatically control system and method | |
| CN205445280U (en) | Door window automatic regulating system based on noise measuring and air pollution detect | |
| CN204392434U (en) | A kind of cam device possessing driving record function | |
| DE102014204980A1 (en) | Method and apparatus for restricting or forcibly activating motor vehicle functions | |
| DE102017204895A1 (en) | Method and device for entertaining a user of a vehicle | |
| CN111724608A (en) | System for monitoring an acoustic scene outside a vehicle | |
| CN106427771A (en) | Sound control method and device | |
| CN204537277U (en) | Port Vehicular intelligent checking system | |
| CN109532654A (en) | A kind of engine voice and goers outside vehicle caution sound integrated control system | |
| CN202329640U (en) | System for applying auxiliary voice recognition technology by mouth shape in vehicular navigation | |
| DE102013001118A1 (en) | Method for operating a motor vehicle | |
| CN205022503U (en) | Intelligence car speech control system | |
| CN108099816A (en) | A kind of vehicle audio input control method for intelligent travel |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: 230601 Anhui Province, Hefei City Industrial Park, the Peach Blossom Road No. 669 Applicant after: Anhui Jianghuai Automobile Group Limited by Share Ltd Address before: 230601 Anhui Province, Hefei City Industrial Park, the Peach Blossom Road No. 669 Applicant before: Anhui Jianghuai Automotive Co., Ltd. |
|
| COR | Change of bibliographic data | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |