CN102815279B - Voice based on embedded system directly start automobile and antitheft method and device - Google Patents
Voice based on embedded system directly start automobile and antitheft method and device Download PDFInfo
- Publication number
- CN102815279B CN102815279B CN201110154304.XA CN201110154304A CN102815279B CN 102815279 B CN102815279 B CN 102815279B CN 201110154304 A CN201110154304 A CN 201110154304A CN 102815279 B CN102815279 B CN 102815279B
- Authority
- CN
- China
- Prior art keywords
- voice
- automobile
- unit
- identification
- feature
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 43
- 238000000605 extraction Methods 0.000 claims description 10
- 230000008878 coupling Effects 0.000 claims description 9
- 238000010168 coupling process Methods 0.000 claims description 9
- 238000005859 coupling reaction Methods 0.000 claims description 9
- 238000012549 training Methods 0.000 claims description 8
- 239000000203 mixture Substances 0.000 claims description 6
- 238000001228 spectrum Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 5
- 238000004458 analytical method Methods 0.000 claims description 4
- 238000001514 detection method Methods 0.000 claims description 4
- 238000004422 calculation algorithm Methods 0.000 claims description 3
- 238000013139 quantization Methods 0.000 claims description 3
- 238000012706 support-vector machine Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 2
- 230000001537 neural effect Effects 0.000 claims 2
- 238000005516 engineering process Methods 0.000 description 13
- 230000000694 effects Effects 0.000 description 9
- 238000006243 chemical reaction Methods 0.000 description 8
- 230000006870 function Effects 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 3
- 238000005070 sampling Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000009499 grossing Methods 0.000 description 2
- 238000011017 operating method Methods 0.000 description 2
- 239000004821 Contact adhesive Substances 0.000 description 1
- 239000000853 adhesive Substances 0.000 description 1
- 230000001070 adhesive effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000019771 cognition Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000030279 gene silencing Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
Landscapes
- Fittings On The Vehicle Exterior For Carrying Loads, And Devices For Holding Or Mounting Articles (AREA)
Abstract
The invention discloses a kind of voice based on embedded system and directly start automobile and antitheft method and device. Wherein voice message order is sent to voice alerting unit by central control unit, passes through loudspeaker play cuing voice after voice alerting unit receives orders; User hears the voice command of saying original typing after suggestion voice, and voice collecting unit is converted to analog electrical signal by voice command and delivers to audio coder & decoder (codec), and audio coder & decoder (codec) carries out being converted to data signal after sample code to it and delivers to central control unit; Central control unit carries out computing and contrast identification to the voice digital signal of input; Identification by after the order of startups automobile is sent to automobile starting control module, automobile starting unit receive this order after start up the car. The present invention, using embedded system as implementation platform, utilizes the uniqueness of Speech acoustics feature to realize car anti-theft function.
Description
Technical field
The present invention relates to the startup of automobile and antitheft, specifically the voice based on embedded system are straightConnect and start automobile and antitheft method and device.
Background technology
Embedded system is application-centered, and taking computer technology as basis, software and hardware can cutting, suitableFor application system, function, reliability, cost, volume, power consumption are had to the dedicated computing of strict demandMachine system. It has, and volume is little, specificity is strong, reliability is high, real-time, execution speed is fast, can expandThe advantages such as malleability is good. It is generally by microprocessor (MPU) or microcontroller (MCU), peripheral hardwareA few part compositions such as equipment, embedded OS, specific application program. Utilize the embedded system canWith implementation pattern identification preferably.
Mode identification technology is the basic technology of artificial intelligence, wherein speech recognition technology, biological identification skillArt and sound groove recognition technology in e are technology important in pattern-recognition. Due to the traditional approach of automobile startingBe to utilize key operatedly, it easily loses, easily repeat and the shortcoming of transreplication makes to utilize mode identification technologyBe convenient for carrying, the advantage such as uniqueness and confidentiality obtained manifesting fully, mode identification technology makes carThe main constraint of having broken away from key, utilizes self exclusive biological characteristic realization startup automobile and automobile burglarFunction.
Automotive theft proof system is essential safety means on present each automobile substantially. Existing antitheftSystem majority is just reported to the police, and cracksman is played to the effect of warning, but does not fundamentally solve carStolen, and rate of false alarm is very high, horn for vehicle sound, setting off firecrackers and thundering etc. all to make its mistakeReport to the police. Some systems can be informed warning message to car in the mode of information or phone by GSM networkingMain, once but rate of false alarm is high, can affect car owner's normal Working Life, cause when vehicle really stolen, car owner but thinks false alarm, does not go to comprehend. Once have car owner by key loss, other people just againCan very easily car be driven away. These burglary-resisting installations fundamentally do not solve stolen the asking of automobileTopic.
Based on this, pass through Application on Voiceprint Recognition with regard to a kind of voice that can utilize embedded system of necessary designTechnology realizes automobile burglar and directly starts the control device of automobile.
Summary of the invention
The object of the invention is the defect existing for prior art, provide based on embedded systemVoice directly start automobile and antitheft method and device, utilize the uniqueness of Speech acoustics feature to realizeCar anti-theft function. The method and device, using embedded system as implementation platform, utilize special pickupDevice picks up speaker's voice, by mode identification technology analysis authentication speaker's identity, and then judgementWhether directly start automobile.
The technical solution adopted for the present invention to solve the technical problems is: based on the language of embedded systemSound directly starts automobile and antitheft method, the steps include:
1), central control unit is sent to voice alerting unit by voice message order, voice messageAfter receiving orders, passes through unit loudspeaker play cuing voice;
2), user hears the voice command of saying original typing after suggestion voice, voice collecting unitVoice command is converted to analog electrical signal and delivers to audio coder & decoder (codec), audio coder & decoder (codec) enters itAfter row sample code, be converted to data signal and deliver to central control unit;
3), central control unit carries out computing to the voice digital signal of input, and and storesIn unit, speech model contrasts identification;
4), identification by after the order of startup automobile is sent to automobile starting control module, automobile opensMoving cell starts up the car after receiving this order.
Voice based on embedded system directly start in automobile and antitheft method, also comprise remote controlReceiving element is decoded after receiving the wireless signal that remote controller sends, then by decoded lifeOrder is sent to central control unit.
The technical solution adopted for the present invention to solve the technical problems is: based on the language of embedded systemSound directly starts automobile and antitheft device, and its structure comprises remote controller, voice collecting unit, raisesThe automobile starting controller of sound device, connection power supply; Wherein automobile starting controller comprises center-controlUnit, audio coder & decoder (codec), remote control reception unit, voice alerting unit; Described center-control listUnit respectively by remote control reception unit, audio coder & decoder (codec), voice alerting unit be connected to remote controller,Voice collecting unit, loudspeaker; Described automobile starting controller also comprises and central control unit phaseThe automobile starting control module, the memory cell that connect.
Described central control unit adopts embedded microprocessor, embedded microcontroller, numeral letterNumber processor one of them.
Embedded microprocessor adopts the chip of S3C2410 or S3C2440 series; Embedded micro-controlDevice processed adopt STM32 series single-chip microcomputer, PIC32 series monolithic, LPC21XX, LPC22XX,LPC23XX series monolithic; Digital signal processor can adopt TMS320C5000,TMS320C6000 family chip.
Described memory cell adopts eeprom memory and SDRAM memory.
Described automobile starting control module is make-and-break ignition circuit or triode electronic ignition electricityRoad.
Described audio coder & decoder (codec) adopts stereo audio encoding/decoding chip TLV320AIC23.
Described remote control reception unit adopts 304M receiver module.
It is ISD4003-04M that described voice alerting unit adopts chip.
Voice based on embedded system of the present invention directly start automobile and antitheft method and deviceCompared to the prior art, there is following beneficial effect:
When traditional automobile starting, adopt mechanical key or electron key, they just have a common shortcomingBe to be easily replicated, just can start up the car as long as anyone takes key. The inventive method and device are adoptedWith sound groove recognition technology in e, self exclusive biological characteristic of employing, there is uniqueness, be not easy to be replicated,Can not lose, need not remember, the advantage such as easy to use.
Central control unit adopts embedded microprocessor, embedded microcontroller, digital signal processorDeng one of them composition. Embedded microprocessor can adopt Samsung ARM9 processor S3C2410 orThe chip of S3C2440 series. Embedded microcontroller can adopt single-chip microcomputer, the PIC32 of STM32 seriesLPC21XX, the LPC22XX of series monolithic, NXP, LPC23XX series monolithic. Digital Signal ProcessingDevice can adopt TMS320C5000, the TMS320C6000 family chip of TI company. Central control unitThe core of whole system, so the chip of above-mentioned employing is all stable performance, most widely used coreSheet, can meet instructions for use completely, has the advantages such as low-power consumption, processing speed be fast. Center-control listThe effect of unit is that speech signal is carried out to calculation process, draws recognition result, while coordinated control systemModules unit.
Memory cell adopts eeprom memory and SDRAM memory, for storing application program, voiceSome variablees and flag bit in model and program operation process. SDRAM and central control unit adopt parallelData-interface, the execution efficiency of raising program.
Voice collecting unit adopts highly sensitive sound pick-up, has low noise, low distortion, volume unskilled labourerMake the features such as voltage range is wide. Its effect is that voice command is converted to analog electrical signal.
Audio coder & decoder (codec) adopts TI company's T LV320AIC23, a high-performance that Ta Shi TI company releasesStereo audio encoding/decoding chip, its analog-to-digital conversion and digital-to-analogue conversion component height are integrated in chip internal,Adopt advanced Sigma-delta oversampling technique, can in the frequency range of 96K, provide at 8K16bit, 20bit, the sampling of 24bit and 32bit, the signal to noise ratio of analog-to-digital conversion and digital-to-analogue conversion respectively canReach 90DB and 100DB. The multichannel buffered serial port seamless link that it can follow DSP, will be through decodingAfter the real-time DSP that is transferred to of audio digital signals process.
Voice cue module adopts ISD4003-04M chip. 4 minutes this monolithic recording playback time, tonequality is good,Adopt CMOS technology, include oscillator, anti-aliasing filter, smoothing filter, audio-frequency amplifier, fromSound is made an uproar and many electrical level flash of high density storage array, and the acoustic information of recording can be preserved under powering-off state100 years (representative value), records 100,000 times repeatedly. It act as provides a kind of Man Machine Interface, promptingThe duty of current system and operating procedure.
Remote control reception unit adopts PT2272 wireless decoding chip, by the wireless encoding chip of remote controller transmittingDecode, then deliver to central control unit. System is provided with remote control cipher, inputs close by remote controllerAfter code, just can carry out corresponding operation. Can register voice password, cancellation and recovery voice by remote controllerRecognition function, directly start automobile, Modify password.
Automobile starting control module adopts different circuit to realize for different automobile types. For common pointFire switch, the normally opened contact of relay is concatenated in ignition lead, voice password identification pass through after by relayThe adhesive of device contact, ignition lead is connected, and automobile starts immediately. For the electronic type ignition switch of luxury car andThe start button of intelligent automobile, controls the point in electronic type ignition switch or in start button with triodeFire signal, voice password identification by after make triode conducting, thereby output ignition signal, automobile is immediatelyStart. The effect of automobile starting control module is to receive after the startup automobile order that central controller sends,Control the startup system of automobile and start up the car, reach voice command and directly start the effect of automobile, saveThe step of manually twisting key or pressing start button, easy to use.
Brief description of the drawings
Fig. 1 is that the voice based on embedded system of the present invention directly start automobile and antitheft deviceStructure flow chart;
Fig. 2 is that the voice based on embedded system of the present invention directly start automobile and antitheft deviceModel training part flow chart;
Fig. 3 is that the voice based on embedded system of the present invention directly start automobile and antitheft deviceIdentification division flow chart.
Detailed description of the invention
Next with reference to Figure of description, the voice based on embedded system of the present invention are directly openedElectrical automobile and antitheft method and device are described in detail below.
Voice based on embedded system directly start automobile and antitheft method, the steps include:
1), central control unit is sent to voice alerting unit by voice message order, voice messageAfter receiving orders, passes through unit loudspeaker play cuing voice;
2), user hears the voice command of saying original typing after suggestion voice, voice collecting unitVoice command is converted to analog electrical signal and delivers to audio coder & decoder (codec), audio coder & decoder (codec) enters itAfter row sample code, be converted to data signal and deliver to central control unit;
3), central control unit carries out computing to the voice digital signal of input, and and storesIn unit, speech model contrasts identification;
4), identification by after the order of startup automobile is sent to automobile starting control module, automobile opensMoving cell starts up the car after receiving this order.
Voice based on embedded system directly start in automobile and antitheft method, also comprise remote controlReceiving element is decoded after receiving the wireless signal that remote controller sends, then by decoded lifeOrder is sent to central control unit.
Voice based on embedded system directly start automobile and antitheft device, and its structure comprises distantThe automobile starting controller of control device, voice collecting unit, loudspeaker, connection power supply; Wherein automobileStart controller and comprise that central control unit, audio coder & decoder (codec), remote control reception unit, voice carryShow unit; Described central control unit is respectively by remote control reception unit, audio coder & decoder (codec), languageSound Tip element is connected to remote controller, voice collecting unit, loudspeaker; Described automobile starting controlDevice also comprises the automobile starting control module, the memory cell that are connected with central control unit.
Central control unit adopts embedded microprocessor, embedded microcontroller, data signal placeOne of them composition such as reason device. Embedded microprocessor can adopt the ARM9 processor of SamsungThe chip of S3C2410 or S3C2440 series. Embedded microcontroller can adopt STM32 seriesSingle-chip microcomputer, PIC32 series monolithic, LPC21XX, the LPC22XX of NXP, LPC23XX seriesSingle-chip microcomputer. Digital signal processor can adopt TMS320C5000, the TMS320C6000 of TI companyFamily chip. Central control unit is the core of whole system, so the chip of above-mentioned employingBe all stable performance, most widely used chip, can meet instructions for use completely, have low-power consumption,The advantages such as processing speed is fast. The effect of central control unit is that speech signal is carried out to calculation process,Draw recognition result, simultaneously the modules unit of coordinated control system.
Memory cell adopts eeprom memory and SDRAM memory, for store application program,Some variablees and flag bit in speech model and program operation process. SDRAM and center-control listUnit adopts parallel data grabbing card, improves the execution efficiency of program.
Voice collecting unit adopts highly sensitive sound pick-up, has low noise, low distortion, volumeThe features such as little operating voltage range is wide. Its effect is that voice command is converted to analog electrical signal.
Audio coder & decoder (codec) adopts TI company's T LV320AIC23, a height that Ta Shi TI company releasesThe stereo audio encoding/decoding chip of performance, its analog-to-digital conversion and digital-to-analogue conversion component height are integrated inChip internal, has adopted advanced Sigma-delta oversampling technique, can arrive 96K's at 8KIn frequency range, provide 16bit, 20bit, the sampling of 24bit and 32bit, analog-to-digital conversion and digital-to-analogueThe signal to noise ratio of conversion can reach respectively 90DB and 100DB. It can be with the multichannel buffer of DSPSerial ports seamless link, by through decoded audio digital signals real-time be transferred to DSP placeReason.
Voice cue module adopts ISD4003-04M chip. 4 minutes this chip recording playback time, soundMatter is good, adopts CMOS technology, includes oscillator, anti-aliasing filter, smoothing filter, audio frequencyAmplifier, automatic silencing and many electrical level flash of high density storage array, the acoustic information of recording can beUnder powering-off state, preserve 100 years (representative value), repeatedly record 100,000 times. It act as provides a kind ofMan Machine Interface, duty and the operating procedure of prompting current system.
Remote control reception unit adopts PT2272 wireless decoding chip, by the wireless encoding of remote controller transmittingChip is decoded, and then delivers to central control unit. System is provided with remote control cipher, by remote controlAfter device input password, just can carry out corresponding operation. Can register voice password, get by remote controllerSlake recovers speech identifying function, directly starts automobile, Modify password.
Automobile starting control module adopts different circuit to realize for different automobile types. For commonIgnition switch, the normally opened contact of relay is concatenated in ignition lead, voice password identification is passed throughAfter by relay contact adhesive, ignition lead connect, automobile starts immediately. For the electronics of luxury carThe start button of formula ignition switch and intelligent automobile, controls in electronic type ignition switch with triodeOr ignition signal in start button, voice password identification by after make triode conducting, thereby defeatedGo out ignition signal, automobile starts immediately. The effect of automobile starting control module is to receive central authorities' controlAfter the startup automobile order that device processed sends, the startup system of controlling automobile starts up the car, and reaches languageSound order directly starts the effect of automobile, saves manually to twist key or by the step of start button,Easy to use.
System detailed operation process:
If use first this system, when start, there is voice message to need advanced lang sound passwordRegistration. Carry out the typing of six same voice passwords according to the prompt tone of system, then system is to recordThe voice password entering carries out calculation process and sets up model of cognition, system in processing procedure and while endSystem has voice message. After prompting is succeeded in registration, illustrate that model has been successfully established and has been saved in to depositIn storage unit. After system is restarted, can use normally voice password directly to start automobile.
When normal use, first open automobile power source, say voice hearing after the prompt tone of systemPassword, system is analyzed identification to voice password, and after identification is confirmed, system is started immediatelyAutomobile. If identification is not said voice password again by pointing out, if continuous three times are not passed through,Will restart system, not pass through for continuous six times, system can lock, and only has by remote controller and inputsPassword just can release.
After inputting password by remote controller, can carry out function setting to system. The function of setting has verticalStart up the car, cancel/recover function of voice identification, Modify password, re-register voice mouthOrder.
Software algorithm part:
The sound groove recognition technology in e that focuses on of the present invention. Utilize sound pick-up and speech coder to speakingPerson's voice are sampled and are quantized, and digitized phonetic entry is arrived to embedded system device, itsProcessing procedure mainly comprises pretreatment, feature extraction, model training, coupling identification and decision-making judgementDeng five part compositions.
Preprocessing part comprises voice signal is carried out to end-point detection, preemphasis, point frame, windowing etc.Four parts.
Feature extraction part is carried out time domain or frequency-domain analysis to the speech frame after windowing, and uses phaseThe characteristic parameter of answering is described the feature of whole voice signal, uses the characteristic parameter institute of each frame voiceThe argument sequence of composition is described. For speech recognition and Speaker Identification, choosing to closing of feature is heavyWant. The feature that can characterize speaker's individual character has: short-time energy, short-time average magnitude, mistake in short-termZero rate, in short-term pitch period and fundamental frequency, linear predictor coefficient (LPC), coefficient of part correlation(PARCOR) feature, line spectrum pair (LSP) feature, short-term spectrum, formant frequency and bandwidth,Cepstrum feature, Mel-cepstrum coefficient (MFCC) and 2 dimension Mel-cepstrum coefficients etc. The present invention adoptsBe Mel-cepstrum coefficient (MFCC).
Model training be by speech parameter after feature extraction through certain algorithm for each treatsThe speaker of identification trains the parameter matching with it. Main pattern drill method has: dynamicallyTime alignment method (DTW), vector quantization method (VQ), HMM method (HMM), artificialNeural network (ANN), SVMs method (SVM), Wavelet Transform (WT) etc. ThisBright employing be HMM method (HMM).
Coupling identification division can adopt Furthest Neighbor or similarity method.
Decision-making judgment part is to preset a threshold value, by the value and the threshold value that draw after coupling identificationCompare, if be less than, think validated user, otherwise be illegal.
If it is legal that speaker's voice are identified as, system output control signal triggering automobile directly opensMoving. Otherwise automobile does not start.
Be illustrated in figure 2 model training part flow chart. First to carry out initially the parameter of systemSet, then by sampling, quantize after voice signal carry out end-point detection, remove quiet part andOf short duration noise is input to useful voice signal after preemphasis, point frame, windowing processCharacteristic extracting module is calculated MFCC cepstrum parameter and difference MFCC cepstrum parameter, by multiple MFCC spiesLevy parameter and carry out cluster and obtain initial HMM parameter, utilize Baum-Welch method to initial HMM parameterCarry out interative computation, until meet after iterated conditional using the HMM parameter value calculating as finallyModel parameter, and input store saving result.
Be illustrated in figure 3 identification division flow chart, speaker only need again say and record password, warpCross pretreatment and feature extraction part obtains MFCC parameter, calculate this characteristic parameter by Viterbi methodWith the similarity of template matches, utilize matching criterior to judge whether speaker is validated user, ifBe can by and start automobile, otherwise will not pass through.
The above embodiment, is the one of the more preferably concrete embodiment of the present invention, thisThe common variation that the technical staff in field carries out within the scope of technical solution of the present invention and replacement all should be wrappedBe contained in protection scope of the present invention.
Claims (2)
1. the voice based on embedded system directly start automobile and antitheft method, the steps include:
1), central control unit is sent to voice alerting unit by voice message order, voice message listAfter receiving orders, unit passes through loudspeaker play cuing voice;
2), user hears the voice command of saying original typing after suggestion voice, voice collecting unit willVoice command is converted to analog electrical signal and delivers to audio coder & decoder (codec), and audio coder & decoder (codec) is adopted itAfter sample coding, be converted to data signal and deliver to central control unit;
Decoding after receiving the wireless signal that remote controller sends in remote control reception unit, then will decodeAfter order be sent to central control unit;
3), central control unit carries out computing to the voice digital signal of input, and and stores singleIn unit, speech model contrasts identification;
4), identification by after the order of startup automobile is sent to automobile starting control module, automobile startingUnit starts up the car after receiving this order;
Wherein, the processing of described embedded system comprise pretreatment, feature extraction, model training,Coupling identification and decision-making judge five part compositions;
Preprocessing part comprises voice signal is carried out to end-point detection, preemphasis, point frame, windowing fourPart;
Feature extraction part is carried out time domain or frequency-domain analysis to the speech frame after windowing, and uses phaseThe characteristic parameter of answering is described the feature of whole voice signal, uses the characteristic parameter institute group of each frame voiceThe argument sequence becoming is described, and for speech recognition and Speaker Identification, characterizes the feature of speaker's individual characterHave: short-time energy, short-time average magnitude, short-time zero-crossing rate, in short-term pitch period and fundamental frequency,Linear predictor coefficient (LPC), coefficient of part correlation (PARCOR) feature, line spectrum pair (LSP) spyLevy, short-term spectrum, formant frequency and bandwidth, cepstrum feature, Mel-cepstrum coefficient (MFCC) and2 dimension Mel-cepstrum coefficients;
Model training be by speech parameter after feature extraction through certain algorithm for each treatsThe speaker of identification trains the parameter matching with it, and pattern drill method has: dynamic time warpingMethod (DTW), vector quantization method (VQ), HMM method (HMM), artificial neural networkMethod (ANN), SVMs method (SVM), Wavelet Transform (WT);
Coupling identification division adopts Furthest Neighbor or similarity method;
Decision-making judgment part is to preset a threshold value, by the value and the threshold value that draw after coupling identificationCompare, if be less than, think validated user, otherwise be illegal;
If it is legal that speaker's voice are identified as, system output control signal triggering automobile directly starts,Otherwise automobile does not start.
2. the voice based on embedded system directly start automobile and antitheft device, it is characterized in that:Its structure comprises the automobile starting controller of remote controller, voice collecting unit, loudspeaker, connection power supply;Wherein automobile starting controller comprise central control unit, audio coder & decoder (codec), remote control reception unit,Voice alerting unit; Described central control unit respectively by remote control reception unit, audio coder & decoder (codec),Voice alerting unit is connected to remote controller, voice collecting unit, loudspeaker; Described automobile starting controlDevice also comprises the automobile starting control module, the memory cell that are connected with central control unit; Described central authoritiesControl module adopt embedded microprocessor, embedded microcontroller, digital signal processor wherein itOne; Embedded microprocessor adopts the chip of S3C2410 or S3C2440 series; Embedded microcontrollerDevice adopts single-chip microcomputer, PIC32 series monolithic, LPC21XX, LPC22XX, the LPC23XX of STM32 seriesSeries monolithic; Digital signal processor can adopt TMS320C5000, TMS320C6000 seriesChip; Described memory cell adopts eeprom memory and SDRAM memory; Described automobile startingControl module is make-and-break ignition circuit or triode electronic igniting circuit; Described audio coder & decoder (codec)Adopt stereo audio encoding/decoding chip TLV320AIC23; Described remote control reception unit adopts 304M to connectReceive module; It is ISD4003-04M that described voice alerting unit adopts chip;
The processing of embedded system device comprises pretreatment, feature extraction, model training, coupling knowledgeOther and decision-making judges;
Pretreatment comprises carries out end-point detection, preemphasis, point frame, windowing four parts to voice signal;
Feature extraction is carried out time domain or frequency-domain analysis to the speech frame after windowing, and with correspondingCharacteristic parameter is described the feature of whole voice signal, forms with the characteristic parameter of each frame voiceArgument sequence is described, and for speech recognition and Speaker Identification, the feature that characterizes speaker's individual character has:Short-time energy, short-time average magnitude, short-time zero-crossing rate, in short-term pitch period and fundamental frequency, linearityPredictive coefficient (LPC), coefficient of part correlation (PARCOR) feature, line spectrum pair (LSP) feature,Short-term spectrum, formant frequency and bandwidth, cepstrum feature, Mel-cepstrum coefficient (MFCC) and 2 dimensionsMel-cepstrum coefficient;
Model training be by speech parameter after feature extraction through certain algorithm for each treatsThe speaker of identification trains the parameter matching with it, and pattern drill method has: dynamic time warpingMethod (DTW), vector quantization method (VQ), HMM method (HMM), artificial neural networkMethod (ANN), SVMs method (SVM), Wavelet Transform (WT);
Coupling identification adopts Furthest Neighbor or similarity method;
Decision-making judgement is to preset a threshold value, and the value drawing after identifying by coupling is compared with threshold value, if be less than, think validated user, otherwise be illegal; If being identified as, close speaker's voiceMethod, system output control signal triggering automobile directly starts, on the contrary automobile does not start.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110154304.XA CN102815279B (en) | 2011-06-10 | 2011-06-10 | Voice based on embedded system directly start automobile and antitheft method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110154304.XA CN102815279B (en) | 2011-06-10 | 2011-06-10 | Voice based on embedded system directly start automobile and antitheft method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102815279A CN102815279A (en) | 2012-12-12 |
| CN102815279B true CN102815279B (en) | 2016-05-04 |
Family
ID=47299836
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110154304.XA Expired - Fee Related CN102815279B (en) | 2011-06-10 | 2011-06-10 | Voice based on embedded system directly start automobile and antitheft method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102815279B (en) |
Families Citing this family (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103903091B (en) * | 2014-03-19 | 2017-10-03 | 江苏科技大学 | A kind of control method of the vehicle ignition control device based on cloud computing platform |
| CN104732972B (en) * | 2015-03-12 | 2018-02-27 | 广东外语外贸大学 | A kind of HMM Application on Voiceprint Recognition based on classified statistics is registered method and system |
| CN106161727A (en) * | 2015-03-24 | 2016-11-23 | 中兴通讯股份有限公司 | The typing of a kind of verbal instructions, recognition methods and device |
| CN104835495B (en) * | 2015-05-30 | 2018-05-08 | 宁波摩米创新工场电子科技有限公司 | A kind of high definition speech recognition system based on low-pass filtering |
| CN105336325A (en) * | 2015-09-25 | 2016-02-17 | 百度在线网络技术（北京）有限公司 | Speech signal recognition and processing method and device |
| CN105227555B (en) * | 2015-09-30 | 2020-06-12 | 广州汽车集团股份有限公司 | Equipment starting method and system |
| CN106128458A (en) * | 2016-08-31 | 2016-11-16 | 成都市和平科技有限责任公司 | A kind of home voice control system based on speech recognition technology and method |
| CN108877146A (en) * | 2018-09-03 | 2018-11-23 | 深圳市尼欧科技有限公司 | It is a kind of that safety automatic-alarming devices and methods therefor is driven based on multiplying for intelligent sound identification |
| TWI691415B (en) * | 2018-11-30 | 2020-04-21 | 宏碁股份有限公司 | Vehicle control system, controller and mobile device |
| CN111402897A (en) * | 2018-12-13 | 2020-07-10 | 长春财经学院 | Voiceprint recognition control cabinet |
| WO2021127975A1 (en) * | 2019-12-24 | 2021-07-01 | 广州国音智能科技有限公司 | Voiceprint detection method, apparatus and device for sound acquisition object |
| CN112966528B (en) * | 2021-03-01 | 2023-09-19 | 郑州铁路职业技术学院 | English speech translation fuzzy matching system |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4856072A (en) * | 1986-12-31 | 1989-08-08 | Dana Corporation | Voice actuated vehicle security system |
| AU703597B2 (en) * | 1994-03-18 | 1999-03-25 | Nuance Communications, Inc. | Speech controlled vehicle alarm system |
| JP2004278225A (en) * | 2003-03-18 | 2004-10-07 | Nissan Motor Co Ltd | Control device for vehicular opening/closing body |
| JP4375428B2 (en) * | 2007-04-09 | 2009-12-02 | 株式会社デンソー | In-vehicle voice guidance device |
| CN101323305A (en) * | 2008-05-14 | 2008-12-17 | 奇瑞汽车股份有限公司 | Vehicle-mounted speech recognition control system and control method |
| CN201856736U (en) * | 2010-10-18 | 2011-06-08 | 长安大学 | Sound recognition anti-theft device for automobile |
| CN202124017U (en) * | 2011-06-10 | 2012-01-25 | 沈阳君天科技股份有限公司 | Device for directly starting automobile by voice and preventing burglary based on embedded system |
-
2011
- 2011-06-10 CN CN201110154304.XA patent/CN102815279B/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN102815279A (en) | 2012-12-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102815279B (en) | Voice based on embedded system directly start automobile and antitheft method and device | |
| CN202124017U (en) | Device for directly starting automobile by voice and preventing burglary based on embedded system | |
| US10381021B2 (en) | Robust feature extraction using differential zero-crossing counts | |
| CN201998942U (en) | Theftproof control device with sound control function for automobile | |
| US9721560B2 (en) | Cloud based adaptive learning for distributed sensors | |
| US9785706B2 (en) | Acoustic sound signature detection based on sparse features | |
| US9412373B2 (en) | Adaptive environmental context sample and update for comparing speech recognition | |
| US6671669B1 (en) | combined engine system and method for voice recognition | |
| US9460720B2 (en) | Powering-up AFE and microcontroller after comparing analog and truncated sounds | |
| TWI584270B (en) | Voice control system and method thereof | |
| CN105448294A (en) | Intelligent voice recognition system for vehicle equipment | |
| CN101393660A (en) | Intelligent gate inhibition system based on footstep recognition | |
| JP2004527006A (en) | System and method for transmitting voice active status in a distributed voice recognition system | |
| DE69531861D1 (en) | VOICE-CONTROLLED VEHICLE ALARM SYSTEM | |
| JP2002536692A (en) | Distributed speech recognition system | |
| CN112093602B (en) | Elevator voice control method and device | |
| CN210489237U (en) | Vehicle-mounted intelligent terminal voice control system | |
| CN208225136U (en) | Unlocking device for non-motor vehicle | |
| CN110539721A (en) | vehicle control method and device | |
| CN111145763A (en) | GRU-based voice recognition method and system in audio | |
| CN112951243A (en) | Voice awakening method, device, chip, electronic equipment and storage medium | |
| CN204791241U (en) | Voice interaction formula access control system | |
| WO2019169685A1 (en) | Speech processing method and device and electronic device | |
| CN114121004B (en) | Voice recognition method, system, medium and equipment based on deep learning | |
| CN101350196A (en) | On-chip system for confirming role related talker identification and confirming method thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| DD01 | Delivery of document by public notice |
Addressee: SHENYANG JUNTIAN TECHNOLOGY Co.,Ltd. Person in charge of patents Document name: payment instructions |
|
| DD01 | Delivery of document by public notice | ||
| DD01 | Delivery of document by public notice |
Addressee: SHENYANG JUNTIAN TECHNOLOGY Co.,Ltd. Person in charge of patents Document name: Notice of Termination of Patent Rights |
|
| DD01 | Delivery of document by public notice | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160504 |