CN107010009A - A kind of safe accessing method of intelligent automobile voice and device - Google Patents

Abstract

The present invention relates to a kind of safe accessing method of intelligent automobile voice, including：The checking information of default two or more different content, in user's registration, obtains the user speech information of each checking information, and extract the vocal print feature of each voice messaging respectively；When user logins, a checking information is pushed at random logins checking for user, obtain the voice messaging of user's input, and extract the vocal print feature of user, the vocal print feature of extraction vocal print feature corresponding with the checking information pushed is compared, when vocal print feature is compared successfully, then user is allowed to login.By pushing checking information at random, effectively user is avoided to be recorded when logining before by other people, so as to which the situation for occurring to be logined by other people occurs, and logined by vocal print, username and password is carried out without user to login, avoid the occurrence of user to forget user name or login the situation generation of password, it is ensured that the safety of user is logined.

Description

A kind of safe accessing method of intelligent automobile voice and device

Technical field

The present invention relates to intelligent automobile field, the more particularly to a kind of safe accessing method of intelligent automobile voice and device.

Background technology

With continuing to develop for automotive engineering, present automobile is increasingly to energy-conserving and environment-protective direction and intelligent interconnection directionFunction in development, automobile also becomes increasingly complex.And in order to preferably manage these functions, and be easy to interact with driver,The car-mounted terminals such as car running computer would generally be provided with automobile.Logining for the car-mounted terminals such as car running computer, generally passes through controlKey-press input or touch-screen input.

In the prior art, the checking of user is mainly verified by user name with logining password.In some realization sidesIn case, also have by speech recognition to realize the input of control instruction.The scheme of the existing control instruction of present speech recognition is mainFor：Control instruction collection is prestored in systems, when in use, the user speech collected from sound acquisition module, andSystem is digitized processing, extracts the characteristic value of voice, then the condition code that the voice command stored with system is concentrated is carried outMatching verification, identifies order, and rear and other controllers carry out information interchange and perform action.

During the present invention is realized, inventor has found that there are the following problems in the prior art, by user name and stepping onEnter password to carry out checking and login easily password occur and forgetting, user name and generation situations such as login the leakage of password.It is existingIn the method for the Voice command of car-mounted terminal, in fact it could happen that lower user is recorded by way of recording by people logins voice messaging,System is stepped on so as to occur being stolen by people, the thing in turn resulting in user's loss is present.

The content of the invention

For this reason, it may be necessary to provide it is a kind of facilitate user carry out safety login, and avoid simultaneously due to by other people record and causeLogin a kind of safe accessing method of intelligent automobile voice and device of the onboard system of intelligent automobile.

To achieve the above object, a kind of safe accessing method of intelligent automobile voice is inventor provided, including：

The checking information of default two or more different content, in user's registration, obtains the use of each checking information respectivelyFamily voice messaging, and extract the vocal print feature of each voice messaging；

When user logins, a checking information is pushed at random and logins checking for user, the voice of user's input is obtainedInformation, and the vocal print feature of user is extracted, the vocal print feature of extraction vocal print feature corresponding with the checking information pushed is carried outCompare, when vocal print feature is compared successfully, then allow user to login.

Further optimization, described " pushing a checking information at random " is the random checking for pushing two or more different contentChecking information outside the checking information that last time pushes in information.

Further optimization, in user's registration, according to the checking information of default two or more different content, passes through vehicle-mounted masterThe microphone of the sound pick-up of machine or the mobile terminal of user obtains the user speech information of each checking information, and sends to cloudHold server to carry out parsing to each user speech information and extract vocal print feature corresponding with each checking information, and vocal print is specialLevy and store into cloud database.

Further optimization, when user logins, according to the checking information pushed at random, passes through the pickup of on-vehicle hostThe microphone of device or mobile terminal obtains user speech information, and user speech information is parsed by cloud serverVocal print feature is extracted, and by the vocal print feature of extraction vocal print feature ratio corresponding with the checking information pushed in database beyond the cloudsIt is right, when the vocal print feature of extraction is compared successfully, then allow user to login, if comparing failure, refusal user logins.

Further optimization, the vocal print feature code includes acoustic feature, lexical characteristics and prosodic features.

The present inventor additionally provides a kind of intelligent automobile voice of another technical scheme and logins device, including registration list safelyMember, login unit；

The registering unit is used for the checking information for presetting two or more different content, in user's registration, obtains respectivelyThe user speech information of each checking information, and extract the vocal print feature of each voice messaging；

The unit of logining pushes a checking information when user logins, at random and logins checking for user, obtains and usesThe voice messaging of family input, and the vocal print feature of user is extracted, the vocal print feature of extraction is corresponding with the checking information pushedVocal print feature is compared, and when vocal print feature is compared successfully, then allows user to login.

Further optimization, described " pushing a checking information at random " is the random checking for pushing two or more different contentChecking information outside the checking information that last time pushes in information.

Further optimization, the registering unit is used in user's registration, according to testing for default two or more different contentInformation is demonstrate,proved, user's language of each checking information is obtained by the microphone of the sound pick-up of on-vehicle host or the mobile terminal of userMessage ceases, and sends to cloud server to each user speech information progress parsing extraction sound corresponding with each checking informationLine feature, and the corresponding vocal print feature of each checking information is stored into cloud database.

Further optimization, the unit of logining is used for when user logins, and according to the checking information pushed at random, leads toThe microphone of the sound pick-up or mobile terminal of crossing on-vehicle host obtains user speech information, and by cloud server to userVoice messaging carries out parsing and extracts vocal print feature, and by the vocal print feature of extraction and the checking information that is pushed in database beyond the cloudsCorresponding vocal print feature is compared, and when the vocal print feature of extraction is compared successfully, is then allowed user to login, if comparing failure, is refusedUser logins.

Further optimization, the vocal print feature code includes acoustic feature, lexical characteristics and prosodic features.

It is different from prior art, above-mentioned technical proposal is by presetting the checking information of two or more different content, in userDuring registration, according to the checking information of default two or more different content, the user speech letter of each checking information is obtained respectivelyBreath, and vocal print feature is extracted from each voice messaging, and these vocal print features are stored in cloud database, and work as userWhen logining, a checking information is pushed at random to user, so as to obtain user according to the checking information pushed at randomUser speech information, and the vocal print feature that user is extracted in parsing is carried out from the user speech information of acquisition, and will data beyond the cloudsCarry out carrying out retrieval matching to the vocal print feature of extraction in storehouse, when the match is successful for the vocal print feature of extraction, logined.By pushing checking information at random, user is logined according to the checking of push, effectively avoid user login before whenTime is recorded by other people, so that the situation for occurring to be logined by other people occurs, and is logined, is carried out without user by vocal printUsername and password is logined, and according to the uniqueness of vocal print, that is, is avoided the occurrence of user and forgotten user name or login the feelings of passwordCondition occurs, and user security is logined on-vehicle host etc..

Brief description of the drawings

Fig. 1 is a kind of a kind of schematic flow sheet of the safe accessing method of intelligent automobile voice described in embodiment；

Fig. 2 is a kind of structural representation that a kind of intelligent automobile voice described in embodiment logins safely device.

Description of reference numerals：

210th, registering unit,

220th, unit is logined,

230th, cloud server.

Embodiment

To describe the technology contents of technical scheme in detail, feature, the objects and the effects being constructed, below in conjunction with specific realityApply example and coordinate accompanying drawing to be explained in detail.

Referring to Fig. 1, a kind of safe accessing method of intelligent automobile voice of the present embodiment, including：

Step S110：The checking information of default two or more different content, in user's registration, obtains each checking respectivelyThe user speech information of information, and extract the vocal print feature of each voice messaging；

Step S120：When user logins, a checking information is pushed at random and logins checking for user, user is obtained defeatedThe voice messaging entered, and the vocal print feature of user is extracted, by the vocal print feature of extraction vocal print corresponding with the checking information pushedFeature is compared, and when vocal print feature is compared successfully, then allows user to login.

Vocal print not only has specificity, and the characteristics of have relative stability.After adult, the sound of people can keep long-termStablize relatively constant.It is demonstrated experimentally that no matter talker is deliberately to imitate other people sound and the tone, or whisper in sb.'s ear is softly talked, i.e.,Make to imitate remarkably true to life, its vocal print is but differed all the time.And Application on Voiceprint Recognition, one kind of biological identification technology.Application on Voiceprint Recognition is justIt is that acoustical signal is converted into electric signal, then is identified with computer.The generation of human language is Body Languages maincenter and pronunciationA complicated physiology physical process between organ, phonatory organ -- tongue, tooth, larynx, lung, nasal cavity that people uses in speechEveryone is widely different in terms of size and form, so the voiceprint map of any two people is all variant.

By presetting the checking information of two or more different content, in user's registration, according to default two or more notWith the checking information of content, the user speech information of each checking information, and the extraction sound from each voice messaging are obtained respectivelyLine feature, and these vocal print features are stored in cloud database, and when user is logining, pushed at random to userOne checking information, so as to obtain user speech information of the user according to the checking information pushed at random, and from the user of acquisitionVoice messaging carries out the vocal print feature that user is extracted in parsing, and the vocal print feature carried out in database beyond the clouds to extraction is carried outRetrieval matching, when the match is successful for the vocal print feature of extraction, is logined.By pushing checking information, Yong Hugen at randomLogined according to the checking of push, effectively avoid user from being recorded when logining before by other people, so as to possibility quilt occurThe situation that other people login occurs, and is logined by vocal print, and carrying out username and password without user logins, and according to vocal printUniqueness, that is, avoid the occurrence of user and forget user name or login the situation of password occur, and it is vehicle-mounted to login user securityMain frame etc..

For example by preset " open sesame ", " cloud degree is specially enjoyed ", " outstanding cloud degree " three checking informations, respectively according to this threeIndividual checking information obtains the voice messaging of user, and the extraction sound corresponding with each checking information from these user speech informationLine feature, when user logins, pushes a checking information, such as pushes the checking information of " outstanding cloud degree ", obtain at randomThe voice messaging of " the outstanding cloud degree " of user, extracts the vocal print feature of user from the voice messaging, and the vocal print of extraction is specialLevy and be compared with the vocal print feature corresponding to " the outstanding cloud degree " of push, when vocal print feature is compared successfully, then allow userLogin, if comparison is unsuccessful, refusal user logins.

In the present embodiment, in order to avoid there is one group of voice messaging pushing at random and the voice messaging of upper one group of pushIdentical, described " pushing a checking information at random " pushed for last time in the random checking information for pushing two or more different contentA checking information outside checking information.Last time pushes in checking information by pushing two or more different content at randomA checking information outside checking information, exclude the checking information that last time pushes, that is, avoid having recorded last time by other people and loginShi Suoyong checking information, can further be reduced by the situation generation of other people accessing systems.For example preset that " sesame is openedDoor ", " cloud degree is specially enjoyed ", " outstanding cloud degree " three checking informations, and after having pushed checking information " open sesame " before, thisThe checking information of secondary push excludes " open sesame ", the random checking letter pushed in " cloud degree is specially enjoyed " or " outstanding cloud degree "Breath.

In the present embodiment, in user's registration, according to the checking information of default two or more different content, by vehicle-mountedThe microphone of the sound pick-up of main frame or the mobile terminal of user obtains the user speech information of each checking information, and sends extremelyCloud server carries out parsing to each user speech information and extracts corresponding with each checking information vocal print feature, and by vocal printCharacteristic storage is into cloud database.Obtained by the microphone of the sound pick-up of on-vehicle host or the mobile terminal of userVoice messaging corresponding with each checking information, the convenient voice messaging for obtaining user, and using the sound pick-up of on-vehicle hostOr the mobile terminal of user obtain the voice messaging of user, the safety of user's registration can be improved when with registration, andVocal print feature is extracted using the convenient and swift voice messaging to parse user of cloud server.

In the present embodiment, when user logins, according to the checking information pushed at random, picking up for on-vehicle host is passed throughThe microphone of sound device or mobile terminal obtains user speech information, and user speech information is solved by cloud serverVocal print feature is extracted in analysis, and by the vocal print feature of extraction vocal print feature corresponding with the checking information pushed in database beyond the cloudsCompare, when the vocal print feature of extraction is compared successfully, then allow user to login, if comparing failure, refusal user logins.In userWhen logining, not only by the sound pick-up of on-vehicle host can login but also can be stepped on by the mobile terminal of userEnter, the convenience that increase user logins so that user can enter when changing mobile terminal or changing intelligent automobileRow is logined, user-friendly.

In the present embodiment, in order to increase the accurate checking of vocal print feature code, vocal print feature code include acoustic feature,Lexical characteristics and prosodic features.By verifying acoustic feature (cepstrum), lexical characteristics in vocal print feature code, (speaker is relatedWord n-gram, phoneme n-gram), prosodic features (using n-gram describe fundamental tone and energy " posture "), and then cause soundThe checking of line condition code is more accurate.

Referring to Fig. 2, login device safely in a kind of intelligent automobile voice of another embodiment, including registering unit 210,Login unit 220；

The registering unit 210 is used for the checking information for presetting two or more different content, in user's registration, obtains respectivelyThe user speech information of each checking information is taken, and extracts the vocal print feature of each voice messaging；

It is described when logining unit 220 and being logined for user, a checking information is pushed at random and logins checking for user, is obtainedThe voice messaging of family input is taken, and extracts the vocal print feature of user, by the vocal print feature of extraction and the checking information pair of pushThe vocal print feature answered is compared, and when vocal print feature is compared successfully, then allows user to login.

By registering unit 210, the checking information of two or more different content is preset, in user's registration, according to defaultTwo or more different content checking information, obtain the user speech information of each checking information respectively, and from each voiceVocal print feature is extracted in information, and these vocal print features are stored in cloud database, by logining unit 220, and works as and usesFamily pushes a checking information at random when logining, to user, so as to obtain user according to the checking information pushed at randomUser speech information, and carry out parsing from the user speech information of acquisition and extract the vocal print feature of user, and will count beyond the cloudsRetrieval matching is carried out according to the vocal print feature carried out in storehouse to extraction, when the match is successful for the vocal print feature of extraction, is stepped onEnter.By pushing checking information at random, user is logined according to the checking of push, effectively avoids user from logining beforeWhen recorded by other people, occur so as to situation about may be logined by other people occur, and logined by vocal print, enter without userRow username and password is logined, and according to the uniqueness of vocal print, that is, is avoided the occurrence of user and forgotten user name or login passwordSituation occurs, and user security is logined on-vehicle host etc..

In the present embodiment, in order to avoid there is one group of voice messaging pushing at random and the voice messaging of upper one group of pushIdentical, described " pushing a checking information at random " pushed for last time in the random checking information for pushing two or more different contentA checking information outside checking information.Last time pushes in checking information by pushing two or more different content at randomA checking information outside checking information, exclude the checking information that last time pushes, that is, avoid having recorded last time by other people and loginShi Suoyong checking information, can further be reduced by the situation generation of other people accessing systems.For example preset that " sesame is openedDoor ", " cloud degree is specially enjoyed ", " outstanding cloud degree " three checking informations, and after having pushed checking information " open sesame " before, thisThe checking information of secondary push excludes " open sesame ", the random checking letter pushed in " cloud degree is specially enjoyed " or " outstanding cloud degree "Breath.

In the present embodiment, the registering unit 210 is used in user's registration, according to default two or more different contentChecking information, the use of each checking information is obtained by the microphone of the sound pick-up of on-vehicle host or the mobile terminal of userFamily voice messaging, and send corresponding with each checking information to the progress parsing extraction of each user speech information to cloud serverVocal print feature, and vocal print feature is stored into cloud database 230.The sound pick-up that registering unit 210 passes through on-vehicle hostIt is convenient to obtain user's or the microphone of the mobile terminal of user carries out obtaining voice messaging corresponding with each checking informationVoice messaging, and sound pick-up or the voice messaging of the mobile terminal progress acquisition user of user using on-vehicle host,The safety of user's registration can be improved during with registration, and utilizes the convenient and swift voice messaging to parse user of cloud serverExtract vocal print feature.

In the present embodiment, the unit 220 of logining is when user logins, according to the checking information pushed at random,User speech information is obtained by the microphone of the sound pick-up of on-vehicle host or mobile terminal, and by cloud server toFamily voice messaging carries out parsing and extracts vocal print feature, and by the vocal print feature of extraction and testing for being pushed in database 230 beyond the cloudsDemonstrate,prove the corresponding vocal print feature of information to compare, when the vocal print feature of extraction is compared successfully, then allow user to login, if comparing failure,Then refusal user logins.By logining unit 220, when user logins, it can both be entered by the sound pick-up of on-vehicle hostRow is logined and can logined by the mobile terminal of user, the convenience that increase user logins so that user is changing shiftingIt can be carried out logining when dynamic terminal or replacing intelligent automobile, it is user-friendly.

In the present embodiment, in order to increase the accurate checking of vocal print feature code, vocal print feature code include acoustic feature,Lexical characteristics and prosodic features.By verifying acoustic feature (cepstrum), lexical characteristics in vocal print feature code, (speaker is relatedWord n-gram, phoneme n-gram), prosodic features (using n-gram describe fundamental tone and energy " posture "), and then cause soundThe checking of line condition code is more accurate.

Although the various embodiments described above are described, those skilled in the art once know basic woundThe property made concept, then can make other change and modification to these embodiments, so embodiments of the invention are the foregoing is only,Not thereby the scope of patent protection of the present invention, the equivalent structure that every utilization description of the invention and accompanying drawing content are made are limitedOr equivalent flow conversion, or other related technical fields are directly or indirectly used in, similarly it is included in the patent of the present inventionWithin protection domain.

Claims (10)

1. a kind of safe accessing method of intelligent automobile voice, it is characterised in that including：

The checking information of default two or more different content, in user's registration, obtains user's language of each checking information respectivelyMessage ceases, and extracts the vocal print feature of each voice messaging；

When user logins, a checking information is pushed at random and logins checking for user, the voice messaging of user's input is obtained,And the vocal print feature of user is extracted, the vocal print feature of extraction vocal print feature corresponding with the checking information pushed is compared,When vocal print feature is compared successfully, then user is allowed to login.

2. the safe accessing method of intelligent automobile voice according to claim 1, it is characterised in that described " random to push oneTesting outside the checking information that checking information " pushed for last time in the random checking information for pushing two or more different contentDemonstrate,prove information.

3. the safe accessing method of intelligent automobile voice according to claim 1, it is characterised in that in user's registration, according toThe checking information of default two or more different content, passes through the sound pick-up of on-vehicle host or the microphone of the mobile terminal of userThe user speech information of each checking information is obtained, and transmission carries out parsing to each user speech information to cloud server and carriedVocal print feature corresponding with each checking information is taken, and vocal print feature is stored into cloud database.

4. the safe accessing method of intelligent automobile voice according to claim 1, it is characterised in that when user logins, according toThe checking information pushed at random, obtains user speech by the sound pick-up of on-vehicle host or the microphone of mobile terminal and believesBreath, and by cloud server to user speech information carry out parsing extract vocal print feature, and by the vocal print feature of extraction withThe corresponding vocal print feature of checking information pushed in cloud database is compared, and when the vocal print feature of extraction is compared successfully, is then allowedUser logins, if comparing failure, refusal user logins.

5. the safe accessing method of intelligent automobile voice according to claim 1, it is characterised in that the vocal print feature code includesAcoustic feature, lexical characteristics and prosodic features.

6. a kind of intelligent automobile voice logins safely device, it is characterised in that including registering unit, login unit；

The registering unit is used for the checking information for presetting two or more different content, in user's registration, obtains respectively eachThe user speech information of checking information, and extract the vocal print feature of each voice messaging；

The unit of logining pushes a checking information when user logins, at random and logins checking for user, obtains user defeatedThe voice messaging entered, and the vocal print feature of user is extracted, by the vocal print feature of extraction vocal print corresponding with the checking information pushedFeature is compared, and when vocal print feature is compared successfully, then allows user to login.

7. intelligent automobile voice logins safely device according to claim 6, it is characterised in that described " random to push oneTesting outside the checking information that checking information " pushed for last time in the random checking information for pushing two or more different contentDemonstrate,prove information.

8. intelligent automobile voice logins safely device according to claim 6, it is characterised in that the registering unit is used forDuring user's registration, according to the checking information of default two or more different content, pass through the sound pick-up of on-vehicle host or userThe microphone of mobile terminal obtains the user speech information of each checking information, and sends to cloud server to each user's languageMessage breath carries out parsing and extracts corresponding with each checking information vocal print feature, and by the corresponding vocal print feature of each checking informationStore into cloud database.

9. intelligent automobile voice logins safely device according to claim 6, it is characterised in that the unit of logining is used forWhen user logins, according to the checking information pushed at random, sound pick-up or the Mike of mobile terminal by on-vehicle hostWind obtains user speech information, and parsing extraction vocal print feature is carried out to user speech information by cloud server, and will carryThe vocal print feature taken vocal print feature corresponding with the checking information pushed in database beyond the clouds is compared, when the vocal print feature of extractionCompare successfully, then allow user to login, if comparing failure, refusal user logins.

10. intelligent automobile voice logins safely device according to claim 6, it is characterised in that the vocal print feature code bagInclude acoustic feature, lexical characteristics and prosodic features.