Dynamic voice label information query system and information query method thereof
Technical field
The present invention relates to the technology of information inquiry, particularly relate to a kind of dynamic voice label information query system andThe technology of information query method.
Background technology
In the time of steering vehicle, if wanting just can access internet in car, user inquires relevant information (such as skyGas, stock information, real-time traffic, periphery interest point information etc.), from convenient and security consideration, user more wishesOnly just can inquire these information by voice command, instead of remove operating handset or other car in drivingThe equipment of carrying.
General method is the system with static key query word sound bank of installing in terminal device, works as useWhen a key query word is said at family, system can become word speech conversion according to the sound bank of key query word, soAfterwards the keyword of text being sent to network inquires about. Be text keyword data owing to sending to the data of network,Its data volume is less, can make the cost of data network transmission process lower, and efficiency is higher, but this methodShortcoming is that the key query word sound bank that terminal device itself can be stored is limited, because system cannot be expected to useThe query demand at family, therefore the key query word sound bank of terminal device storage itself must be enough large, can coverCover most user's query demand; In addition, if user expects that the information of a renewal is (such as new issueStock), new information keyword does not exist in key query word sound bank, and system is with regard to this keyword of None-identified,Or identification error, user just cannot inquire the information of wanting.
Another kind method is that terminal device directly sends to network agent server, net user's acoustic information dataNetwork proxy server is converted to keyword to be checked, then network acoustic information data (such as wav file)This keyword decorrelation query site of server by utilizing is to relevant information. The benefit of this method is that equipment itself does not needConfigure a very large memory; And because can having a very large keyword, looks into network agent serverAsk sound bank, therefore also very little to user's query demand restriction. But shortcoming is each for terminal deviceWhen inquiry, all need to upload larger voice data (such as wav file), cause the cost of data network transmission processHigher, the transmission time is longer, also needs proxy server to convert text data at every turn, and whole process efficiency is not high,Proxy server software design is more complicated.
Summary of the invention
For the defect existing in above-mentioned prior art, technical problem to be solved by this invention is to provide a kind of useConvenient, little to user's query demand restriction, and also inquiry cost is low, the dynamic voice label letter that search efficiency is highBreath inquiry system and information query method thereof.
In order to solve the problems of the technologies described above, a kind of dynamic voice label information query system provided by the present invention, bagDraw together client device and content server, it is characterized in that: in described client device, be provided with microprocessor, communicationModule, GPS navigation module, speech recognition engine, data memory module and microphone;
Described communication module connects content server through internet, and described GPS navigation module receives GPS with wireless modeNavigation information, described microphone connects the audio input port of microprocessor, and described microprocessor connects and controls each mouldPiece operation;
In described data memory module, deposit fixing voice label list and dynamic voice label list, described fixingIn voice label list, there are download instruction voice label and multiple queries instruction voice label, described dynamic voice markSign in list and have multiple road name voice labels;
Described speech recognition engine is built in microprocessor or deposits in data memory module, for identifying MikeThe voice that wind picks up;
In described content server, deposit the road name list in each city.
The information query method of dynamic voice label information query system provided by the invention, is characterized in that, concreteStep is as follows:
1) client device picks up voice by microphone, and by speech recognition engine by the voice that pick up and dataFixing voice label list and dynamic voice label list in memory module are compared, if compare unsuccessful heavyDuplicate step;
2) generate a download request or inquiry request according to the comparison result of step 1, or upgrade existing inquiry and askAsk;
If in step 1, the download instruction voice label in the voice that pick up and fixing voice label list is compared successfully,Be download instruction by the speech recognition of picking up, and generate a download request that includes this instruction, then pass throughGPS navigation module gathers the current location information of client device, and gathered current location information is added intoIn download request;
If in step 1, the query statement voice label in the voice that pick up and fixing voice label list is compared successfully,Be the query statement corresponding with comparing successful query statement voice label by the speech recognition of picking up, and checkWhether there is inquiry request, please if do not exist inquiry request directly to generate an inquiry that includes this query statementAfter asking, return to step 1, if exist inquiry request first to cancel this inquiry request, and then one of generation includes thisAfter the inquiry request of query statement, return to step 1;
If in step 1, the road name voice label in the voice that pick up and dynamic voice label list is compared successfully,Check whether there is inquiry request, if do not exist inquiry request to return to step 1; If having inquiry request willThe speech recognition of picking up is the road name text corresponding with comparing successful road name voice label, and shouldText is added in inquiry request;
3) client device sends to content server by communication module by download request or inquiry request;
Content server is received after download request, first determines client device according to the current location information in this requestCity, place, then sends to client device by the road name list in this city; Client device is received contentAfter the road name list that server is passed back, first empty the dynamic voice label list in data memory module, then willEach road name in road name list converts road name voice label one by one to, and deposits moving after emptying inIn the list of state voice label;
Content server is received after inquiry request, carries out this request taking the road name text in this request as keywordIn query statement, and Query Result is sent to client device; Client device receives that content server passes backQuery Result after, play Query Result or show Query Result by graphics mode by audio frequency mode.
Dynamic voice label information query system provided by the invention and information query method thereof, have the following beneficial effect that hasReally:
1) client device picks up voice by microphone, and by speech recognition engine voice command recognition, makesUser in car only need to open one's mouth and can carry out information inquiry, very easy to use;
2) client device can be according to the road name voice in the list of user's current location Regeneration dynamics voice labelLabel, when user arrives another one place, is just easy to obtain local road name voice label, profit like thisWith local road name voice label, user just can inquire real-time traffic, the navigation letter of local roadNear point of interest breath or road is little to user's query demand restriction;
3) in local memory module, only deposit the voice label of some fixed instruction and the road name voice mark of localitySign, local road name voice label can need to dynamically update according to user, and its data volume is less, therefore rightThe storage capacity requirement of local memory module is not high;
4) because voice label is all stored in client device, user's voice data can be fast by speech recognitionEngine identification, and be converted into fast the text keyword that will inquire about, send inquiry by network to content serverRequest only has very little network flow data, relatively sends voice data to network, and the inquiry cost of this design is low,Search efficiency is high.
Brief description of the drawings
Fig. 1 is the structured flowchart of the dynamic voice label information query system of the embodiment of the present invention;
Fig. 2 is the information inquiry flow chart of the dynamic voice label information query system of the embodiment of the present invention;
Fig. 3 is the Regeneration dynamics voice label list of the dynamic voice label information query system of the embodiment of the present inventionFlow chart;
Fig. 4 be the dynamic voice label information query system of the embodiment of the present invention Nanjing Road, inquiry Shanghai City in real timeThe flow chart of traffic.
Detailed description of the invention
Below in conjunction with brief description of the drawings, embodiments of the invention are described in further detail, but the present embodiment being not used inRestriction the present invention, every employing analog structure of the present invention and similar variation thereof, all should list protection model of the present invention inEnclose.
As shown in Figure 1, a kind of dynamic voice label information query system that the embodiment of the present invention provides, comprises manyIndividual client device 1 and multiple content server 2, is characterized in that: in each client device 1, be provided with micro-placeReason device 11, communication module 12, GPS navigation module 13, speech recognition engine 14, data memory module 15 and wheatGram wind 16;
Described communication module 12 connects each content server 2 through internet 3, described GPS navigation module 13 is with nothingLine mode receives GPS navigation information, and described microphone 16 connects the audio input port of microprocessor 11, described micro-Processor 11 connects and controls each module and move;
In described data memory module 15, deposit fixing voice label list and dynamic voice label list, described solidDetermine in voice label list, to have download instruction voice label and multiple queries instruction voice label, described dynamic voiceIn list of labels, there are multiple road name voice labels;
Described speech recognition engine 14 is built in microprocessor or deposits in data memory module, for identifying wheatGram voice that wind 16 picks up;
In described content server, deposit the road name list in each city.
In the embodiment of the present invention, described GPS navigation module 13 is that model is the GPS navigation chip of SirFIII, instituteStating speech recognition engine 14 is ScanSoftVoCon3200 software engines, and described microprocessor 11 is that model isThe micro-chip processor of PowerPC5200, described communication module 12 is that model is the Bluetooth chip of CSRBlueCore6,When client device 1 uses by this Bluetooth chip be built-in with Bluetooth transmission module outside that can connecting InternetCommunication apparatus interconnection, and then be connected to internet by external communications equipment;
In the embodiment of the present invention, described communication module 12 can be also gsm module, CDMA module or ADSL modulationDemodulator;
The information inquiry side of the dynamic voice label information query system that as shown in Figure 2, the embodiment of the present invention providesMethod, concrete steps are as follows:
1) client device picks up voice by microphone, and by speech recognition engine by the voice that pick up and dataFixing voice label list and dynamic voice label list in memory module are compared, if compare unsuccessful heavyDuplicate step;
2) generate a download request or inquiry request according to the comparison result of step 1, or upgrade existing inquiry and askAsk;
If in step 1, the download instruction voice label in the voice that pick up and fixing voice label list is compared successfully,Be download instruction by the speech recognition of picking up, and generate a download request that includes this instruction, then pass throughGPS navigation module gathers the current location information (such as longitude and latitude) of client device, and by gathered working asFront position information is added in download request;
If in step 1, the query statement voice label in the voice that pick up and fixing voice label list is compared successfully,Be the query statement corresponding with comparing successful query statement voice label by the speech recognition of picking up, and checkWhether there is inquiry request, please if do not exist inquiry request directly to generate an inquiry that includes this query statementAfter asking, return to step 1, if exist inquiry request first to cancel this inquiry request, and then one of generation includes thisAfter the inquiry request of query statement, return to step 1;
If in step 1, the road name voice label in the voice that pick up and dynamic voice label list is compared successfully,Check whether there is inquiry request, if do not exist inquiry request to return to step 1; If having inquiry request willThe speech recognition of picking up is the road name text corresponding with comparing successful road name voice label, and shouldText is added in inquiry request;
3) client device sends to content server by communication module by download request or inquiry request;
Content server is received after download request, first determines client device according to the current location information in this requestCity, place, then sends to client device by the road name list in this city; Client device is received contentAfter the road name list that server is passed back, first empty the dynamic voice label list in data memory module, then willEach road name in road name list converts road name voice label one by one to, and deposits moving after emptying inIn the list of state voice label;
Content server is received after inquiry request, carries out this request taking the road name text in this request as keywordIn query statement, and Query Result is sent to client device; Client device receives that content server passes backQuery Result after, play Query Result or show Query Result by graphics mode by audio frequency mode.
As shown in Figure 3, when the road name that does not have user to inquire about in the dynamic voice label list of client deviceWhen voice label (such as user is usually in the road list of use Pekinese, Beijing, once and user has arrived Shanghai, useThe condition of road surface in Shanghai is wanted to inquire about in family, must use the road list in Shanghai), now user say voice command " underCarry link name ", client device is compared out in this voice command and fixing voice label list by speech recognition engineDownload instruction voice label match, be identified as download instruction by it, then generate one and include this instructionDownload request, and in download request, add the current location information that GPS navigation module gathers (such as longitude andLatitude), the form of this download request is: download instruction+current location information is (such as city, the current place of user isShanghai, the current location information that GPS navigation module gathers is within the scope of Shanghai, and now the form of download request is:Download instruction+Shanghai); And then by communication module, download request is sent to content server, content serverAccording to the download request of receiving (download instruction+current location information), with the corresponding city of current location informationThe road name list in (such as Shanghai) sends to client device; Client device receive road name list itAfter, convert one by one each road name in road name list to road name voice label, and deposit in and emptyAfter dynamic voice label list in;
As shown in Figure 4, when user wants to inquire about the real-time traffic situation of Nanjing Road in Shanghai, and downloaded ShanghaiRoad name list, now user says voice command " real-time traffic ", client device draws by speech recognitionHold up the real-time traffic query statement voice label of comparing out in this voice command and fixing voice label list and match,Be identified as real-time traffic query statement by it, then generate an inquiry request that includes this instruction, and with languageSound mode or graphics mode prompting user " please say road name ", and client device is compared out by speech recognition engineRoad name voice label " Nanjing Road " in this voice command and dynamic voice label list matches, by itBe identified as content for the text of " Nanjing Road ", and the text is added in inquiry request, now inquiry requestForm is: real-time traffic query statement+Nanjing Road; And then by communication module, inquiry request is sent to content clothesBusiness device, content server is carried out and is carried out real-time traffic query statement taking " Nanjing Road " as keyword, and will inquire about knotFruit sends to client device; Client device is received after the Query Result that content server passes back, by audio frequency sideFormula is play Query Result or is shown Query Result by graphics mode.