CN106920553A - A kind of speech recognition control system and its identification control method - Google Patents
A kind of speech recognition control system and its identification control method Download PDFInfo
- Publication number
- CN106920553A CN106920553A CN201710279717.8A CN201710279717A CN106920553A CN 106920553 A CN106920553 A CN 106920553A CN 201710279717 A CN201710279717 A CN 201710279717A CN 106920553 A CN106920553 A CN 106920553A
- Authority
- CN
- China
- Prior art keywords
- sound
- exynos
- development boards
- speech recognition
- stm32 single
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 24
- 238000011161 development Methods 0.000 claims abstract description 39
- 238000004891 communication Methods 0.000 claims description 12
- 230000008569 process Effects 0.000 claims description 9
- 230000015572 biosynthetic process Effects 0.000 claims description 6
- 238000003786 synthesis reaction Methods 0.000 claims description 6
- 239000000203 mixture Substances 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims 1
- 238000005516 engineering process Methods 0.000 description 11
- 230000000875 corresponding effect Effects 0.000 description 10
- 238000013461 design Methods 0.000 description 7
- 239000000284 extract Substances 0.000 description 4
- 238000001914 filtration Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 238000002054 transplantation Methods 0.000 description 2
- 241000209140 Triticum Species 0.000 description 1
- 235000021307 Triticum Nutrition 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 238000012772 sequence design Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/28—Constructional details of speech recognition systems
- G10L15/30—Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
- G10L2015/223—Execution procedure of a spoken command
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Circuit For Audible Band Transducer (AREA)
Abstract
A kind of speech recognition control system and its identification control method, the system include sound pick-up, the development boards of Exynos 4412, news rumours sound high in the clouds and STM32 single-chip microcomputers.Method is comprised the following steps：The external sound of microphone pickup simultaneously changes into data signal and is passed to the development boards of Exynos 4412；The result that Exynos4412 development boards send information to interrogate on rumours sound high in the clouds after treatment is returned on the development boards of Exynos 4412 with json text formattings；Then matching result is analyzed and extracted and changes into keyword by the development boards of Exynos 4412, and information then is passed into STM32 single-chip microcomputers with GB2312 yards of form；Last STM32 single-chip microcomputers carry out task and match and perform the corresponding task after matching.The system and method not only greatly reducing the probability of misrecognition, and strong adaptability, can extensively be embedded in the small machines the inside in life.
Description
Technical field
The invention belongs to Embedded speech recognition controlled technical field, a kind of speech recognition control system is specifically related to
And its identification control method.
Background technology
Speech recognition controlled technology is widely used, especially in robot control field, current speech recognition controlled
Method is usually the api function that Recording Process is carried using Windows first, and being obtained by recorded program needs the audio of identification
File；Then audio file is input in news rumours sound recognition function interface and is processed, export the text corresponding with audio
This；Controlled according to the keyword in text is recognized with Zigbee again.There is problems with the technology：1st, the technology is to be based on
In Windows operating system, Windows operating system is very big to hardware requirement, not as embedded system, it is impossible to be embedded into
Small machines such as router etc., therefore bad adaptability, it is impossible to widely use；2nd, this technology exists to identification text key word
Defect, the likelihood ratio of misrecognition is larger.
The content of the invention
The present invention is in view of the shortcomings of the prior art, there is provided a kind of speech recognition control system and its identification control method；Should
System and method not only greatly reducing the probability of misrecognition, and strong adaptability, can be embedded in extensively small-sized in living
Machine the inside.
In order to achieve the above object, a kind of speech recognition control system of the invention, it is main to include for picking up external sound
And the analog signal of the sound is synthesized the sound pick-up for being processed into data signal, come from sound pick-up sound figure as host process
The development boards of Exynos 4412 of signal, for audio digital signal being analyzed into matching and result being converted into json texts
The news rumours sound high in the clouds of form, and processed at the information from the development boards of Exynos 4412 and execution as control process device
Manage the STM32 single-chip microcomputers of corresponding task after the information；The sound pick-up, the development boards of Exynos 4412 and STM32 single-chip microcomputers according to
Secondary communication connection, and interrogate rumours sound high in the clouds then based on the development boards of Exynos 4412 as host platform respectively with sound pick-up,
STM32 single chip communications are connected；The operating system is Linux QT operating systems, and then conduct should for the development boards of Exynos 4412
The platform of operating system.
Preferably, be delivered to the analog signal of sound in the DSP of microphone array module XFM10412 by the sound pick-up
Synthesis is processed into data signal.
Preferably, the sound pick-up is by four linear microphones, base plate filter circuit, interface circuit, and microphone
Array module is constituted.The microphone array module is the winged microphone array module XFM10412 of University of Science and Technology's news.
Preferably, the development boards of the Exynos 4412 are by the matching knot of the json text formattings from news rumours sound high in the clouds
Fruit is analyzed and extracts and changes into keyword, is then passed to information serial communication mode with GB2312 yards of form
STM32 single-chip microcomputers, afterwards STM32 single-chip microcomputers corresponding will be appointed with what GB2312 yards of form carry out that task matches and perform after matching
Business, so as to complete the function of speech recognition controlled.
A kind of identification control method of speech recognition control system, mainly includes the following steps that：
First, the external sound of microphone pickup and the synthesis of the analog signal of the sound is processed into after data signal by string
Port communications are passed to the development boards of Exynos 4412；, used as slave, Exynos4412 development boards are then as main for sound pick-up now
Machine；
Secondly, Exynos4412 development boards are based on the speech recognition application programming interface of the winged exploitation of news in Linux QT operating systems
Upper execution, the result that the information transmitted from sound pick-up is sent on news rumours sound high in the clouds after treatment is returned with json text formattings
Return in operating system；
The development board of 3rd, Exynos 4412 carries out the matching result of the json text formattings from news rumours sound high in the clouds
Analyze and extract and change into keyword, it is mono- that information serial communication mode then is passed into STM32 with GB2312 yards of form
Piece machine；, used as slave, Exynos4412 development boards are then as main frame for STM32 single-chip microcomputers now；
Finally, STM32 single-chip microcomputers will carry out task matching with GB2312 yards of form, perform again afterwards corresponding after matching
Task, so as to complete the function of speech recognition controlled.
Preferably, the sound pick-up is by four linear microphones, base plate filter circuit, interface circuit, and microphone
Array module is constituted.The microphone array module is the winged microphone array module XFM10412 of University of Science and Technology's news.
Compared with prior art, the present invention mainly has the advantage that：1st, using Linux QT as operating system platform,
Linux QT operating systems are low to hardware requirement, so adaptability is good, have a wide range of application；2nd, with reference to the technology that news rumours sound is ripe
In system, multiple times of filtration information can substantially reduce false recognition rate in itself；3rd, GB2312 yards is changed into json text keywords,
And control is completed with GB2312 yards of matching corresponding task in STM32 single-chip microcomputers, can more accurately perform task with the method.
Brief description of the drawings
Fig. 1 is the composition frame chart of speech recognition control system of the present invention；
Fig. 2 is the algorithm flow chart of speech recognition control system of the present invention.
Wherein, 1 is sound pick-up, and 2 is the development boards of Exynos 4412, and 3 is news rumours sound high in the clouds, and 4 is STM32 single-chip microcomputers.
Specific embodiment
Below in conjunction with the drawings and specific embodiments, the present invention will be described in detail, but not as a limitation of the invention.
Referring to Figures 1 and 2, a kind of speech recognition control system of the embodiment of the present invention, it is main to include for picking up external sound
The analog signal of the sound is simultaneously synthesized the sound pick-up 1 for being processed into data signal for sound, and the sound of sound pick-up 1 is come from as host process
The development boards 2 of Exynos 4412 of data signal, for audio digital signal being analyzed into matching and result being converted into json
Text formatting is returning to the news rumours sound high in the clouds 3 in operating system, and is processed as control process device and come from Exynos
The information of 4412 development boards 2 and perform process the information after corresponding task STM32 single-chip microcomputers 4；The sound pick-up 1, Exynos
4412 development boards 2 and STM32 single-chip microcomputers 4 are communicated to connect successively, and are interrogated rumours sound high in the clouds 3 and then developed based on Exynos 4412
Plate 2 is communicated to connect with sound pick-up 1, STM32 single-chip microcomputers 3 respectively as host platform；The operating system is operated for Linux QT
System, the development boards 2 of Exynos 4412 are then as the platform of the operating system.
Synthesis is processed into data signal during the analog signal of sound is delivered to the DSP of module by the sound pick-up 1, described to pick up
Sound device 1 is by four linear microphones, base plate filter circuit, interface circuit, and the winged microphone array module of University of Science and Technology's news
XFM10412 is constituted.The development boards 2 of the Exynos 4412 are by the matching knot of the json text formattings from news rumours sound high in the clouds 3
Fruit is analyzed and extracts and changes into keyword, is then passed to information serial communication mode with GB2312 yards of form
STM32 single-chip microcomputers 4, afterwards STM32 single-chip microcomputers 4 will with GB2312 yards of form carry out task match and perform matching after it is corresponding
Task, so as to complete the function of speech recognition controlled.
Referring to Figures 1 and 2, a kind of identification control method of speech recognition control system, mainly includes the following steps that：
First, sound pick-up 1 picks up external sound and passes through after the analog signal synthesis of the sound is processed into data signal
Serial communication is passed to the development boards 2 of Exynos 4412；Sound pick-up 1 now is then made as slave, Exynos4412 development boards 2
It is main frame；
The sound pick-up 1 is by four linear microphones, base plate filter circuit, interface circuit, and the winged wheat of University of Science and Technology's news
Gram wind array module XFM10412 composition.Be delivered to for the analog signal of sound to synthesize in the DSP of module and be processed into by the sound pick-up 1
Data signal.
Secondly, Exynos4412 development boards 2 are based on interrogating the speech recognition application programming interface for flying exploitation and being operated in Linux QT and are
Performed on system, the information transmitted from sound pick-up 1 is sent to result on news rumours sound high in the clouds 3 after treatment with json text lattice
Formula is returned in operating system；
The development board 2 of 3rd, Exynos 4412 enters the matching result of the json text formattings from news rumours sound high in the clouds 3
Row is analyzed and extracted and changes into keyword, and information serial communication mode then is passed into STM32 with GB2312 yards of form
Single-chip microcomputer 4；, used as slave, Exynos4412 development boards 2 are then as main frame for STM32 single-chip microcomputers 4 now；
Finally, STM32 single-chip microcomputers 4 will carry out task matching with GB2312 yards of form, perform the phase after matching again afterwards
Task is answered, so as to complete the function of speech recognition controlled.
Software for Design in the present invention is divided into design, the control journey of STM32 single-chip microcomputers that news fly speech recognition application programming interface
Sequence design, Linux QT operating system transplantations and driving transplanting development design.Its CTC's rumours sound recognition application designs and is
Fly the interface function of itself using news to develop, modification addition code is adapted to the requirement of oneself, and this design is most original
Speech recognition full text result extracts keyword with JSON forms, the misrecognition result that raw tone identification may be brought
Multiple times of filtration is carried out, the probability of misrecognition is substantially reduced, along with the accuracy high identification that news scud end calculates, general speech is known
Other accuracy can accomplish 99%.The design of STM32 single-chip microcomputer control programs is to write serial communication function code and control electricity
The code of machine.Wherein, this is to control corresponding motor to make corresponding actions again after carrying out task matching by GB2312 yards；Linux
QT operating system transplantations and driving transplanting development design are to transplant Linux QT operating systems in Exynos4412 development boards, due to
The system transplanted lacks WIFI and drives with driver of sound card, it is necessary to be driven exploitation according to WIFI chips, driver of sound card is carried out
Storehouse transplanting etc..
Compared with prior art, the present invention mainly has the advantage that：1st, using Linux QT as operating system platform,
Linux QT operating systems are low to hardware requirement, so adaptability is good, have a wide range of application；2nd, with reference to the technology that news rumours sound is ripe
Multiple times of filtration information can substantially reduce false recognition rate to system in itself again；3rd, GB2312 yards is changed into json text keywords,
And control is completed with GB2312 yards of matching corresponding task in STM32 single-chip microcomputers, can more accurately perform task with the method.
Below the present invention is described in detail, but it will be apparent that those skilled in the art can carry out various changing
Become and improve, without departing from the scope of the present invention that appended claims are limited.
Claims (5)
1. a kind of speech recognition control system, it is characterised in that：It is main to include for picking up external sound and by the mould of the sound
Intend the sound pick-up that signal synthesis is processed into data signal, as Exynos of the host process from sound pick-up audio digital signal
4412 development boards, for audio digital signal being analyzed into matching and result being converted into the news rumours sound of json text formattings
High in the clouds, and process the information from the development boards of Exynos 4412 as control process device and perform corresponding after the treatment information
The STM32 single-chip microcomputers of task；The sound pick-up, the development boards of Exynos 4412 and STM32 single-chip microcomputers are communicated to connect successively, and
News rumours sound high in the clouds is then connected with sound pick-up, STM32 single chip communications respectively based on the development boards of Exynos 4412 as host platform
Connect；The operating system is Linux QT operating systems, and the development boards of Exynos 4412 are then as the platform of the operating system.
2. a kind of speech recognition control system according to claim 1, it is characterised in that：The sound pick-up is by four lines
Property microphone, base plate filter circuit, interface circuit, and microphone array module composition.
3. a kind of speech recognition control system according to claim 1, it is characterised in that：The Exynos 4412 is developed
The matching result of the json text formattings from news rumours sound high in the clouds is analyzed and extracted and changes into keyword by plate, then with
Information serial communication mode is passed to STM32 single-chip microcomputers by GB2312 yards of form, and STM32 single-chip microcomputers will be with GB2312 afterwards
The form of code carries out task and matches and perform the corresponding task after matching, so as to complete the function of speech recognition controlled.
4. the identification control method of a kind of speech recognition control system as claimed in claim 1, it is characterised in that mainly include
Following steps：
First, the external sound of microphone pickup and led to by serial ports after the synthesis of the analog signal of the sound is processed into data signal
Letter is passed to the development boards of Exynos 4412；
Secondly, Exynos4412 development boards are held based on the speech recognition application programming interface that news fly exploitation in Linux QT operating systems
OK, the result that the information for being transmitted from sound pick-up is sent on news rumours sound high in the clouds after treatment is returned to json text formattings
On the development boards of Exynos 4412 of embedded Linux QT operating systems；
Be analyzed for the matching result of the json text formattings from news rumours sound high in the clouds by the development board of the 3rd, Exynos 4412
And extraction changes into keyword, and information serial communication mode then is passed into STM32 single-chip microcomputers with GB2312 yards of form；
Finally, STM32 single-chip microcomputers will carry out task matching with GB2312 yards of form, perform corresponding after matching again afterwards
Business, so as to complete the function of speech recognition controlled.
5. a kind of voice identification control method according to claim 4, it is characterised in that：The sound pick-up is by four lines
Property microphone, base plate filter circuit, interface circuit, and microphone array module composition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710279717.8A CN106920553A (en) | 2017-04-21 | 2017-04-21 | A kind of speech recognition control system and its identification control method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710279717.8A CN106920553A (en) | 2017-04-21 | 2017-04-21 | A kind of speech recognition control system and its identification control method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106920553A true CN106920553A (en) | 2017-07-04 |
Family
ID=59567567
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710279717.8A Pending CN106920553A (en) | 2017-04-21 | 2017-04-21 | A kind of speech recognition control system and its identification control method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106920553A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109637539A (en) * | 2019-01-29 | 2019-04-16 | 浪潮金融信息技术有限公司 | A kind of audio recognition method of the What You See Is What You Get based on the Iflytek unlimited time |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102708863A (en) * | 2011-03-28 | 2012-10-03 | 德信互动科技（北京）有限公司 | Voice dialogue equipment, system and voice dialogue implementation method |
| CN103336788A (en) * | 2013-06-05 | 2013-10-02 | 上海交通大学 | Humanoid robot added Internet information acquisition method and system |
| CN105446146A (en) * | 2015-11-19 | 2016-03-30 | 深圳创想未来机器人有限公司 | Intelligent terminal control method based on semantic analysis, system and intelligent terminal |
| CN105491080A (en) * | 2014-09-16 | 2016-04-13 | 比亚迪股份有限公司 | Vehicle control method and system based on mobile terminal |
| CN205487330U (en) * | 2015-12-28 | 2016-08-17 | 天津天大天星智能物联技术有限公司 | Controller based on pronunciation array |
| CN105895100A (en) * | 2016-06-29 | 2016-08-24 | 广东美的厨房电器制造有限公司 | Kitchen voice control device, system and method |
-
2017
- 2017-04-21 CN CN201710279717.8A patent/CN106920553A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102708863A (en) * | 2011-03-28 | 2012-10-03 | 德信互动科技（北京）有限公司 | Voice dialogue equipment, system and voice dialogue implementation method |
| CN103336788A (en) * | 2013-06-05 | 2013-10-02 | 上海交通大学 | Humanoid robot added Internet information acquisition method and system |
| CN105491080A (en) * | 2014-09-16 | 2016-04-13 | 比亚迪股份有限公司 | Vehicle control method and system based on mobile terminal |
| CN105446146A (en) * | 2015-11-19 | 2016-03-30 | 深圳创想未来机器人有限公司 | Intelligent terminal control method based on semantic analysis, system and intelligent terminal |
| CN205487330U (en) * | 2015-12-28 | 2016-08-17 | 天津天大天星智能物联技术有限公司 | Controller based on pronunciation array |
| CN105895100A (en) * | 2016-06-29 | 2016-08-24 | 广东美的厨房电器制造有限公司 | Kitchen voice control device, system and method |
Non-Patent Citations (1)
| Title |
|---|
| 严毓培等: "智能家居服务型机器人的设计与开发", 《电子世界》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109637539A (en) * | 2019-01-29 | 2019-04-16 | 浪潮金融信息技术有限公司 | A kind of audio recognition method of the What You See Is What You Get based on the Iflytek unlimited time |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101923857A (en) | Extensible audio recognition method based on man-machine interaction | |
| CN102737629A (en) | Embedded type speech emotion recognition method and device | |
| CN103198829A (en) | Method, device and equipment of reducing interior noise and improving voice recognition rate | |
| CN109935226A (en) | A kind of far field speech recognition enhancing system and method based on deep neural network | |
| CN109389976A (en) | Intelligent household electrical appliance control method and device, intelligent household electrical appliance and storage medium | |
| CN105848062B (en) | The digital microphone of multichannel | |
| CN110349582A (en) | Display device and far field speech processing circuit | |
| CN111145746A (en) | Man-machine interaction method based on artificial intelligence voice | |
| CN108447483A (en) | Speech recognition system | |
| CN102890931A (en) | Method for increasing voice recognition rate | |
| CN102671383A (en) | Game implementing device and method based on acoustic control | |
| CN107818778A (en) | A kind of interactive system based on intelligent sound mouse | |
| CN106920553A (en) | A kind of speech recognition control system and its identification control method | |
| CN103903617A (en) | Voice recognition method and electronic device | |
| CN208724111U (en) | Far field speech control system based on television equipment | |
| CN109243458A (en) | A kind of speech recognition system for intelligent robot | |
| CN206021901U (en) | A kind of speech recognition of wisdom bank outlets tangible machine people and speech synthetic device | |
| CN111968411A (en) | Unmanned aerial vehicle swarm scheduling system and method based on voice recognition | |
| CN110265014A (en) | A kind of method, apparatus and translator of voice control | |
| CN202600936U (en) | Remote controller with voice function | |
| CN110148407A (en) | Sound control method for Intelligent bracelet | |
| CN206209693U (en) | A kind of multi-channel audio signal parallel acquisition device | |
| CN204760038U (en) | Recording pen with recording and text writing function | |
| CN205647914U (en) | Stereo set is far controlled to intelligence | |
| CN102698434A (en) | Device and method for implementing game based on conversation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170704 |
|
| WD01 | Invention patent application deemed withdrawn after publication |