CN108449479A - A kind of medical wireless communication means and system based on speech recognition - Google Patents
A kind of medical wireless communication means and system based on speech recognition Download PDFInfo
- Publication number
- CN108449479A CN108449479A CN201810404846.XA CN201810404846A CN108449479A CN 108449479 A CN108449479 A CN 108449479A CN 201810404846 A CN201810404846 A CN 201810404846A CN 108449479 A CN108449479 A CN 108449479A
- Authority
- CN
- China
- Prior art keywords
- medical care
- voice
- request information
- equipment
- identification module
- 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
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M1/00—Substation equipment, e.g. for use by subscribers
- H04M1/26—Devices for calling a subscriber
- H04M1/27—Devices whereby a plurality of signals may be stored simultaneously
- H04M1/271—Devices whereby a plurality of signals may be stored simultaneously controlled by voice recognition
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; 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 OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Processing of the speech or voice signal to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
- G10L21/0208—Noise filtering
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Processing of the speech or voice signal to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
- G10L21/0208—Noise filtering
- G10L21/0216—Noise filtering characterised by the method used for estimating noise
- G10L21/0232—Processing in the frequency domain
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
- G10L25/51—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; 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
Abstract
A kind of medical wireless communication means and system based on speech recognition, the Call- Control1 and call function of Nurse Mobile extension facility are realized by medical care sound identification module, the both hands that doctor and nurse can be liberated to greatest extent, keep the work of doctor and nurse more convenient;Nurse Mobile equipment is identified phonetic order by built-in medical care sound identification module, and recognition result is matched to the database in Nurse Mobile equipment, and corresponding instruction is executed according to match condition;The Call- Control1 and call function of Nurse Mobile extension facility are realized by medical care sound identification module, the both hands of medical staff can be liberated to greatest extent, keep the work of medical staff more convenient, more portable Nurse Mobile equipment is provided for medical staff, substantially increases the working efficiency of medical staff.
Description
Technical field:
The present invention relates to a kind of medical wireless communication means and system based on speech recognition.
Background technology:
Currently, in traditional medical mobile devices terminal, if it is desired to call some doctor or nurse, need with touch screen or
It is the mode of button selection, first finds call object, call operation then is carried out to it again.When medical mobile devices terminal
On software interface it is excessive or when interface is more complex, user's operation step grows proportionately, and has seriously affected user experience.
Invention content:
In order to compensate for the shortcomings of the prior art, the present invention provides a kind of medical wireless communication means based on speech recognition
And system, the Call- Control1 and call function of Nurse Mobile extension facility are realized by medical care sound identification module, it can be maximum
The both hands of the liberation doctor and nurse of limit, keep the work of doctor and nurse more convenient;Solving in the prior art can not root
The problem of corresponding Nurse Mobile equipment control function being provided according to user speech content.
Used technical solution is the present invention to solve above-mentioned technical problem:
In a first aspect, an embodiment of the present invention provides the medical wireless communication means based on speech recognition, including:First moves
The voice call request information that dynamic medical care equipment is sent out based on the first user, activates first in the first movement medical care equipment
Medical care sound identification module;The first medical care sound identification module for obtaining and identifying voice call request information, and with
Speech form exports recognition result;
The first medical care sound identification module that the first movement medical care is set is according to the voice sent out to first user
The recognition result of call request information determines the letter of the corresponding called device owning user of institute's voice call request information
Breath carries out the medical care Affiliates List of the called device owning user information and first movement medical care equipment storage
Matching;
If successful match, the contact person corresponding second of first movement medical care equipment to the successful match moves
Dynamic medical care equipment, initiates calling;
If matching is unsuccessful, the first movement medical care equipment sends out prompt;
The second Nurse Mobile equipment is listened request information based on the voice that second user is sent out, and activation described second moves
The second medical care sound identification module in dynamic medical care equipment;
Second medical care sound identification module of the second Nurse Mobile equipment identifies the voice that the second user is sent out
Listen request information, and exports recognition result with speech form;
Listened request information based on the voice that the second Nurse Mobile equipment sends out the second user, establish with
Voice connection between the first movement medical care equipment.
Preferably, the sound identification module of the first movement medical care equipment or the language of the second Nurse Mobile equipment
Sound identification module identifies that voice call request information or voice listen request information by same speech recognition algorithm respectively.Institute
Voice call request information is the request made a phone call to so-and-so doctor or so-and-so nurse that doctor or nurse send out;The voice
The information that listens request is medical care or the solicited message answered or hung up that nurse sends out.
Preferably, the speech recognition algorithm is:
Voice call request information or voice is obtained to listen request information;Such as:So-and-so doctor needs to call so-and-so shield
Then scholar sends out the voice call requests information such as " making a phone call to so-and-so nurse ", " call so-and-so nurse ";Or Nurse Mobile is set
Standby when above having incoming call, so-and-so doctor or so-and-so nurse send out the voices such as " answering " or " sheet " and listen request information;
Institute's voice call request information or voice are listened request information, corresponding first text message is converted to;
The first keyword set is extracted from the text message according to preset rules, wherein first keyword set
It include at least one keyword in conjunction;
Calculate the similarity of first keyword set and the second keyword set, wherein second keyword set
It is combined into the keyword set that corresponding second text message extracts in presupposed information library, wherein the presupposed information library Zhong Bao
Containing multiple second text messages and the incidence relation for receiving group information.
Preferably, by following formula calculate separately the first text message A and the second text message B morphology similarity,
The long similarity of sentence, word order similarity, Distance conformability degree:
Morphology similarity:
The long similarity of sentence:
Word order similarity:
Distance conformability degree:
The similarity of the first text message A and the second text message B are calculated according to following formula:
Sim (A, B)=α1WordSim(A,B)+α2LenSim(A,B)+α3OrderSim(A,B)+α4DisSim(A,B)
Wherein, Len (A) is the number of keyword in the corresponding keyword sets of the first text message A, and Len (B) is second
The number of keyword in the corresponding keyword sets of text message B;SameWc (A, B) is the corresponding keys of the first text message A
The number of same keyword in set of words and the corresponding keyword sets of the second text message B;OnceWs (A, B) is indicated first
All occur in text message A and the second text message B and primary keyword set only occurs;| OnceWs (A, B) is OnceWs
The number of keyword in (A, B) set, the word distance value that k be the trained DisSim (A, B) arrived when being 0.5, Max () are
It is maximized function, Abs () is the function that takes absolute value, α1For the weight coefficient of morphology similarity WordSim (A, B), α2For sentence
The weight coefficient of long similarity LenSim (A, B), α3For the similarity of word order similarity OrderSim (A, B), α4For apart from similar
Spend the weight coefficient of DisSim (A, B).
Preferably, after the voice call request that the first movement medical care equipment is sent out based on the first user, and
Before activating the first medical care sound identification module in the first movement medical care equipment;
And the voice sent out based on second user in the second Nurse Mobile equipment is listened request after information, and
Before activating the second medical care sound identification module in the second Nurse Mobile equipment;
The method includes:
It obtains institute's voice call request or the voice to listen request corresponding first voice signal, wherein described
First voice signal is the voice signal that input carries noise in real time;
First voice signal is transformed into frequency domain, obtains corresponding second voice signal of frequency domain;
Logarithmic spectrum posteriori SNR is calculated according to second voice signal, the logarithmic spectrum posteriori SNR is present frame
Ratio between the logarithm and the logarithm of former frame noise power estimation value of the power spectrum of second voice signal;
Noise power spectrum estimated value is obtained according to the logarithmic spectrum posteriori SNR based on weighted noise estimation algorithm;
The gain function that Wiener filtering is generated according to the noise power spectrum estimated value, according to the gain function to described
Second voice signal is filtered, and generates third voice signal, wherein the third voice signal is removal noise in frequency domain
Voice signal;
The third voice signal is transformed into time domain, obtains the 4th voice signal, wherein the 4th voice signal is made
To remove the voice signal of noise in time domain.
Preferably, the logarithmic spectrum check back signal ratio is calculated by following formula:
Wherein, POST_SNR (k, w) is the logarithmic spectrum posteriori SNR on second w-th of frequency point of voice signal kth frame, | y
(k,w)|2It is noise power on kth w-th of frequency point of -1 frame for the power spectrum of the second voice signal on w-th of frequency point of kth frame
Estimated value, a are constant.
Second aspect, the medical wireless communication system based on speech recognition that an embodiment of the present invention provides a kind of, including the
One Nurse Mobile equipment, the second Nurse Mobile equipment;
The first movement medical care equipment is used for the voice call request information that is sent out based on the first user, activation described the
The first medical care sound identification module in one mobile device;
The first medical care sound identification module is used to obtain and identify voice call request information, and defeated with speech form
Go out recognition result;
And the first medical care sound identification module of the first movement medical care medical care equipment, it is additionally operable to according to described first
The recognition result for the voice call request information that user sends out determines the corresponding called device of institute's voice call request information
The information of owning user joins the medical care of the called device owning user information and first movement medical care equipment storage
It is that roster matches;
The first movement medical care equipment is additionally operable to correspond to the contact person of the successful match in successful match
The second Nurse Mobile equipment, initiate calling;And unsuccessful prompt is being sent out matching;
The second Nurse Mobile equipment is used to listen request information 407 based on the voice that second user is sent out, and activates institute
State the second medical care sound identification module in the second Nurse Mobile equipment;
The second user is sent out for identification for second medical care sound identification module of the second Nurse Mobile equipment
Voice listens request information, and exports recognition result with speech form;
Second Nurse Mobile equipment is additionally operable to listen request information 407 based on the voice for sending out the second user, builds
The vertical voice between the first movement medical care equipment is connect.
Preferably, a kind of medical wireless communication system based on speech recognition, including:The system comprises first movement doctors
Protect equipment, the second Nurse Mobile equipment;Wherein, the first movement medical care equipment includes the first medical care sound identification module, institute
It includes the second medical care speech ciphering equipment module to state the second Nurse Mobile equipment;
The first movement medical care equipment is used for the voice call request information that is sent out based on the first user, activation described the
One medical care sound identification module;The first medical care sound identification module for obtaining and identify voice call request information, and
Recognition result is exported with speech form, and for the identification according to the voice call request information sent out to first user
As a result, the information of the corresponding called device owning user of institute's voice call request information is determined, by the called device
Owning user information is matched with the medical care Affiliates List that the first movement medical care equipment stores;
The first movement medical care equipment is additionally operable in the successful match, is corresponded to the contact person of the successful match
The second Nurse Mobile equipment, initiate calling;And for when matching unsuccessful, sending out prompt;
The second Nurse Mobile equipment is used to listen request information based on the voice that second user is sent out, activation described the
Two medical care sound identification modules;The second medical care sound identification module voice that the second user is sent out for identification is answered
Solicited message, and recognition result is exported with speech form;
The second Nurse Mobile equipment is additionally operable to, and is listened request information, is built based on the voice that the second user is sent out
The vertical voice between the first movement medical care equipment is connect.
Preferably, a kind of medical wireless communication system based on speech recognition, first sound identification module or institute
The second sound identification module is stated, respectively by same speech recognition algorithm, identifies that corresponding voice call request information or voice connect
Listen solicited message.
Preferably, a kind of medical wireless communication system based on speech recognition, the speech recognition algorithm are:
Voice call request information or voice is obtained to listen request information;
Institute's voice call request information or voice are listened request information, corresponding first text message is converted to;
The first keyword set is extracted from the text message according to preset rules, wherein first keyword set
It include at least one keyword in conjunction;
Calculate the similarity of first keyword set and the second keyword set, wherein second keyword set
It is combined into the keyword set that corresponding second text message extracts in presupposed information library, wherein the presupposed information library Zhong Bao
Containing multiple second text messages and the incidence relation for receiving group information.
The present invention using the above scheme, Nurse Mobile equipment by built-in medical care sound identification module to phonetic order into
Row identification, and recognition result is matched to the database in Nurse Mobile equipment, execute corresponding instruction according to match condition;I.e.
The Call- Control1 and call function that Nurse Mobile extension facility is realized by medical care sound identification module, can solve to greatest extent
The both hands for putting medical staff keep the work of medical staff more convenient, and more portable Nurse Mobile equipment is provided for medical staff,
Substantially increase the working efficiency of medical staff.
Description of the drawings:
Fig. 1 is the flow of a kind of medical wireless communication means and system based on speech recognition provided in an embodiment of the present invention
Figure;
Fig. 2 is the stream of another medical wireless communication means and system based on speech recognition provided in an embodiment of the present invention
Cheng Tu;
Fig. 3 is the stream of another medical wireless communication means and system based on speech recognition provided in an embodiment of the present invention
Cheng Tu;
Fig. 4 is a kind of medical wireless communication system schematic diagram based on speech recognition provided by the invention.
Specific implementation mode:
In order to clarify the technical characteristics of the invention, below by specific implementation mode, and its attached drawing is combined, to this hair
It is bright to be described in detail.
Fig. 1 is a kind of structure of medical wireless communication means and system based on speech recognition provided in an embodiment of the present invention
Schematic diagram.
S101, the voice call request information that first movement medical care equipment is sent out based on the first user, activation described first
The first medical care sound identification module in Nurse Mobile equipment;The first medical care sound identification module is for obtaining and identifying language
Sound call request information, and recognition result is exported with speech form.
The first medical care sound identification module of S102, the first movement medical care equipment are sent out according to first user
Voice call request information recognition result, determine the corresponding called device owning user of institute's voice call request information
Information, by the called device owning user information and the first movement medical care equipment storage medical care Affiliates List
It is matched;
S103, if successful match, the contact person of first movement medical care equipment to the successful match is corresponding
Second Nurse Mobile equipment initiates calling;
S104, if matching is unsuccessful, the first movement medical care equipment sends out prompt;
S105, the second Nurse Mobile equipment are listened request information based on the voice that second user is sent out, described in activation
The second medical care sound identification module in second Nurse Mobile equipment;
The second medical care sound identification module of S106, the second Nurse Mobile equipment identify what the second user was sent out
Voice listens request information, and exports recognition result with speech form;
S107 is listened request information based on the voice that the second Nurse Mobile equipment sends out the second user,
The voice established between the first movement medical care equipment connects.
Below by taking doctor's application scenarios as an example, the medical nothing provided in an embodiment of the present invention based on speech recognition is elaborated
Line communication means, as shown in Figure 2.
Fig. 2 is the stream of another medical wireless communication means and system based on speech recognition provided in an embodiment of the present invention
Cheng Tu;
S201, the voice call request information that first movement medical care equipment is sent out based on the first user, activation described first
The first medical care sound identification module in Nurse Mobile equipment;The first medical care sound identification module is for obtaining and identifying language
Sound call request information, and recognition result is exported with speech form;When doctor needs to call some doctor or some nurse,
Doctor sends out voice request " call so-and-so doctor " to first movement medical care equipment, activates first in first movement medical care equipment
Medical care sound identification module, the first medical care sound identification module identify content corresponding with the first medical care sound identification module simultaneously
Recognition result is exported with speech form, that is, it is keyword to pre-set " calling ", is detected after " calling " that " calling " is subsequent
Voice is converted into word.
The first medical care sound identification module of S202, the first movement medical care equipment are sent out according to first user
Voice call request information recognition result, determine the corresponding called device owning user of institute's voice call request information
Information, by the called device owning user information and the first movement medical care equipment storage medical care Affiliates List
It is matched;The recognition result of speech form is matched to the voice sent out with the first user and asked by the first medical care sound identification module
The medical care contact person of corresponding first movement medical care equipment storage is asked, i.e., is returned by conversion feedback " so-and-so doctor " and advance
The contact name field of acquisition carries out lookup comparison, and matched this information seeks to the contact person of calling.
S203, if successful match, the contact person of first movement medical care equipment to the successful match is corresponding
Second Nurse Mobile equipment initiates calling.
S204, if matching is unsuccessful, the first movement medical care equipment sends out prompt;" do not find so-and-so doctor ".
Fig. 3 is the stream of another medical wireless communication means and system based on speech recognition provided in an embodiment of the present invention
Cheng Tu.
S301, the second Nurse Mobile equipment are listened request information based on the voice that second user is sent out, described in activation
The second medical care sound identification module in second Nurse Mobile equipment;When having incoming call in the second Nurse Mobile equipment, second uses
Family sends out phonetic order and activates the second medical care sound identification module in the second Nurse Mobile equipment, i.e. microphone that will acquire
Sound passes to news and flies speech recognition library.
The second medical care sound identification module of S302, the second Nurse Mobile equipment identify what the second user was sent out
Voice listens request information, and exports recognition result with speech form;Second medical care sound identification module identifies and the second medical care
The corresponding content of sound identification module simultaneously exports recognition result with speech form, and recognition result is matched to database, matching at
Work(then executes instruction, and matching is unsuccessful, sends out prompt, that is, interrogates and fly speech recognition library according to the progress of pre-set keyword
Match, be matched to after different keywords to interrogate and fly speech recognition library different instructions is passed into application software, application software according to
Instructions match is answered or is hung up, and is directly executed and is answered or hang up processing.
S303 is listened request information based on the voice that the second Nurse Mobile equipment sends out the second user
407, the voice established between the first movement medical care equipment connects.
Fig. 4 is a kind of medical wireless communication system schematic diagram based on speech recognition provided by the invention.
This system is made of first movement medical care equipment 401 and the second Nurse Mobile equipment 402, the first movement doctor
The voice call request information 404 that shield equipment 401 is used to send out based on the first user 403, activates the first movement equipment 401
In the first medical care sound identification module 405;
The first medical care sound identification module 405 is used to obtain and identify voice call request information 404, and with voice
Form exports recognition result;
And the first medical care sound identification module 405 of the first movement medical care equipment 401, it is additionally operable to according to described the
The recognition result for the voice call request information 404 that one user 403 sends out determines that institute's voice call request information 404 corresponds to
Called device owning user information, by the called device owning user information and the first movement medical care equipment
The medical care Affiliates List of 401 storages is matched;
The first movement medical care equipment 401 is additionally operable to the contact person couple to the successful match in successful match
The the second Nurse Mobile equipment 402 answered initiates calling;And unsuccessful prompt is being sent out matching;
The second Nurse Mobile equipment 402 is used to listen request information 407 based on the voice that second user 406 is sent out,
Activate the second medical care sound identification module 408 in the second Nurse Mobile equipment 402;
Second medical care sound identification module 408 of the second Nurse Mobile equipment 402 second user for identification
406 voices sent out listen request information 407, and export recognition result with speech form;
Second Nurse Mobile equipment 402 is additionally operable to listen request information based on the voice for sending out the second user 406
407, the voice established between the first movement medical care equipment 401 connects.
Preferably, a kind of medical wireless communication system based on speech recognition, including:The system comprises first movement doctors
Protect equipment 401, the second Nurse Mobile equipment 402;Wherein, the first movement medical care equipment 401 is known including the first medical care voice
Other module, the second Nurse Mobile equipment include the second medical care speech ciphering equipment module;
The voice call request information 404 that the first movement medical care equipment 401 is used to send out based on the first user 403,
Activate the first medical care sound identification module 405;The first medical care sound identification module is for obtaining and identifying that voice is exhaled
It is solicited message 404, and recognition result is exported with speech form, and for according to the language sent out to first user 403
The recognition result of sound call request information 404 is determined and is used belonging to 404 corresponding called device of institute's voice call request information
The information at family contacts the medical care that the called device owning user information is stored with the first movement medical care equipment 401
Roster matches;
The first movement medical care equipment 401 is additionally operable in the successful match, to the contact person of the successful match
Corresponding second Nurse Mobile equipment initiates calling;And for when matching unsuccessful, sending out prompt;
The second Nurse Mobile equipment is used to listen request information 407 based on the voice that second user is sent out, and activates institute
State the second medical care sound identification module;The second medical care sound identification module voice that the second user is sent out for identification
The information that listens request 407, and recognition result is exported with speech form;
The second Nurse Mobile equipment is additionally operable to, and is listened request information 407 based on the voice that the second user is sent out,
The voice established between the first movement medical care equipment 401 connects.
Preferably, a kind of medical wireless communication system based on speech recognition, first sound identification module or institute
The second sound identification module is stated, respectively by same speech recognition algorithm, identifies corresponding voice call request information 404 or voice
The information that listens request 407.
Preferably, a kind of medical wireless communication system based on speech recognition, the speech recognition algorithm are:
It obtains voice call request information 404 or voice listens request information 407;
Institute's voice call request information 404 or voice are listened request information 407, corresponding first text is converted to
Information;
The first keyword set is extracted from the text message according to preset rules, wherein first keyword set
It include at least one keyword in conjunction;
Calculate the similarity of first keyword set and the second keyword set, wherein second keyword set
It is combined into the keyword set that corresponding second text message extracts in presupposed information library, wherein the presupposed information library Zhong Bao
Containing multiple second text messages and the incidence relation for receiving group information.
The step of method described in conjunction with the examples disclosed in this document or algorithm, can use hardware, processor to execute
The combination of software module or the two is implemented.Software module can be placed in random access memory (RAM), memory, read-only memory
(ROM), electrically programmable ROM, electrically erasable ROM, register, hard disk, moveable magnetic disc, CD-ROM or technical field
In any other form of storage medium well known to interior.
Above-described specific implementation mode has carried out further the purpose of the present invention, technical solution and advantageous effect
It is described in detail, it should be understood that the foregoing is merely the specific implementation mode of the present invention, is not intended to limit the present invention
Protection domain, all within the spirits and principles of the present invention, any modification, equivalent substitution, improvement and etc. done should all include
Within protection scope of the present invention.
Claims (10)
1. a kind of medical wireless communication means based on speech recognition, which is characterized in that including:
The voice call request information that first movement medical care equipment is sent out based on the first user activates the first movement medical care to set
The first medical care sound identification module in standby;The first medical care sound identification module is for obtaining and identifying voice call request
Information, and recognition result is exported with speech form;
First medical care sound identification module of the first movement medical care equipment is exhaled according to the voice sent out to first user
The recognition result for being solicited message determines the information of the corresponding called device owning user of institute's voice call request information,
By the medical care Affiliates List progress of the called device owning user information and first movement medical care equipment storage
Match;
If successful match, the corresponding second movement doctor of contact person of first movement medical care equipment to the successful match
Equipment is protected, calling is initiated;
If matching is unsuccessful, the first movement medical care equipment sends out prompt;
The second Nurse Mobile equipment is listened request information based on the voice that second user is sent out, activation the second movement doctor
Protect the second medical care sound identification module in equipment;
Second medical care sound identification module of the second Nurse Mobile equipment identifies that the voice that the second user is sent out is answered
Solicited message, and recognition result is exported with speech form;
Listened request information based on the voice that the second Nurse Mobile equipment sends out the second user, establish with it is described
Voice connection between first movement medical care equipment.
2. a kind of medical wireless communication means based on speech recognition according to claim 1, it is characterised in that:Described
The sound identification module of the sound identification module of one Nurse Mobile equipment or the second Nurse Mobile equipment, respectively by same
One speech recognition algorithm identifies that corresponding voice call request information or voice listen request information.
3. a kind of medical wireless communication means based on speech recognition according to claim 2, it is characterised in that:Institute's predicate
Sound recognizer is:
Voice call request information or voice is obtained to listen request information;
Institute's voice call request information or voice are listened request information, corresponding first text message is converted to;
The first keyword set is extracted from the text message according to preset rules, wherein in first keyword set
Including at least one keyword;
Calculate the similarity of first keyword set and the second keyword set, wherein second keyword set is combined into
The keyword set that corresponding second text message extracts in presupposed information library, wherein comprising more in the presupposed information library
A second text message and the incidence relation for receiving group information.
4. a kind of medical wireless communication means based on speech recognition according to claim 3, it is characterised in that:By such as
Lower formula calculate separately the morphology similarity of the first text message A and the second text message B, the long similarity of sentence, word order similarity,
Distance conformability degree:
Morphology similarity:
The long similarity of sentence:
Word order similarity:
Distance conformability degree:
The similarity of the first text message A and the second text message B are calculated according to following formula:
Sim (A, B)=α1WordSim(A,B)+α2LenSim(A,B)+α3OrderSim(A,B)+α4DisSim(A,B)
Wherein, Len (A) is the number of keyword in the corresponding keyword sets of the first text message A, and Len (B) is the second text
The number of keyword in the corresponding keyword sets of information B;SameWc (A, B) is the corresponding keyword sets of the first text message A
Close the number of same keyword in keyword set corresponding with the second text message B;OnceWs (A, B) is indicated in the first text
All occur in information A and the second text message B and primary keyword set only occurs;| OnceWs (A, B) | for OnceWs (A,
B the number of keyword in) gathering, k be word distance values of the DisSim (A, B) of training arrived when being 0.5, and Max () is takes most
Big value function, Abs () are the function that takes absolute value, α1For the weight coefficient of morphology similarity WordSim (A, B), α2For sentence appearance
Like the weight coefficient of degree LenSim (A, B), α3For the similarity of word order similarity OrderSim (A, B), α4For Distance conformability degree
The weight coefficient of DisSim (A, B).
5. a kind of medical wireless communication means based on speech recognition according to claim 2, it is characterised in that:Described
After the voice call request that first movement medical care equipment is sent out based on the first user, and the first movement medical care is being activated to set
Before the first medical care sound identification module in standby;
And the voice sent out based on second user in the second Nurse Mobile equipment is listened request after information, and activating
Before the second medical care sound identification module in the second Nurse Mobile equipment;
The method includes:
It obtains institute's voice call request or the voice to listen request corresponding first voice signal, wherein described first
Voice signal is the voice signal that input carries noise in real time;
First voice signal is transformed into frequency domain, obtains corresponding second voice signal of frequency domain;
Logarithmic spectrum posteriori SNR is calculated according to second voice signal, the logarithmic spectrum posteriori SNR is present frame second
Ratio between the logarithm and the logarithm of former frame noise power estimation value of the power spectrum of voice signal;
Noise power spectrum estimated value is obtained according to the logarithmic spectrum posteriori SNR based on weighted noise estimation algorithm;
The gain function that Wiener filtering is generated according to the noise power spectrum estimated value, according to the gain function to described second
Voice signal is filtered, and generates third voice signal, wherein the third voice signal is the voice of removal noise in frequency domain
Signal;
The third voice signal is transformed into time domain, obtains the 4th voice signal, wherein the 4th voice signal as when
The voice signal of noise is removed in domain.
6. a kind of medical wireless communication means based on speech recognition according to claim 5, it is characterised in that:It is described right
Number spectrum check back signal ratio is calculated by following formula:
Wherein, POST_SNR (k, w) is the logarithmic spectrum posteriori SNR on second w-th of frequency point of voice signal kth frame, | y (k,
w)|2It is the estimation of noise power on kth w-th of frequency point of -1 frame for the power spectrum of the second voice signal on w-th of frequency point of kth frame
Value, a is constant.
7. a kind of medical wireless communication system based on speech recognition, it is characterised in that:Including first movement medical care equipment, second
Nurse Mobile equipment;
The voice call request information that the first movement medical care equipment is used to send out based on the first user, activation described first move
The first medical care sound identification module in dynamic equipment;
The first medical care sound identification module is exported with speech form and is known for obtaining and identifying voice call request information
Other result;
And the first medical care sound identification module of the first movement medical care equipment, it is additionally operable to basis and first user is sent out
Voice call request information recognition result, determine the corresponding called device owning user of institute's voice call request information
Information, by the called device owning user information and the first movement medical care equipment storage medical care Affiliates List
It is matched;
The first movement medical care equipment is additionally operable in successful match to the contact person of the successful match corresponding
Two Nurse Mobile equipment initiate calling;And unsuccessful prompt is being sent out matching;
The second Nurse Mobile equipment is used to listen request information based on the voice that second user is sent out, and activation described second moves
The second medical care sound identification module in dynamic medical care equipment;
The second medical care sound identification module voice that the second user is sent out for identification of the second Nurse Mobile equipment
Listen request information, and exports recognition result with speech form;
Second Nurse Mobile equipment is additionally operable to listen request information based on the voice for sending out the second user, establish with it is described
Voice connection between first movement medical care equipment.
8. a kind of medical wireless communication system based on speech recognition, which is characterized in that including:The system comprises first movements
Medical care equipment, the second Nurse Mobile equipment;Wherein, the first movement medical care equipment includes the first medical care sound identification module,
The second Nurse Mobile equipment includes the second medical care speech ciphering equipment module;
The voice call request information that the first movement medical care equipment is used to send out based on the first user, activation first doctor
Protect sound identification module;The first medical care sound identification module is used to obtain and identify voice call request information, and with language
Sound form exports recognition result, and for the identification knot according to the voice call request information sent out to first user
Fruit determines the information of the corresponding called device owning user of institute's voice call request information, by the called device institute
Belong to user information to be matched with the medical care Affiliates List that the first movement medical care equipment stores;
The first movement medical care equipment is additionally operable in the successful match, to the contact person of the successful match corresponding
Two Nurse Mobile equipment initiate calling;And for when matching unsuccessful, sending out prompt;
The second Nurse Mobile equipment is used to listen request information based on the voice that second user is sent out, activation second doctor
Protect sound identification module;The second medical care sound identification module voice that the second user is sent out for identification listens request
Information, and recognition result is exported with speech form;
The second Nurse Mobile equipment is additionally operable to, and is listened request information based on the voice that the second user is sent out, establish with
Voice connection between the first movement medical care equipment.
9. a kind of medical wireless communication system based on speech recognition according to claim 8, it is characterised in that:Described
One sound identification module or second sound identification module identify corresponding voice respectively by same speech recognition algorithm
Call request information or voice listen request information.
10. a kind of medical wireless communication system based on speech recognition according to claim 9, it is characterised in that:It is described
Speech recognition algorithm is:
Voice call request information or voice is obtained to listen request information;
Institute's voice call request information or voice are listened request information, corresponding first text message is converted to;
The first keyword set is extracted from the text message according to preset rules, wherein in first keyword set
Including at least one keyword;
Calculate the similarity of first keyword set and the second keyword set, wherein second keyword set is combined into
The keyword set that corresponding second text message extracts in presupposed information library, wherein comprising more in the presupposed information library
A second text message and the incidence relation for receiving group information.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810404846.XA CN108449479A (en) | 2018-04-28 | 2018-04-28 | A kind of medical wireless communication means and system based on speech recognition |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810404846.XA CN108449479A (en) | 2018-04-28 | 2018-04-28 | A kind of medical wireless communication means and system based on speech recognition |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108449479A true CN108449479A (en) | 2018-08-24 |
Family
ID=63202587
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810404846.XA Pending CN108449479A (en) | 2018-04-28 | 2018-04-28 | A kind of medical wireless communication means and system based on speech recognition |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108449479A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110536033A (en) * | 2019-08-01 | 2019-12-03 | 阿克索健康科技(广州)有限公司 | A kind of wisdom ward interphone extension system and control method based on speech recognition |
CN112398927A (en) * | 2020-11-04 | 2021-02-23 | 重庆品博科技有限公司 | Cloud platform network call processing system and method |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102131156A (en) * | 2011-01-27 | 2011-07-20 | 上海华美络信息技术有限公司 | Hospital instant communication system and instant communication method |
CN104103278A (en) * | 2013-04-02 | 2014-10-15 | 北京千橡网景科技发展有限公司 | Real time voice denoising method and device |
CN107273350A (en) * | 2017-05-16 | 2017-10-20 | 广东电网有限责任公司江门供电局 | A kind of information processing method and its device for realizing intelligent answer |
CN107895578A (en) * | 2017-11-15 | 2018-04-10 | 百度在线网络技术(北京)有限公司 | Voice interactive method and device |
-
2018
- 2018-04-28 CN CN201810404846.XA patent/CN108449479A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102131156A (en) * | 2011-01-27 | 2011-07-20 | 上海华美络信息技术有限公司 | Hospital instant communication system and instant communication method |
CN104103278A (en) * | 2013-04-02 | 2014-10-15 | 北京千橡网景科技发展有限公司 | Real time voice denoising method and device |
CN107273350A (en) * | 2017-05-16 | 2017-10-20 | 广东电网有限责任公司江门供电局 | A kind of information processing method and its device for realizing intelligent answer |
CN107895578A (en) * | 2017-11-15 | 2018-04-10 | 百度在线网络技术(北京)有限公司 | Voice interactive method and device |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110536033A (en) * | 2019-08-01 | 2019-12-03 | 阿克索健康科技(广州)有限公司 | A kind of wisdom ward interphone extension system and control method based on speech recognition |
CN112398927A (en) * | 2020-11-04 | 2021-02-23 | 重庆品博科技有限公司 | Cloud platform network call processing system and method |
CN112398927B (en) * | 2020-11-04 | 2022-10-14 | 重庆品博科技有限公司 | Cloud platform network call processing system and method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10554826B2 (en) | Method and apparatus for adjusting volume of user terminal, and terminal | |
US9769296B2 (en) | Techniques for voice controlling bluetooth headset | |
CN107240405B (en) | Sound box and alarm method | |
CN103888581B (en) | A kind of communication terminal and its method for recording call-information | |
CN108198569A (en) | A kind of audio-frequency processing method, device, equipment and readable storage medium storing program for executing | |
CN108962260A (en) | A kind of more human lives enable audio recognition method, system and storage medium | |
CN107945790A (en) | A kind of emotion identification method and emotion recognition system | |
CA2416592A1 (en) | Method and device for providing speech-to-text encoding and telephony service | |
CN101442579A (en) | Mobile terminal with speech recognition calling subscriber information | |
CN111683317B (en) | Prompting method and device applied to earphone, terminal and storage medium | |
CN113938785A (en) | Noise reduction processing method, device, equipment, earphone and storage medium | |
WO2021184837A1 (en) | Fraudulent call identification method and device, storage medium, and terminal | |
CN106850931A (en) | The method and mobile intelligent terminal of Barassment preventing telephone | |
CN107978320A (en) | One kind call method of adjustment, device, equipment and medium | |
CN108449479A (en) | A kind of medical wireless communication means and system based on speech recognition | |
CN104202485B (en) | A kind of safety call method, device and mobile terminal | |
CN108766416B (en) | Speech recognition method and related product | |
CN101354886A (en) | Apparatus for recognizing speech | |
CN112259112A (en) | Echo cancellation method combining voiceprint recognition and deep learning | |
CN107274895A (en) | A kind of speech recognition apparatus and method | |
CN113271430B (en) | Anti-interference method, system, equipment and storage medium in network video conference | |
CN116343756A (en) | Human voice transmission method, device, earphone, storage medium and program product | |
CN204117590U (en) | Voice collecting denoising device and voice quality assessment system | |
CN106598536A (en) | Record startup method and apparatus for electronic device, and electronic device | |
CN108537019A (en) | A kind of unlocking method and device, storage medium |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180824 |
|
RJ01 | Rejection of invention patent application after publication |